brain mind and the signifying body open linguistics

Brain, Mind, and the Signifying Body

Open Linguistics Series

Series EditorRobin Fawcett, University of Wales, Cardiff

This series is 'open' in two senses. First, it provides a forum for works associated with any schoolof linguistics or with none. Most practising linguists have long since outgrown the unhealthyassumption that theorizing about language should be left to those working in the generativistformalist paradigm. Today large and increasing numbers of scholars are seeking to understandthe nature oflanguage by exploring one or other of various cognitive models oflanguage, or interms of the communicative use of language, or both. This series is playing a valuable part in reestablishing the traditional 'openness' of the study of language. The series includes manystudies that are in, or on the borders of, various functional theories oflanguage, and especially(because it has been the most widely used of these) Systemic Functional Linguistics. Thegeneral trend of the series has been towards a functional view of language, but this simplyreflects the works that have been offered to date. The series continues to be open to allapproaches, including works in the generativist-formalist tradition.

The second way in which the series is 'open' is that it encourages studies that open out 'core'linguistics in various ways: to encompass discourse and the description of natural texts; toexplore the relationships between linguistics and its neighbouring disciplines - psychology,sociology, philosophy, cultural and literary studies - and to apply it in fields such as education,language pathology and law.

Recent titles in this seriesAnalysing Academic Writing, Paul]. ThibaultClassroom Discourse Analysis,Frances ChristieConstruingExperience through Meaning: A Language-based Approach to Cognition, M. A. K. Halliday

and Christian M. I. M. MatthiessenCulturallySpeaking: Managing RapportthroughTalk across Cultures, Helen Spencer-Oatey (ed.)Development of Language, Geoff Williams and Annabelie Lukin (eds)Educating Eve: The 'Language Instinct' Debate, Geoffrey SampsonEmpiricalLinguistics, Geoffrey SampsonGenre and Institutions:Social Processes in the Workplace and School; Frances Christie and J. R. Martin

(eds)The Intonation Systems of English, Paul TenchLanguage, Education and Discourse, Joseph A. Foley (ed.)Language Policy in Britain and France: The Processes of Policy, Dermis AgerLanguage Relations across Bering Strait: Reappraising the Archaeological and Linguistic Evidence,

Michael FortescueLearning throughLanguage in EarlyChildhood, Clare PainterMultimodalDiscourse Analysis, Kay L. O'Halloran (ed.)Pedagogy and the Shaping of Consciousness: Linguistic and Social Processes, Frances Christie (ed.)Register Analysis: Theory and Practice, Mohsen Ghadessy (ed.)Relations and Functions within and around Language, Peter H. Fries, Michael Cummings, David

Lockwood and William Spruiell (eds)Researching Language in Schools and Communities: FunctionalLinguistic Persepctives, Len Unsworth

(ed.)SummaryJustice: JudgesAddressJuries, Paul RobertshawSyntactic Analysis and Description: A Constructional Approach, David G. LockwoodThematicDevelopments in English Texts, Mohsen Ghadessy (ed.)Ways of Saying: Ways of Meaning. Selected Papers of RuqaiyaHasan, Carmen Cloran, David Butt and

Geoffrey Williams (eds)Words, Meaning and Vocabulary: An Introduction to Modern English Lexicology, HowardJackson and

Etienne Ze ArnvelaWorkingwith Discourse: Meaning beyond the Clause, J. R. Martin and David Rose

Brain, Mind, and theSignifying BodyAn Ecosocial Semiotic Theory

Paul J. Thibault

with a foreword by

M. A. K. Halliday

ContinuumThe Tower Building, 11 York Road, London SEl 7NX15 East 26th Street, New York NY 10010

www.continuumbooks.com

© Paul J. ThibauIt 2004Foreword © M. A. K Halliday 2004

All rights reserved. No part of this publication may be reproduced ortransmitted in any form or by any means, electronic or mechanical, includingphotocopying, recording, or any information storage or retrieval system,without prior permission in writing from the publishers.

First published 2004

British Library Cataloguing-in-Publication DataA catalogue record for this book is available from the British Library.

ISBN 0-8264-6965-5 (hardback)

Library of Congress Cataloging-in-Publication DataA catalogue record for this book is available from the British Library.

Typeset by Kenneth Burnley in Wirral, CheshirePrinted and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wilts

Contents

List of FiguresList of TablesForeword byM. A. K. HallidayPreface

PART I

IX

X

XI

xiii

Chapter 1: Introduction

The Body-brain System, Meaning-making Activity, and EcosocialEnvironment: Building a New, Unified Discourse 3

2 The Conceptual Framework of the Ecosocial Semiotic Perspective 83 Gibson's Ecological Theory of Perception and the Three-level Scalar

Hierarchy View of Organism-plus-Environment Transactions 114 The Brain as Regulator of Sensori-motor Activity: Implications for

Social Semiosis 185 Topological and Typological Modes of Semiotic-material

Interdependence 236 Contextualization and Meta-redundancy 267 A Critique of the Causal View of Brain-mind Relations 308 Defining and Extending the Notion of Meaning in Terms of the

Three-level Hierarchy and the Specification Hierarchy 349 The Signifying Body: Rethinking the Stratified view of Semiosis in

Terms of the Three-level Scalar Hierarchy 3910 Systemic-functional Linguistic Theory: Bringing Together the Intra-

organism and Inter-organism Perspectives on Meaning-making 4611 Reconnecting the Semiotic Concept of Value to the Body-brain System

and to Meaning-making Activity 4912 A Thumbnail Sketch of the Book 54

PART 11

Chapter 2: Sensori-motor Activity, Movement, and Social Meaning-making:Rethinking the Expression Plane of Semiosis

1 Some Early Signposts from Saussure and Hjelmslev: The ExpressionPlane as Embodied Articulatory Movement 59

2 Energy Exchange and the Complementarity of InteractingBody-brains 68

vi CONTENTS

3 The Stratified Model of Semiosis: The Problem of ConceptualAbstractness and Scalar Homogeneity 69

4 Blackboxing the Sensori-motor Dimension: Language Seen asModality-independent Centralized Processing Mechanism 72

5 The Intentional Character of 'Inner' and 'Outer' Sensori-motorActivity: Towards a Unified Account 75

6 The Symbolic Possibilities of Bodily Movement 777 Articulatory Movement Seen as Actional Semiotic, Not Physical

Behaviour 788 Inner and Outer Body States and Social Semiosis 819 The Semiotic Mediation and Entraining of Embodied Bio-kinematic

Potential 8310 Metafunctional Diversity on the Expression Plane 8611 The Expression Plane is the Interface between Body and Ecosocial

Environment 8712 The Metafunctional Basis of Vocal-tract Articulatory Activity 9013 Subjectivity, Agency, and the Prosodic Realization of Interpersonal

Meaning 9414 Vocal-tract Gestures and Grammar: Symptom and/or

Supervenience? 9815 The Intentional Character of Articulatory Activity 10016 Embodying the Metafunctions: The Example of Vocal-tract

Articulatory Activity 10317 The Metafunctional Basis of Space and of Bodily Movement in

Ecosocial Space-time 105

Chapter 3: Body Dynamics, Meaning-making, and Scale Heterogeneity:Expression and Content as Cross-scalar Semiotic Processes Embeddingthe Body-brain in its Ecosocial Environment

The Dynamical Character of Expression and Content and the Cross-coupling of Diverse Scalar Levels of Semiotic Organization 108

2 The Expression Stratum and the Principle of Alternation 1093 Stratification in Relation to Expression and Content 1164 The Integration of Iconic, Indexical and Symbolic Modes of

Meaning in Phonology 1185 Rhythm and the Foot 1206 The Metafunctional Organization of Phonology as seen from

the Perspective of Rhythm and the Foot 1227 The Integration of Iconic, Indexical and Symbolic Modes of

Meaning in Lexicogrammar 1268 Language as Particle, Wave, and Field 1349 The Brain as Selective Recognition System: Language and Edelrnan's

Theory of Recategorization 13910 The Integration oflndividuals to their Semiotic Trajectories 14611 The Intrinsically Time-bound Nature of Semiosis: The Integration of

Temporal Dynamics Across Scales of Expression and Content 148

CONTENTS vii

12 Scalar Heterogeneity and the Phonological and LexicogrammaticalRank Scales 154

13 The Emergence of Linguistic Categories from the Child's PrimaryForceful Interactions with its Environment 162

Part III

Chapter 4: The Semiotic Basis of Consciousness

1 First-person and Third-person Accounts of Consciousness 1712 The Representation of Subjective Experience in Consciousness in

Relation to the Higher-scalar Environment of the Individual 1733 Locating the Seat of Consciousness 1764 The Meaning-making Capacity of the Body-brain Complex through

the Discrimination of Difference 1845 Language Functions and the Cortical Organization of the Brain:

Implications for Higher-order Consciousness 1896 Experiential Meaning and the Assimilation of the Phenomena of

Experience to Knowable Categories 1957 Interpersonal Meaning as Exploratory and Orienting Activity in

Relation to the Ground 1988 The Textual Metafunction as Semiotic Means for Giving Unity and

Wholeness to Meaning-making 2009 Experiential and Interpersonal Meaning in Gaze 201

10 Proto-interpersonal Meaning and the Child's Exploration of itsEnvironment 202

11 Bogdan's Theory of Mental Sharing and Topical Predication 20512 Interpersonal Meaning, Value, and Action 20913 Procedural Knowledge, Declarative Knowledge and the Semiotic

Spiral towards Symbolic Consciousness 21214 Interpersonal Meaning, Goal-seeking Activity, and the Goal

Hierarchy 21515 Consciousness and Semiotic Stratification 21716 An Alternative Reading of Descartes in the Internalist Perspective

of Interpretive Activity 22417 The Entropic Character of Meaning 22618 Consciousness as the Contextualization of Experience in the

Perspective of the Self 22719 The Embedding of Consciousness in a Higher-scalar System of

Interpretance 231

Chapter 5: The Metafunctional Character of Consciousness: SomeCorrelations Between the Neurobiological and Semiotic Dimensions

1 The Contextual Character of Consciousness 2362 The Metafunctions and the Shape of Consciousness 2383 Vague Contours of the Metafunctions in the Infant's Early

Perceptual-motor Engagements with the Environment 241

VIII CONTENTS

4 The Structure of (Self)Consciousness in Perceptual Awareness 2465 Damasio's Neurobiological Theory of Consciousness 2486 The Proto-metafunctional Structure of Core and Extended

Consciousness 2577 Integrating the Interaction System and the Meaning System

Perspectives on (Self) Consciousness 2648 Minding the Gap between Minds: Mirror Neurons and Interpersonal

Meaning 2679 Inner Speech as Linguistically Realized Higher-order Thinking 271

10 The Metafunctional Character of Inner Speech as LinguisticallyConstituted Thought 272

11 The Re-grounding of the Perspectives of Self and Other inSymbolic Consciousness 276

12 Text, Social Meaning-making Practices and Higher-orderConsciousness 277

Chapter 6: Brain, Meaning, and Consciousness

1 Biological and Socio-cultural Factors Form a Single System ofComplexly Related and Interacting Factors: Putting Time andActivity Back into the Picture 281

2 Re-interpreting Flohr's Brain-based Theory of PhenomenalAwareness: A Three-level Hierarchy View of the Emergence ofProto-meaning in the Brain 283

3 Brain, Meaning, and Symbolic Consciousness 2894 The Embodiment of the Material and the Conscious Modes in

Expression and Content 2905 Periodicity and the Intrinsic Temporal Organization of the

Expression Stratum of Speech on Diverse Scalar Levels 2956 The Creation of Symbolic Objects of Consciousness in Semantic

Neural Space 2967 Contextualizing Relations, the Principle of Meta-redundancy, and

the Brain as Contextualizing Tool 3008 A Definition of Consciousness in Terms of the Semiosis that

Occurs Within the Brain 3029 The Three-level Hierarchy, Neuronal Activity, and the Emergence

of the Self-perspective 310

Epilogue 314References 318Name Index 331Subject Index 334

List of Figures

2.1 Example of gestural scores for /p<en/ ('pan') and /b<en/ ('ban');borrowed from Browman and Goldstein (1995: 189) 91

3.1 Syllable structure of sail, showing moraic and non-moraic elements 1133.2 Trajectory of wave of stressed and unstressed syllables in clause

complex, showing alternation of stressed and unstressed syllables assyntagm unfolds in time 114

3.3 Trajectory of tonicity wave in clause complex 1143.4 Rhythmic field of the word sailed, showing Ictus and Remiss phases

of its temporal trajectory 1233.5 Dependency relations between Ictus and Remiss phases in never 1243.6 Thematicity wave in the English clause complex, showing Theme

and Rheme as two peaks of informational quanta 1293.7 Multiple modalizing fields acting on the syntagmatic domain of a

proposition, showing the scope over which each field's modalizedinfluence extends 131

3.8 Simplified determiner system, showing the categorial distinctionbetween the values 'specific' and 'non-specific', as symbolized byEnglish theand a 133

3.9 Expression stratum as combinatorial hierarchy, showingreorganization of higher levels as recouplings of elements onlower levels thereby allowing for the emergence of new systemicpossibilities and, hence, new system-environment couplings 143

3.10 The integration of initial at-oneness with the world to the expansionof information-meaning at more specified levels of semiosis alongan individuating trajectory 149

3.11 Linguistic categorization and its emergence from pre-linguisticforceful interactions between body and environment 168

4.1 Integration-cum-presupposition hierarchy of iconic, indexical, andsymbolic modes of grounding 211

4.2 Classification of input and output impulses involved in the symbolictransduction of stimulus information and sensori-motor activity insemiosis (adapted and modified from Gibson 1983 [1966]: 46) 219

6.1 Reentrant loop of conscious experience, showing relationshipbetween self and object of experience 288

6.2 Content and expression and their relations to the central andperipheral nervous systems; adapted from Peng (1994) 294

6.3 Open-ended hierarchy of semiotic triplicates of levels showing theintegration of perceptual, conceptual, and symbolic phenomenaof conscious experience 307

List of Tables

2.1 The robustness of the agent and its interfacing with the expressionand content strata of language 64

2.2 Stratified model of language, showing vertical hierarchy of differentlevels of abstraction 69

2.3 Metafunctional analysis of exchange unit: phonological,lexicogrammatical and discourse semantic strata; tonic segmentsin upper case 96

3.1 The expression stratum of spoken language in relation to thethree-level hierarchy 110

3.2 Semiotic properties of icon, index, and symbol 1353.3 The metafunctional organization of both phonology and

lexicogrammar on the expression and content strata illustratingthe operation of mixed-mode semiosis, combining both discretetypological-categorial distinctions and continuous topologicalvariation on both strata 152

3.4 The content stratum of language in relation to the three-levelhierarchy 158

3.5 Emergence of embodied category formation and its differentiationinto 'learning about' and 'acting on' through forceful interactionwith environment 166

4.1 Lateralization of language functions in the left and righthemispheres according to Deacon 191

4.2 A comparison of four accounts of language form and function incontext, showing an emerging consensus concerning themultifunctional nature of language and its contextual motivation 194

4.3 Proto-experiential meaning in gaze vector 2014.4 Proto-interpersonal meaning in gaze vector, showing ground

functions 2024.5 Multimodal eo-deployment of perceptual-semiotic resources in the

exchange between the child and his mother in Halliday's example 2084.6 Types of knowledge and stages of semiotic development in early

infant semiosis 2136.1 Metafunctional analysis of the imperative clause (you) look at the mess 2986.2 Scalar hierarchy of anatomical structures of perception and relative

functions 304

Foreword

It is a privilege to be invited to introduce a work of this range and importance.Paul Thibault's book is appearing at a time when the disciplinary bordersinherited from the previous century - no longer felt as enabling, but rather asconstraining, as boundaries rather than borders - are tending to fuzz out anddisappear; and new strategies of thought, new dimensions of knowledge areemerging. This book makes a significant contribution to the ongoing dialoguearound these critical themes.

One feature that contributes to the strength and effectiveness of PaulThibault's treatment of his topic is the way he engages with the work of leadingscholars of the past hundred years whose ideas can be seen as having in somesense anticipated the directions of change. In this respect he is following up hisown earlier (1997) study of Ferdinand de Saussure, the scholar usually claimedas 'the father of modern linguistics' and certainly the linguist most frequentlyrevisited and commented on during the entire twentieth-century period. In thepresent book, Paul Thibault discusses the theoretical contributions of Saussure'smost important successor, Louis Hjelmslev, as well as those of other scholars suchas Karl Biihler and Gregory Bateson who helped to shape the pattern ofknowledge for their own and the succeeding time. By picking up on their workPaul Thibault provides an intellectual context for interpreting the theoreticaladvances made by contemporary scholars who, in their turn, are redefining theparameters of our own thinking - people such as Gerald Edelman, Jay Lemke,Antonio Damasio and Terrence Deacon.

The central organizing concept running through the book is that of meaning,or semiosis - the realm of our existence that is distinct from, and complementary to, the material realm. Paul Thibault theorizes meaning as 'trajectories'taking place in time, in the form of activities on a number of different scalarlevels from neuronal activity in the individual brain through to movement acrosshuman populations; but manifested critically in the social meaning-makingpractices of the members of human groups. Here too there is an echo of aprevious motif in his work, that of 'Social Semiotics as Praxis' (1991) developedagainst the background of theory of narrative. All such activities are grounded inbiological processes: 'meaning-making is to be explained in terms which areconsistent with what we understand about the biological basis of semiosis'(present volume p. 24) - but this is not to say that the whole of semiosis can beexplained by reduction to biology.

Xll FOREWORD

The most powerful manifestation of semiosis is human language; and in thisbook Paul Thibault's central concern is the interpretation of language itself as aself-organizing complex system. His concept of the 'signifying body' encapsulates the notion of the human body as the locus where meaning is made, throughthe operation of the brain at the critical interfacing of the semiotic with thematerial - the content plane with the ecosocial environment, the expressionplane with the physiological resources of the production and reception ofspeech. Thus all linguistic processes are grounded in processes of the body. Butlanguage cannot be reduced to bodily processes, and much of the discussioncentres on the organization of language as it appears in depth with this dualperspective.

Paul Thibault defines his starting point in these terms: ' ... the intrinsic organization oflanguage has evolved in the species (phylogenesis) and develops in theindividual (ontogenesis) so that it cross-eouples both with the biological makeupof the body-brain and with the socio-cultural organization of our ecosocial environment in ways that closely relate to the kinds of social activities that humanbeings perform' (present volume p. 48). To me as a linguist perhaps the mostsignificant feature of Paul Thibault's many-faceted approach is the way he buildsup the intellectual context both for language and for linguistics, language in itsrelation to the human condition, linguistics in its relation to human knowledge.In a sense these are two aspects of the same contextualizing process, sincelanguage figures as the centrepiece of semiosis, or meaning-making activity, andall construction of knowledge is the making of meaning.

Paul Thibault's project is not so much interdisciplinary as transdisciplinary,coaligning the resources for thinking about his topic along a new thematicdiscussion. This is the kind of thinking that is needed so that new questions canbe asked wherever there are problems to be solved.

M. A. K. HALLIDAY

Hong KongNovember 2003

Preface

The writing of this book has arisen out of the conviction that there is an urgentneed for a materialist ecosocial semiotics which is able to reconnect body-brainprocesses and interactions both to the social and cultural practices which directlyact upon and affect human bodies, as well as to the ways in which bodily andbrain processes directly participate in and are a constitutively inseparable part ofour meaning-making activity.

This book is concerned with the role of the body-brain complex in our socialmeaning-making practices. In recent years, the role of the body and the brain inour social meaning-making practices has been a source of considerable interestand discussion. However, it is my contention that these discussions do not moveour understanding of the body's central role in meaning-making beyond thediscourse- and language-centred models of textual 'representation' whichcontinue to inform most accounts of the constitutive role of the body in socialsemiosis. A different orientation is called for. This orientation is what I refer to,following Jay Lemke, as an ecosocial semiotics of human meaning-makingactivity.

Like human meaning-making activity itself, this book is very much a hybridphenomenon. This reflects my conviction that the science of human meaningmaking activity is necessarily a transdisciplinary theory and praxis. Whileworking on the lectures of the course 'Saussure and Beyond' that I wrote forthe Cyber Semiotics Institute between 1996 and 1998 (WWW site: http:/ /www.chass.utoronto.ca/epc/srb). it became increasingly clear to me that manyof the fundamental questions concerning ecosocial semiotic theory can only beanswered by engaging in a dialogue with the foundational concepts andquestions of both the social and the life sciences. In some ways, my web course'Saussure and Beyond' is a kind of Prolegomenon to the present study.

In the past two decades or so, the dialogue between the human and lifesciences has been given new direction and impetus by the development of thetheory of complex adaptive dynamic open systems. Such systems are characterized by the fact that their component parts give rise to newly emergent levels oforganization that are not reducible to the sum of their lower-level components.Instead, the higher, global levels take on dynamical characteristics that have constraining effects on the lower, local levels. Furthermore, dynamic open systemsbring time and history to the centre of the theoretical enquiry. Time, irreversibility, the embedding of systems in higher-scalar environments, and individuation, are all important features of this new way of understanding the world,including the part we play in it. Thus, the prediction and control of phenomena

PREFACE xiv

and the universalizing principles and modes of explanation in the natural andsocial sciences that have been the hallmarks of the'classical' approach to sciencein the past three or four centuries no longer hold sway in this new view of theworld and our place in it. Meaning, the interpretation of meaning across manydifferent scalar levels, and our being necessarily inside such systems of interpretance, will be seen to play a key role in this perspective.

The new perspectives that the theory of complex dynamic systems makesavailable to the science of human meaning-making activity allow us to ask newquestions. What are the foundational concepts of this approach? What is therelationship between the physical-biological and the social-cultural dimensionsof our being? What does this approach entail for the relationship between theexpression and content strata of semiotic systems? Does meaning originate in'mind' or 'brain', or is it the result of complex, time-bound semiotic trajectoriesthat loop between the individual's intrinsic dynamics and the ecosocial environment in which the individual is embedded? What is an ecosocial environment?How does the body-brain complex relate to meaning-making activity? How doesconsciousness relate to meaning-making activity? What is the status of thediscourse- and language-centred models of textual 'representation' in relation tothe centrality of activity? How do the material and the semiotic dimensions ofmeaning-making activity combine with each other? What is the political significance of an ecosocial semiotic theory?

It is only through the attempt to understand the constitutive inseparability ofthe semiotic-discursive and physical-material cross-couplings and dynamics thatwe can adequately theorize our and others' embodiments, our subjectiveexperience of our ecosocial environments, our perceptions of our inner statesand sensations and the meanings we attribute to these both in our internaldialogues in 'inner' speech activity, as well as in our interactions with others, andthe materiality of the body as playing a central, not marginal, role in socialmeaning-making activity.

In this volume, I have sought to engage with the relevant perspectives andtheoretical insights of a range of disciplines - e.g. linguistics, semiotics, biology,and psychology - in order to understand their relevance to the ecosocialsemiotic framework of this book and to reconstitute their perspectives andinsights within that framework. In doing so, I have tried to organize the diverseinsights and theoretical perspectives that I engage with in this book into acoherent set of principles and analytical tools that will serve to extend theecosocial semiotic framework into new areas of enquiry and practice.

A second book, which I have completed and which will appear after thepresent one, will focus more specifically on the cultural, social, interpersonal anddevelopmental aspects of the ways in which body-brain systems are cross-coupledto their ecosocial semiotic environments. In many ways, the arguments of the twobooks are complementary in terms of the overall ecosocial semiotic perspective.

I am grateful to Professor Gugliemo Cinque, former Dean of the Facolta diLingue e Letterature Straniere of the University of Venice, and to Professor PaoloBalboni, then Director of the Dipartimento di Scienze del Linguaggio of the sameUniversity. Both Gugliemo and Paolo generously supported my application forstudy leave from the University of Venice during 2002, as well as facilitating itspassage through the University bureaucracy. This leave was crucial in allowingme to complete this project.

PREFACE xv

This book was completed in August 2002 during my stay as a Visiting Professorin the Department of English in Lingnan University in Hong Kong. I wish toexpress my appreciation and gratitude to my colleagues at Lingnan Universityand especially to the Head of the Department of English, Professor Barry Asker,for providing so much support and assistance during the period of my stay atLingnan University and also for making my stay there such an enjoyable andrewarding one. Wendy Wong, the departmental secretary, was a constant sourceof invaluable secretarial and practical assistance.

I have been privileged to be able to present some of the arguments developedin this book to colleagues and students in universities and research institutes inthe following countries: Australia, Belgium, Canada, China, Hong Kong, Italy,Japan, Norway, San Marino, Singapore, Spain, the United Kingdom and theUnited States. It is impossible to thank all ofyou by name, but I am most gratefulto all of you for providing a forum for my ideas, as well as for your criticaldiscussion and input.

I also wish to express my gratitude to the following individuals with whom Ihave had discussions, or whose advice, support and friendship have beencritically important for the completion of this project: John Alexander, FelipeAlcantara Iglesias, Anthony Baldry, Marco Battacchi, Olga Battacchi,Jim Benson,Paul Bouissac, David Butt, Ray Cattell, Magda Cortelli, Stephen Cowley, MichaelCummings, Fan Dai, Kristin Davidse, Ersu Ding, Paolo Fabbri, Robin Fawcett,Gillian Francis, Peter Fries, Nancy Fries, Andrew Goatly, Bill Greaves, MichaelGregory, Morag Harris, Mike Ingham, Marcel Kinsbourne, Lisa Leung, MarcLorrimar, Bob Lumsden, Eva Maagere, Bill McGregor, Blair McKenzie, Ng LaiPing, Kay O'Halloran, Carlo Prevignano, Helen Price, Duane Savage-Rumbaugh,Susan Savage-Rumbaugh, Stuart Shanker,Jared Tagliatela, Godfey Tanner, CarolTaylor Torsello, Gordon Tucker, Eija Ventola, and David Wallace.

I owe a special debt of gratitude - both intellectual and personal- to MichaelHalliday, Ruqaiya Hasan and Jay Lemke for the critically important dialoguesthat I have had with them on the subject matter of this book over many years.Without their constant encouragement and support, the development of theideas in this book would have been considerably poorer.

I also wish to thank my mother, Sylvia, my father, Philip, my brother, Mark,and my sister, Linda, for believing in me and helping to make it all happen.

To my daughter, Ilaria, my profound appreciation and gratitude for herunderstanding and love.

I have dedicated this book to the memory of Morag Harris. My deep sense ofloss at her passing cannot adequately be expressed in words.

Finally, to Ordy, for filling my life with her radiance and love.

PAUL J. THIBAULT

BolognaOctober 2003

In memory of

Morag Anne Harris

17 November 1954-11 December 2000

Part I

This page intentionally left blank

1 Introduction

1. The Body-brain System, Meaning-making Activity, and EcosocialEnvironment: Building a New, Unified Discourse

This book is concerned with the relations between the body-brain system,meaning-making activity, and the ecosocial environment in which the former areembedded. An increasing number of studies in recent years show the inadequacies of the view that the brain is the seat of disembodied mental processes, orthat cognitive and meaning-making activities stop at the skin of the individualorganism. Moreover, the idea that the 'inner' mental life of the individual can bedisjoined from our meaning-making (semiotic) activity itself needs to be reexamined. Traditionally, Western cultural traditions have evolved two radicallydistinct discourse traditions for talking about notions such as mind, cognition,and the mental, on the one hand, and society, communication, and the semiotic,on the other (see Gee 1992; Lemke 1995a: 81-82; Manghi 1988; Walkerdine1997). Furthermore, the material basis of both 'inner' cognitive processes and'outer' meaning-making, or semiotic, activity has been disjoined from ourdiscourses about the mental (e.g. thinking) and meaning (e.g. discourse,language). Another problem stems from the way in which semiosis is narrowlydefined in terms of linguistically informed notions of text. One negative consequence of this restriction is the failure to see not only the continuities that linkhuman semiosis with the semiosis of other species, but also the continuities thatexist between perception, non-linguistic conceptual thinking, mental imaging,consciousness, and semiosis.

A second negative consequence lies in the way in which linguistic models oftext have been used as models of embodied meaning-making activity - activitywhich is embedded in its ecosocial environment. Textual products and recordsare produced and/or used in different kinds of social activities (see Lemke1984a: 78-80; Thibault 1991a: chapters 2-4,1994, 2003a: 58). The more fundamental notion in the theoretical framework of the present study is that of activity.Texts are integral parts of activities in particular ecosocial contexts, yet they havea secondary and derived status with respect to the activities in which they aremade and in which they participate. Activity is primary in the present framework(see also Goodwin 1996a, 1996b; Lave 1997).

Activity extends beyond the individual into the ecosocial environment.Meaning-making activity is a trajectory-in-time. In saying this, I am adapting thenotion of an ontogenetic trajectory in the work of Salthe (1993: 181-5). The locus

4 BRAIN, MIND, AND THE SIGNIFYING BODY

of meaning is not the organism. It cannot be reduced to the semiotic forms thatare eo-deployed in a given meaning-making activity or their physical-materialsubstrate. Nor is it the object text that may result from this activity. Rather, thelocus of meaning is the trajectory. It is useful to consider meaning-makingactivity as a semiogenetic trajectory that reaches back in time as well as forwardsinto the future. The relevant viewpoint here is that of the selves who jointlyengage in such activities and the perspectives that they implicate. Furthermore,semiogenetic trajectories can be seen as embedded in much larger-scalehistorical-biographical (ontogenetic) and ecosocial-eultural trajectories at thesame time as they integrate the much smaller scales of the neural and sensormotor dynamics of the body-brain to their own larger-scale dynamics andtimescales. A trajectory is a persistence-in-time that arises through the organization of processes. A meaning-making, or semiogenetic, trajectory is, therefore,an organization of semiotic processes that persist on some timescale. Theexistence of the trajectory on its level is dependent on both higher-scalarcontextual constraints and lower-scalar affordances, or enabling conditions, aswell as on meaning which is stored within the system (e.g. the biologicalorganism which supports the trajectory) as biological initiating conditions(Salthe 1993: 107).

A semiogenetic, or meaning-making, trajectory does not require preestablished plans or rules locked within the central nervous system of theorganism that govern, cause, or otherwise specify from the outset what meaningscan be made and how. All that is required to get things started is a motivation orintention which provides the initial impetus. The self-organizing processes thatenable a semiogenetic trajectory to emerge in time during the course of theactivity will do the rest. That is, the reentrantly correlated neural groups that arebuilt up from diverse sources of information both within the brain, in theenvironment of the agent, the agent's past history of interactions, and so on,constitute attractors. These attractors interact with the motivation or intention(seeJuarrero 1999: 199) in ways which open up possibilities in the individual'sneurally organized semantic state space. Attractors, by definition, enact constrained pathways or trajectories through this state space. Given the recursivenature of cerebral organization as a network of network of networks ... etc., asdiscussed by Kinsbourne (2001: 244-5; see below), more specific networks arenested within more schematic ones. A vague motivation or intention to dosomething, as Juarrero (1999: 199) shows, is embodied in a very broad, shallowattractor.

A broad or shallow attractor allows for very many lower-level possibilities forits realization in motor behaviour, depending on which sub-networks of neuralpatterns are activated. If, for example, a linguistic strategy for realizing theintention is activated - e.g. an imperative utterance directing someone to bringabout the desired state of affairs - rather than, say, a strategy based on directphysical action such as reaching for the desired object and grasping it in one'shand, then some neural pathways and motor processes rather than others will beactivated. These patterns of activation do not require conscious planning ordeliberation on the part of the agent. Being an agent does not entail this. Rather,being an agent depends on the overall context of the activity: the activation of

INTRODUCTION 5

one pathway rather than some other sets up particular patterns of constraints onvarious levels of organization in the overall context as it unfolds in time. Theagent's initiating conditions include a body-brain habitus of dispositions and waysof orienting to and being 'in' particular kinds of situations that the agent hasaccumulated over time, how the agent feels about them, the agent's perceptualmotor abilities at a given point in time, and the categories - perceptual-motor,conceptual, and semiotic - that the agent has established in his or her neuraldynamics on the basis of previous engagements with the environment.

Once a particular trajectory has been activated through the agent's neuraldynamics such that a given motor process - the neuromuscular activity of vocaltract articulation - is entrained to the patterns of activation of the agent's neuraldynamics, then higher-scalar contextual constraints come into play. Contextsensitive semiotic constraints provide global principles of organization wherebythe action-trajectory is given a determinate organization and is modulated byboth higher-scalar ecosocial constraints and lower-scalar (e.g. biological)enabling conditions as it unfolds in time towards its completion. The discoursegenres and social activity-structure types which are characteristic of a particularsocial network are context-sensitive constraints in this sense. They are not fixedrecipes for action, but systems of higher-order contextual constraints whichcross-couple and entrain semiotic-discursive and physical-material processes totheir principles of organization (see Andersen 2000). In so doing, they skew theprobability distributions of the semiotic-discursive selections and their crosscoupling to physical-material, including bodily, processes, such that someoutcomes are preferred and hence are more likely than others. It is in this waythat meaning-making activity is given a determinate shape and vectoriality as aninstance of an act of a given type.

Rather than relying on an intuitive and common-sense appeal to intentions,wants, desires, and so on, as internal mental states that are prior to selforganizing meaning-making activity, intentions are emergent properties of thisactivity. Agents, in their meaning-making activity, attribute intentions tothemselves and to others as a way of regulating and interpreting the contributions of each other to the eo-construction of the activity itself. In this view,intentions, rather than belonging to an epistemically private inner mental realm,are emergent discourse constraints and meanings that agents deploy in order,dialogically, to co-ordinate their respective contributions to the evolvingdiscourse as it unfolds in real-time. The semiogenetic trajectory, therefore,affords the integration of the body-brain's material interactivity with its here-nowenvironment with past events and occasions of interacting with others, as well aswith anticipated possible future outcomes. For example, a near-six-month-oldchild's high-pitched squeak is contextually integrated with the perceptual pickup of information about an environmental event (the scattering of somepigeons), the child's turning towards and attending to this event, as well as theparent's response to the child's vocalization (Halliday 1993: 95-6; see chapter 4,section 10, pp. 202-5, for discussion). The child's vocalization is itself an environmental event, which the mother both hears and interprets. She interprets thevocalization as a meaningful act that requires a response from her. At the sametime, the child's high-pitched squeak is contextually integrated with previous


experiences of the child's vocalizations and the parents' history of interpretingthem in context-specific ways.

The notion of the meaning-making (semiogenetic) trajectory thus provides away of conceptualizing how body-brains contextually integrate informationderiving from different perceptual modalities (e.g. seeing, hearing, moving) inthe here-now of their own and others' bodily activities to events, activities, and soon, on other space-time scales, both in the actual past and in the anticipatedfuture, involving either the same or other participants (see Thibault 2000a:303-6; 2003b). It is in this way that infants are integrated, through their activeparticipation in such events, to the higher-scalar ecosocial processes anddynamics where semiotic entities such as, for example, 'wants', 'beliefs','intentions' and 'desires' are recognized and used as the basis both for adjustingto and accounting for the behaviour of others qua selves to whom we attributesuch things as wants, beliefs, intentions, and desires (see Matthiessen 1993;Thibault 1993; see also Zelazo 1999).

On this view, intentions are not, as I pointed out above, epistemically privatemental states that belong to an ontologically distinct domain of 'mind' with independent causal powers. Instead, they are meaning-making resources that belongto the higher-scalar system of in terpretance of some ecosocial semiotic system. Assuch, agents use them for the purposes of interpreting and orienting to eachother in discourse, in the process adjusting their own activity on the basis of thebelief in our culture in the notion that persons are in possession of intentionsqua epistemically private mental states that can be read off behaviour as evidencefor the existence of a private mental realm that language simply serves to reveal.This account is essentially a realist one: intentions have a separate existence inthe mind with respect to the signs that refer to them. In my account, intentions,desires, wants, and so on are semiotic categories that agents appropriate fromthe interpersonal moral orders in which their actions are recognized,understood, and accounted for (Harre 1983). As such, they are not independentof the semiotic resources that we use in discourse to construe and enact them.They function in meaning-making activity to interpret and orient our own andothers' activities.

Semiosis is also a physical-material phenomenon. That is, semiosis is enactedand materially supported by the biologically based physical interactions thatoccur between individual organisms and their physical-material environments.From this point of view, semiosis is integrated with the material processes of ourbodies and brain in interaction with their external environments. Semioticactivity both originates in physical-material (e.g. biological) interactions and, atthe same time, has material effects on the environment, including other individuals and their bodies.

In particular, meaning-making activity, or semiosis, is cross-coupled tophysical-material processes which constitute the substrate of semiosis. By thesame token, the activity is guided and modulated along its trajectory by higherscalar semiotic constraints that arise in the symbolic (e.g. semantic) neuralspace of the individual until the activity's completion. Semiotically mediatedactivity (meaning-making activity) always has both a semiotic-discursive and aphysical-material dimension (see also Hasan 1999: 245, who points out that

INTRODUCTION 7

context can 'never be dissociated from the material and institutional aspects ofa culture'). Meaning-making is a semiotic-discursive phenomenon in the sensethat its dynamical processes enact meaningful patterns and relations thatintegrate the particular occasion or situation to the systems of semantic andother semiotic categories that the members of a given community or socialnetwork recognize and interpret as meaningful. Furthermore, these categoriesserve to link the particular event and the particular individuals who participatein that event to other activities, other practices, other individuals and social institutions on diverse space-time scales that go beyond the particular here-nowevent. These linkages are possible because the relevant systems of categorieshave evolved in historical time and have been adapted to the diverse needs andpurposes of the individuals who constitute a particular community or socialnetwork.

Those who argue in favour of formalist models of language maintain thatlanguage is autonomous with respect to both its physical-material basis in thebody-brain complex of the individual organism and the wider social and culturalpractices and meaning systems that characterize and constitute a given humancommunity. In this book, I shall emphasize the continuity between semioticprocesses and our other embodied resources for integrating organisms to theirenvironments. Thus, perception, non-linguistic conceptual thinking, reasoning,sensori-motor exploration, memory and many forms of pre- or non-linguisticconsciousness can be seen to be components of a much larger, phylogeneticallyolder, complex of resources whereby organisms are integrated with and engagein meaningful transactions with their inner and outer environments on diversespace-time scales. Language and other semiotic resources (e.g. depiction,movement, and gesture) are not discontinuous with, or even necessarily qualitatively distinct from, these other systems in all respects.

On the other hand, this does not mean that there are no differences betweenthese older cognitive resources and social semiotic ones. Instead, our symbolicresources for making meanings with others in social contexts have evolved aspart of this larger, older complex. The symbolic categories of language and othersemiotic systems can be seen to be the further specification of these prior, noncultural systems. This position is not dissimilar to the recent arguments advancedby Hauser, Chomsky, and Fitch (2002) that the 'language faculty in the broadsense (FLB) , is shared with other species, whereas the 'language faculty in thenarrow sense (FLN) , 'may be unique to humans' (Hauser et al.: 1578). Asidefrom the question of recursion, which these authors argue belongs to the FLN,their position argues for considerable continuity between animal communication and human language with respect to the FLB.

I shall discuss in more detail later in this book the reasons why the mentalistdiscourse for talking about notions such as 'cognition', 'mental representation',and 'internal symbol manipulation' (Schank and Abelson 1977), cannot serve asa useful or adequate conceptual framework for understanding the ways in whichindividuals are integrated to their material and semiotic environments on manydifferent scalar levels of organization. The mentalist discourse starts with theassumption of an individual mind which must solve problems and so on, in anexternal environment 'out there'. Early cognitive science focused on the internal


mechanisms of cognition, especially on the role of the information-processingstrategies, the mental representations, and the computations performed onthese representations. There was little or no interest in the neural processes thatformed the physical substrate of these cognitions. The prevailing view was thatneural processes merely provided the physical medium through which cognitionwas implemented. Cognition was seen as independent of the physical mediumand its workings. This view was based on the models and metaphors of humancognition that the digital computer both inspired and made available. However,the body-brain complex, unlike the digital computer, is a dynamic open systemwhich is both sensitive to and constrained by its contexts. In this book, I take theview that the biology of the body-brain complex cannot be separated from thestructure and function of meaning-making activity. Furthermore, the discourseof mind has remained silent about the larger-scale ecosocial systems in whichindividuals and their activities are embedded and which mediate and link theactivities of individuals together as component subsystems which function inlarger-scale, systemically organized wholes (see Thibault 1986a).

Quite aside from the individual-centred, even ideologically charged,assumption of a mind which is separated from a body and an individual actorwho is separate from society, the causal and ontological priority assigned to thatreified entity that we usually refer to as 'mind' leads us into waysof thinking andacting that are inconsistent with theoretical assumptions of this book. In the firstplace, no 'entity' exists independently of some context. Individual organisms areintegrated to and are constituent and functioning parts of larger-scale systemsand processes at the same time as they themselves are the integrating contextsfor entities and processes on still smaller scales. Rather than a constitutiveseparation of mind, body, and environment, the focus of the present study willbe on the ways in which individuals and their interactions with both their innerand outer environments are mediated by higher-scalar systems of interpretanceand the social practices in and through which these systems of interpretance aredeployed in particular contexts. Individuals, on their particular scale, are constitutive parts or subsystems of these higher-scalar environments. They do notsimply exist 'in' an environment which is external to them (Salthe 1993: 20).Rather, the semiotically mediated activities in which they engage activelyconstitute their environments by virtue of the continuous exchanges of matter,energy, and information-meaning between organism and environment (see alsoBateson 1973a: 285-91; Gibson 1986 [1979]: 8).

2. The Conceptual Framework of the Ecosocial Semiotic Perspective

The conceptual framework of the present study is an ecosocial semiotic one.This term derives from and is a further development of the concept of ecosocialdynamics proposed by Lemke (1995b). Human socio-cultural systems, Lemkepoints out, are systems of social practices (actions, activity formations) linked tosocially and historically specific meaning formations (thematic and axiological)in which the former are construed as meaningful (Lemke 1983, 1985; Thibault1986b, 1989, 1990, 1991b, 2003a). Furthermore, every social practice 'is simulta-

INTRODUCTION 9

neously also an instance of some material process' (Lemke 1995b: 106). Socialpractices of all kinds are simultaneously material processes in a complex, hierarchically organized, developing and evolving ecosocial system. Socio-culturalsystems are not, then, comprised of meaningful human practices alone. They arealso material-physical processes. From the perspective of an ecosocial dynamics,there is a dialectically dual relationship of complementarity between thesemiotic and the material dimensions of the couplings of the two that areenacted in social meaning-making activity. These couplings form the basis of thedynamics of an ecosocial system. Lemke has formulated the fundamental unityof these material and semiotic couplings in the constitution of a unitaryecosocial system as follows:

Such systems are hierarchically organized at many different scales throughcomplex couplings of processes which feed back to one another to produceentirely surprising, emergent phenomena (self-organization). In thedynamics of complex, tightly coupled systems with strong multiple feedbackloops, even small local changes can produce surprising global effects.Semiotic formations, which slightly bias the linkages of material processesaccording to their semiotic meanings for a human culture, are essentialelements in the material dynamics of human communities. There cannot betwo systems here, one material and the other cultural, each changingaccording to separate laws, relatively independent of one another. There canonly be one unitary ecosocial system, material and semiotic, with a single unifieddynamics. (Lemke 1995b: 107; italics in original)

In other words, there are the dynamics of the meaning-making (semioticdiscursive) practices of the community; there are also (and simultaneously) thematerial-physical dynamics of the material resources and the matter and energyprocesses and flows which embody, constrain, and, in part, organize the former.Ecosocial semiotic formations are actions-formations-cum-material-processesthat are organized in terms of their semiotic relations and salience as socialpractices in a human community. Only some of the many possible crosscouplings between the semiotic-discursive and the physical-material are regularlyand typically made and recognized in a given community. The dialectically dualcharacter of these cross-eoupling relations also means that neither component isprivileged or viewed as dominant. Nor is the one reducible to the other.

Social meaning-making practices are not, therefore, uniquely or one-sidedlycontextualized in relation to the paradigmatically organized sets of potentiallymeaningful differences that are typically recognized as being semiotically salientin a given community. Equally important, socially meaningful acts are contextualized in and through their relations to what Lemke (1984b: 113-21; 1995c:166-75) has called the Interaction System of the community. That is, thematerial, eco-biological and physical relations to other system processes in agiven ecosocial system. It is only through the attempt to understand the constitutive inseparability of these semiotic-discursive and physical-material crosscouplings and dynamics that we can adequately theorize our own and others'embodiments, our subjective experience of our ecosocial environments, our


perceptions of our inner states and sensations and the meanings we attribute tothese both in our internal dialogues in 'inner' speech activity, as well as in ourinteractions with others, and the materiality of the body as playing a central, notmarginal, role in social meaning-making.

In the perspective that I have outlined above, one of the central questions isthe following: how does patterned meaning-making activity, or semiosis, emergeand evolve in time? The fundamental importance of time and process is nowbecoming clearer after decades of theorizing and research which privilegedplans and/or causal mechanism that were assumed to be in the organism fromthe beginning and which could be postulated as the driving forces of cognitive,developmental, and semiotic processes (see Thelen and Smith 1994). Meaningmaking activity involves very many heterogeneous elements, on many differentspace-time scales, ranging from the neurological processes in the human brainto the persistence in historical time of an entire culture's ways of makingmeaning and the resources for doing so. As we saw above, meaning-makingarises from the cross-coupling of both physical-material and semiotic-discursiveprocesses. The time-bound nature of these cross-couplings is fundamental. AsThelen and Smith show, the time-bound cross-couplings of elements from verymany different levels of organization aredevelopment and cognition. The possibilities of combination of the two domains are seemingly endless: there are verymany degrees of freedom. However, the cross-coupling of the two gives rise topattern and organization because the many heterogeneous elements that areinvolved self-organize around fewer, simpler, more global principles of organization that give the resulting whole coherence along its temporal trajectory. As weshall see, the reduction of the many degrees of freedom to produce the patternand organization that we recognize and interpret as meaningful according to

our particular observer perspectives is the result of very complex, nonlinearinteractions among the processes and relations on the many different scalarlevels of organization that are involved.

Ecosocial systems are stable, far-from-equilibrium dynamic open systems.Contrary to our expectations, based on the behaviour of closed systems, stable,far-from-equilibrium systems increase their organization and complexity in time.The far-from-equilibrium conditions that are necessary for this increase inorganization and complexity are maintained through continuous transactions ofenergy, matter, and information between the system and its environment. Suchsystems were called dissipative structures by Prigogine and Stengers (1985[1984]: 12-14, 143-4; Salthe 1993: 106-11) because they maintain (andincrease) their own order and complexity by obtaining needed energy, matter,and information from the environment and then dissipating back to theenvironment some of this waste energy in the form of higher-entropy outputsthat export entropy or disorder to the environment. Ecosocial semiotic systemsand the human individuals who participate in these are, on their respectivescalar levels, dynamic open systems. Such systems - physical, chemical,biological, social - are comprised of very many interacting components ondifferent scalar levels that interact in nonlinear and heterogeneous ways.Moreover, they are dissipative structures: their existence-in-time and theirstructural integrity are maintained by their constant transactions of matter,

INTRODUCTION 11

energy, and information between the system and its environment. Dynamic opensystems are, therefore, dependent on their contextualized relations with their'external' environment. Their persistence-in-time cannot be explained in termsof more fundamental causal principles on lower levels. Instead, their persistencein-time is the organization of processes which depend on their transactions withthe system's environment.

Dynamic open systems are also self-organizing systems (Varela and Maturana1980). The nonlinear interactions among the heterogeneous array ofcomponents that comprise the system enable patterned arrangements that didnot previously exist to emerge spontaneously and to behave in ways that were notpredictable on the basis of the very many elements that comprise the system.Instead, the system as a whole is attracted to a smaller number of globalprinciples that come to constrain the behaviour of the system as a whole suchthat the individual components of the system are attracted to one preferredstable configuration rather than to other possible ones. The enormous potentialvariability of the system gives rise to points of localized instability such that concentrations of matter, energy, and information attract other elements to theiremerging patterns. In turn, these local asymmetries may be amplified and hencebifurcate such that a new mode of organization is selected by the system as itspreferred state. Meaning is not 'caused' by mind or brain states and processesthat arise from within the individual. Instead, the body-brain complex providesinitial (biological) conditions qua material affordances that interact withprocesses on other levels such as the ecosocial one. A heterogeneous diversity ofprocesses and relations on many different space-time scales cooperate to giverise to the meanings that we recognize and interpret in our interactions withothers.

3. Gibson's Ecological Theory of Perception and the Three-level ScalarHierarchy View of Organism-plus-Environment Transactions

Ecosocial semiotic theory also implicates a very different idea of perception, ascompared to the traditional account of perception in psychological theory. Traditional theories of perception, as Grene (1990:18) points out in her discussionofJ.J. Gibson's (1986 [1979]) ecological theory of perception, are based on thenotion of association: organisms of all types 'build up associations from manyinstances' prior to perception. This notion is akin to the nativist claim thatindividual stimulus-receptor cells are endowed with an intrinsic, latent andnatively provided content, which is 'triggered' by appropriate, externally derived'primitive sensations' (see Maze 1991: 171-2 for a useful critique). Cibson'secological theory of perception provides an alternative, which is entirelycompatible with the conceptual framework of ecosocial semiotics.

Gibsori's account is not a representational theory of perception. For Gibson,perception occurs through the organism's active exploration of its Umwelt.Organisms do not receive sense data from the outside and then elaborate a representation of this, which, in turn, forms the basis of a perception. Instead, they'can discriminate information that is species-specifically relevant to their survival


in their environment' (Grene 1990:18). Gibson places the emphasis on theexploratory activities of embodied, active participants in their material environments. For instance, the participants in human social dyads are dialogicallyoriented both to significant others and the environment (the non-self), asconstituted in and through the dyad. The active and embodied nature ofperception, as defined by Gibson, accords with this view.

In dialogically orienting to the other and/or to the environment, theindividual learns to orient him- or herself, to localize, and to converge uponrelevant distinctions, or semiotically construable differentiations, in its Umwelt.In the process, the individual does not so much give meaning to these as receivemeaning from them in and through the exchange processes which constituteboth. Thus, the 'intentional' and active orientation to this principle of alterityrepresents a predisposition to receive relevant meanings from the outside. At thesame time, the autopoetic (Le. self-organizing) nature of these processes meanthat in the making of more and more relevant differentiations, the members ofthe dyad both develop to higher levels of organizational complexity and at thesame time expand their Umwelt (Harre 1990: 300).

Gibson has rejected the notion of perception as a passive neural representation in the form of a data structure which is stored wholly in the brain. Instead,he has written of the 'mutuality' of animal and environment (Gibson 1986[1979]: 8)~ The relation between the two terms - animal and environment - isone of complementarity: the one implies and could not exist without the other.The environment of an animal, Gibson points out, cannot be reduced to orexplained in terms of the physical environment. The environment of livingthings is also informationally constrained, although for reasons that areexplained in section 8, pp. 34-9 below, it is more appropriate to rethink theconcept 'information' in terms of the 'meaning' or the 'significance' of the environment of a given animal. The reductionist view that the environment consistsof physical constraints perse 'neglects the fact that an environment is ambient fora living object in a different way from the way that a set of objects is ambient fora physical object' (Gibson 1986 [1979]: 8). As Gibson also points out, the animalis both a perceiver and a behaver in its environment.

Gibson's theory of affordances is built on this premise. The affordances of agiven animal's environment are what it offers or affords the animal. Again, thecomplementarity, or mutuality, of animal and environment is crucial here. Theenvironment is not the same for different species of animal. Animal andenvironment are mutually adapted to each other so that different species orientto different environmental affordances. According to Gibson, the perceptualpick-up of information about environmental events provides the animal withinformation about that event and what it affords the animal for action orbehaviour (Gibson 1986 [1979]: 127). In this way, the perceptual informationwhich is picked up provides the organism with information about environmentalevents and how the given environmental event can enable the organism to act orbehave in particular ways. The importance of Gibson's theory for our presentpurposes is twofold.

First, it shows that organisms perceive the environment in ways that closelyand internally relate to the ways they act or behave in the environment. Secondly,

INTRODUCTION 13

the particular affordances to which the animal has eo-adapted provide theanimal with information which is already an intrinsic part of its environment.The perceptual pick-up of environmental stimulus information is already primedby the ways in which perception mediates action. There is no requirement herethat all of the relevant information be stored in the animal in the form of passiveand behaviour-neutral databases that the animal consults in order to infer orreason about the appropriate course of action to undertake. Instead, the environmental stimulus information which is ambient to the organism providesinformation that is directly constitutive of the animal's action repertoire. That is,the organism does not process information or data about an objective environment 'out there' and then infer an appropriate course of action. Rather, theaffordances of the environment directly specify possible courses of action for theorganism. In specifying what an environmental event is, they also specify whatthe animal can do in relation to that event.

We need not worry about all the details of Gibson's theory to understand themain point for our present purposes. Perception and action are closely linked toeach other at the same time as they implicate, on-line, and constantly respondto, neural, motor, and contextual (environmental) factors. The traditionalconception of the brain as a central processing unit that 'represents' or 'models'an external world has no place here. Rather, the brain is primarily concernedwith the regulation of sensori-motor activity. That is, the brain regulatesrelations between the body and the environment at the same time as these bodyenvironment transactions act on and shape body-brain functions. The brain isthen a contextual organ which regulates the organism's engagements with itsenvironment. Gibson's theory helps us to see that brain, body, and environmentare interlinked as functioning components in a larger-scale system that cannot beexplained in terms of or reduced to anyone component part of the larger whole.

Gibson's notion of affordances is important here for two reasons. First, affordances are environmental phenomena (objects, events, and so on) that areambient to or surround the animal in some way that is potentially significant tothe animal. Secondly, the affordances of an animal's environment are afunctional component of the trajectory that extends from the organism's centralnervous system through its bodily (sensori-motor) activity and into its environment and then loops back again to the organism. The point is that affordancesare functional components of a trajectory that extends beyond the biologicalorganism and which selectively imports features of the organism's environmentback into the organism's internal dynamics. The notion of the trajectory is afundamental one in this book (see Salthe 1993: 181-8). Trajectories, as we shallsee later, occur on many different space-time scales. In the present context, bothcognition and meaning-making activity can be thought of as time-bound trajectories that extend from the body-brain complex into the environment only toloop back again such that the information- or meaning-potential of the environment is selectively re-contextualized in the internal dynamics of the individual(see also Bateson 1973a: 285-91;Juarrero 1999: 212-13).

Now, the traditional model of the brain as the source of reason and rationality and hence as the central executive or governor which oversees and controlsbehaviour does, of course, require that brains make contact with the external


world. The difference between the approach I am proposing here and the traditional one can be explained as follows. First, in the approach I am arguing infavour of here, the brain perseis not the sole locus of cognitive activity, as in thetraditional approach. Instead, cognition and/or semiosis are a distributedactivity implicating a trajectory which links brain, body, the available semioticresource systems, and ecosocial environment. The seat of cognition and semiosisis the trajectory; it is not the brain per se. Meaning-making activity is not unilaterally caused by a central executive inside the brain. Rather, it is the emergentconsequence of the ways in which both the intrinsic dynamics of the organism e.g. patterns of neural activation, neuromuscular activity, cognitive capacitiessuch as memory, attention, and so on - interact with contextual factors, relevantintertexts, the available material and semiotic resources such that both intrinsicand extrinsic factors are jointly recruited and eo-deployed in the real-timeassembling of the meaning-making trajectory. Secondly, Gibson's theory ofaffordances is one theory which shows that the environment of an animal ismeaningful for the animal in non-trivial ways. That is, meanings are located andstored at the higher-scalar level of the environment which surrounds the animal,rather than being located in or caused by lower-scalar biological mechanismsthat are entirely within the organism.

In this way, we can begin to glimpse the need for a theory which can discussdifferent scalar levels that are implicated in the organism's transactions with theaffordances in its environment. Salthe points out that, for most purposes, a basicthree-level system is 'minimally needed in order to preserve the complexity'(1993: 46). The three-level system therefore avoids the temptation of reducingall levels to a single, privileged level at the same time as it preserves the recognition that there are processes taking place on different scalar levels. Higher-scalarlevels provide the integrating contexts for lower-scalar ones. They act ascontextual constraints or boundary conditions on the lower levels. Lower-scalarlevels constitute enabling conditions of higher-scalar ones. Generally speaking,the intermediate level is selected as the focal level.

The interactions between the organism and some affordance can be taken asthe focal level in the present account. Below this level, there are the biologicalproperties of organisms and the physical-material and possibly biologicalproperties of the affordances that organisms interact with and which predisposethem to some kinds of organism-affordance engagements rather than others.Above the focal level of these organism-affordance transactions, there is the environment of the organism. Importantly, it is in the environment of the organismthat the meanings of environmental events are located and interpreted. This isso because meanings, in the extended sense I am using the term here, are notsolely located in the individual organism, but are distributed across an entirepopulation and its environmental niche. It is the environment as repository ofmeanings that mediates and brings into relation an individual organism quamember of that species and a particular environmental affordance in a giventime and place. But it is the environment of the animal, as defined above, and,therefore, the possibilities of perception and action, in the first instance, that itaffords the animal which mediates this relation and which makes possible itsinterpretation for the organism. To paraphrase Bateson (1973b: 392), there is

INTRODUCTION 15

contextual redundancy (see section 6, pp. 26-30) in the organism-plusenvironment system such that the morphology and behaviour of the organismprovide information concerning the environment that the organism inhabits. Inthe processes of eo-adaptation and eo-evolution that create the organism-plusenvironment system, the environment is imported into the organism's internaldynamics such that it entrains these to its own dynamics (see above).

The relevant system of relations is hierarchically organized in terms of threelevels. The three-level hierarchical system which I have derived can beformalized as follows:

L+1: Environment of animal qua system of interpretance which brings intorelation (mediates) animal and its affordances and provides the higherscalar principles whereby these affordances and their relation to theanimal can be interpreted in ways that afford perception and action;

L: The focal level of the animal's engagements with the affordances that itencounters in its environment;

L-l: The biological and other physical-material properties of organisms andenvironmental affordances that enable them and predispose them toengage in transactions with one another and selected aspects of theirenvironment.

The system of relations that I have outlined here provides a synoptic descriptionof a cognitive and/or semiotic process that extends beyond the individualorganism and into its environment. The three-level scalar hierarchy view showsthat the organism's transactions with the affordances it encounters andinterprets are embedded in and mediated by a higher-scalar environment whichcannot be defined in terms of any given individual perse. Furthermore, the lowerscalar neural processes in the brain and central nervous system, along with thebody's sensori-motor activity, are neither the source nor the 'cause' of meaningsand their interpretation. Instead, the physical properties of body and brain andthe dynamical processes which these enable on their timescale are integrated toand contextualized by processes occurring on the still larger timescales of theprogressively higher-scalar levels Land L+1.

This also means that lower-scalar neural and bodily dynamics are entrained tothe dynamics of the higher scales and their processes. It is only when suchentrainment to a higher-scalar semantic or other meaningful property takesplace that we can say that bodily activities and the physical properties of objectsqua artefacts are involved in the instantiation of operations that are meaningfulon higher levels. In other words, it is only when such activities and properties arefunctional in the processes of enacting, storing, and transmitting informationwhich the organism can construe as meaningful in the service of its particulargoals that we can say they are performing cognitive and/or semiotic functions.

The focal level L is the level on which the individual's agency is manifestedand enacted in specific encounters with environmental affordances and theactions they enable the agent to carry out. The agent is not the biological


organism. Rather, the agent represents the integration of the lower-scalar neuraland bodily (sensori-motor) resources of its organism on level L-l to the perspective of the self and the agency which manifests its selfness on level L. Theindividual is an agent in its own processes of individuation because it can deployand change its knowledge, experience, and perspectives through its own activity.The agent can use its own neurological and bodily processes, as well as those ofothers, as a tool or an affordance in its own activities. The agent is not definedin terms of its neurological processes or its sensori-motor activities. These arelower-scalar (L-l) phenomena whereas the agent is a higher-scalar phenomenonon the focal level L. The lower-scalar phenomena are sub-personal with respectto the agent. Agents contextually integrate these lower-scalar affordances to theirown level, in the process completing them by integrating them to the personalperspective ofa determinate self (Salthe 1993: 49; see also Prodi 1989). At thesame time, the agent and its activities on level L are constrained and mediatedby the higher-order system of interpretance on level L+1.

The agency which is manifested on the focal level L is an emergentphenomenon which results from the complex, non-linear relations between (l)the 'upwardly' emergent affordances of the body-brain on level L-l and howthese are contextually integrated to the next higher-scalar level (L) and (2) the'downwardly' acting system of contextual constraints on the higher level L+l ofthe system of interpretance which is in operation. Thus, the agent and its agencyon the focal level are constituted by the ways in which higher-order constraintsand lower-order enabling conditions (affordances) give rise to an emergentintermediate level of organization which is a consequence of the ways in whichthe semiotic and material processes in the observed system of relations areorganized on the intermediate (focal) level.

Thus, bodily activities such as gaze orientation, pointing, gesturing, speaking,facial expressions, body movement, neural dynamics, and so on provideresources which afford joint social meaning-making with others. The extensionof these possibilities to extra-somatic artefacts such as texts qua material objects,tools, toys, calculating machines, and so on, means that typological semioticresources not only spread throughout the ecosocial environment and colonizeit, as Lemke (2000a: 203) has put it, but they also adapt to and colonize the brainitself (Deacon 1998 [1997]: 327). The brain is not the 'cause' of these processes;rather, it is a functional participant in a semiotic-material trajectory that extendsacross potentially many diverse space-time scales. It is by virtue of his or herbeing embedded in a higher-scalar ecosocial environment and the meaningmaking trajectories which this enables that we can say that the individual, in thesense of the person who participates in social meaning-making with others, is notthe biological organism, but the trajectory-in-time. This trajectory extends fromthe neural dynamics of the individual in ways that regulate, entrain andmodulate motor dynamics so that the meaningful content of the neuraldynamics is transduced into an instance of a given act-token. Writing on howintentions organize and channel action trajectories that originate in semanticneural space,]uarrero has formulated this question as follows:

INTRODUCTION 17

Thinking of prior intentions as self-organized dynamics within conscioussemantic space, of proximate intentions as control knobs governing thatspace, and of act-types as attractors through that space allows us to conceptualize act-tokens as individual behavioural trajectories within that semanticspace. These trajectories originate in intentional dynamics that both activateand then channel and regulate motor processes top-down such that the latterpreserve and execute the invariant relations that embody the intention'smeaningful content. (Juarrero 1999: 212)

Semantic neural space is itself the result of the ways in which the higher-scalarsystems of interpretance emanating from the ecosociallevel have entrained thebrain's neurological processes to their own dynamics. In this way, the body-brainis contextually in tegrated to the dynamical (ecosocial semiotic) processes of thehigher-scale. Kinsbourne (2001: 246) argues that meaning and experience in thebrain are not assembled piecemeal fashion from smaller bits that are aggregatedinto larger wholes, bottom-up fashion. Rather, the global network of neurons, ofwhich the newborn has a full complement, progressively differentiates out of aless differentiated whole into an increasingly more differentiated network ofnetworks of networks ... (Kinsbourne 2001: 246) by virtue of the ways in whichthe brain is functionally embedded in and entrained by ecosocial dynamics, aswell as by its own internal dynamics. The brain is then sculpted by the dynamicsof the topological and typological functions of the body and its extra-somatic,artefactual extensions and the ways these operate across different space-timescales in the ecosocial environment.

This approach suggests that we need to develop a new discourse for talkingand thinking about the ways in which brain, body, and ecosocial semiotic environment are embedded in and are functioning participants in higher-scalarsystems that link all three components in complex, hierarchically organized andnon-linear interactions across the many levels of relations and space-time scalesthat are involved (Togeby 2000). Such an approach needs to be combined withan ecosocial theory of meaning-making activity. This also suggests that the shape- the intrinsic patterned organization - of our semiotic resource systemsconstitute principles of environmental organization that can be seen as fundamental components of brain organization itself. Neural dynamics and neuralarchitecture will thus be seen to be in tune with the dynamics of the body andthe ecosocial environment. Consider in this respect the following observationsmade by Kinsbourne on the left-right hemispherical organization of the brain:

We learn that most people's left cerebral hemisphere is concerned withlanguage, their right with spatial orientation. More broadly still, the left catersto sequential analysis and the generating of action sequences, the right tosetting such activities into a spatial framework. Most general of all, the lefthemisphere controls motivated approach sequences (handling, eating, and soon), progressively focusing and acting upon the target; the right hemisphereis more involved with the person's movement through the intervening spaceand the spatial background of the target. The left hemisphere's activities canbe context free, whereas the activities of the right are context bound. The


hemispheres are complementary in their functioning. (Kinsbourne 2001:236)

These observations by Kinsbourne show the diverse, yet complementary, rolesthat the two hemispheres play in the regulation of sensori-motor activity. The lefthemisphere is 'analytical' and 'focal'; the right hemisphere is 'general' and'global'. It is not difficult here to see how the metafunctional organization oflanguage that is postulated by systemic-functional theory can be related to theproperties of the global hemispherical network, encompassing both left andright hemispheres. According to systemic-functional theory, language form isorganized in terms of a number of diverse functional regions which are simultaneously configured in the forms of language on various levels of organization(see section 10, pp. 46-9, below). Given that social semiosis is an extension andfurther elaboration of the body-brain's neural and sensori-motor activity, we canpostulate that the brain is attuned to the metafunctional organization oflanguage and other semiotic modalities (see Chapter 5).

For example, the 'analytical' and 'focal' properties of the left hemisphere thatKinsbourne identifies in the above quotation can be related to the typologicalcategorial bias of the experiential and logical metafunctions. On the other hand,the 'general' and 'global' properties of the right hemisphere can be related tothe topological-eontinuous bias of the interpersonal and textual metafunctions.This claim has nothing whatsoever to do with the notion that language is codedin the brain in the form of a neural code, seen as the ultimate neural substrateof contextualized, linguistically mediated activity. However, it does claim thatbrain dynamics, on their particular scale, are entrained to the dynamics of semiotically mediated activity and in ways which enable brains to regulate the body'smeaning-making activity so that it will be both integrated with and appropriateto its contexts. The biological architecture of the brain is consistent with theorganization of ecosocial semiotic processes, though this does not mean that thelatter can be explanatorily or causally reduced to the former.

4. The Brain as Regulator of Sensori-motor Activity: Implications for SocialSemiosis

The brain is not a symbolic processor, but a regulator of sensori-motor activity.As we shall see later, sensori-motor activity can be seen as proto-metafunctionalin its organization. Social semiosis emerges through the entrainment ofembodied sensori-motor activity to the higher-scalar meaning-making practicesof the ecosocial system. It is through brain-regulated sensori-motor activity thatindividuals extend their selves and, hence, their trajectories into their ecosocialenvironment at the same time that that environment is selectively incorporatedinto the individual's body-brain dynamics in the form of higher-scalar or topdown contextual constraints. The cross-modal basis of the brain's regulation ofsensori-motor activity is, in this way, extended into semiotically mediated formsof activity (meaning-making). That is, the original cross-modal basis of the bodybrain's sensori-motor engagements with its environment gets extended into

INTRODUCTION 19

increasingly specified and socially constrained multimodal forms of meaningmaking involving both somatic and extra-somatic resources and theircombinations.

By the same token, the body-brain's lower-scalar sensori-motor activity constitutes bottom-up affordances which are entrained to the higher-scalar, top-downconstraints of the social meaning-making practices and conventions in which thelower-scalar body-brain dynamics of individuals are embedded. It is the interaction and combination of these two levels which enables newly emergentintermediate levels such as, for example, the dyads that characterize the 'primaryintersubjectivity' of mother-newborn interaction, to emerge (Thibault 2000a).Moreover, it is the embedding of body-brain dynamics in and their entraining tothe higher-scalar dynamics of the socia-cultural processes that are characteristicof some community or social network that allows us to recognize and interpretthe emergent properties of brain-regulated sensori-motor activity as being bothsemiotically salient and socially functional. In this way, the metafunctional organization of linguistically mediated and other forms of social semiosis can be seenas an emergent consequence of higher-scalar patterns on the ecosocial scale.

The logic of the three-level hierarchy view is evident here. Thus, theemergence of social semiosis in the individual is not the result of a biological ormental programme on level L-l. On this level, biological initiating conditions inthe form of the values (Edelman 1992; Thelen and Smith 1994: 185-6) thatnudge the individual organism along some developmental pathways rather thanothers (e.g. orienting to faces as a source of meaning and affect; attending tohuman voices, and so on) are entrained by higher scalar factors on the ecosociallevel. At the focal level of the (mother-infant) dyads in which infant semiosis isobservable and at which level the infant engages with our perspectives, the infantmoves from dynamic attractor to attractor as development unfolds in time.

Lower-scalar sensori-motor activity (L-l) constitutes affordances which standin no single or fixed relation to higher scalar meanings. For example, the neurological processes which regulate sensori-motor activity and emergent socialmeanings stand in no fixed relation to these meanings. Neural firings in thebrain do not generate meanings. Nor do they represent symbolic tokens that arestored in the brain. Rather, lower-scalar neural activity self-organizes to formhigher, more global patterns of brain activity that embody the brain states thatwe recognize and experience, as well as attribute to others, as thoughts,intentions, feelings, meanings, and so on. These higher-scalar mental states donot stand in any fixed relationship to the neurological processes that realizethem. It does not therefore matter which neurological patterns on level L-lrealize higher-scalar mental states (Juarrero 1999: 190). In this sense, the lowerscalar neurological processes are affordances which are entrained to thedynamics of higher-scalar meanings.

The lower-scalar neurological processes afford their selective reorganizationas more global patterns which are meaningful on the higher-scalar perspectiveof the self. Juarrero argues that behaviour is meaningful action depending on'whether or not the neural processes transmit information as mental' (1999:191). In this way, arm movement induced by an involuntary spasm and armmovement that is construable as an instance of an act-type such as greeting


someone are distinguished. Only the latter's trajectory, Juarrero (1999: 191)points out, originates in and takes place in semantic space. The main point hereis the role that meaning plays in regulating and bringing about the given motorbehaviour qua act-token. Neurological processes therefore do not causemeaningful action. Rather, they are affordances in the sense that they constitutean information potential which is selectively reorganized on higher levels as'meaning', in the broadest sense of the term, in the perspective of a self.

As we shall see in more detail later in this study, the information potential ofthe lower-scalar neurological and sensor-motor dynamics (level L-l) are reorganized by the next-higher level (level L) as semiotically salient differences orrelations - typological or topological - that are recognizable and interpretablefor the relevant system of interpretance on level L+1. Strictly speaking, it is notpace Juarrero (see above) a matter of whether the neurological processes'transmit' a meaningful content or not. Rather, it is a question of whetherlower-scalar neurological processes constitute information which is able to bereorganized in ways that are meaningful to higher-scalar levels of organization.The principles which I have outlined here have been formulated by Lemke(1999) as the Principle of Alternation. We shall see later that Lernke's principleis a powerful and elegant theoretical tool that plays a key role in building thetheoretical bridges between brain, body, meaning, and ecosocial environment.Furthermore, it fits perfectly with Edelman's theory of neuronal group selection.There is no homunculus or other ultimate causal or executive factor in the brainwhich directs the process of meaning-making. Rather, neural dynamics, on theirtimescale, interact with and provide support to events on other timescales.Meaning is the emergent outcome of the interaction across these processes ondiverse levels.

The crucial factor here is the operation of different timescales on thedifferent levels that constitute the given system. Salthe points out: 'Processes atdifferent scales change direction, cycle, or go to completion at different rates(have different rate constants), with lower-scale ones behaving faster and havingsmaller relaxation times than higher-scale ones (which, however, can travel atgreater absolute speeds)' (1993: 46). Salthe's observations pertain to the organization of scalar hierarchical systems (see section 3, pp. 14-16). With respect tothe physical brain processes that realize it, a mental content is a higher-scalarprocess; it has a slower, larger timescale with respect to the much faster scale ofthe neurological processes that are its substrate. The latter occur beyond thelevel of our awareness, whereas mental contents on the higher-scalar levelconstitute the phenomena of awareness.

This suggests that faster, lower-scale neurological processes are reorganizedacross different scalar levels as perceptual and mental contents that are salient toand congruent with the timescale on which the self experiences and interpretsmeanings of all kinds (see also Finnemann 2000: 300). Again, we see here theimpossibility of insulating one timescale from the other, for the very fasttimescale of the neurological processes, say, is not isolated from the slower,higher-scalar level of phenomenal awareness and experience, but is integrated toit such that no insulation of the dynamical relations and processes on each levelcan be postulated. This integration across levels suggests that the emergence of

INTRODUCTION 21

meaningful contents in the brain is based on the principle of the heterarchical,rather than strictly, hierarchical organization of the different scalar levels thatare involved.

Kinsbourne (2001: 240) argues on the basis of extensive neuropsychologicalobservation and research that cerebral organization can be understood as 'anarchitecture of nested functional units'. To quote Kinsbourne:

We emerge with an understanding of cerebral organization as an architectureof nested functional units. Within an overall sequential analytic mode, the lefthemisphere generates word sequences in one of its parts, action sequences inanother, and sequential identification (for instance, of letters) in yet another.Within subareas of the respective areas, it enables both recognition andexpression. In subareas of those subareas, it enables material-specificprocessing, for example, of action words, color names, names of animals, orthe connotations of 'right' and 'left.' Within its overall spatial relationalmode, the right hemisphere enables orientation in ambient space in one ofits parts and pattern perception in another. In subareas of these subareas, itpermits subskills like map reading, face recognition, or identifying sketchedshapes ('perceptual closure'). A recursive organization emerges; superordinate processing modes differentiate into distinctive domains of functioning,and these again into different specific applications. I envisage a relativelyundifferentiated whole that progressively differentiates (in child development? in evolution?) into the rich set of specifically human potential skillsand alternative strategies. (Kinsbourne 2001: 240)

The brain is organized as a global network of networks of networks of networks... (Kinsbourne 2001: 241). Furthermore, the higher mental functionsconstitute a further development of the brain's original adaptive function as aregulator of sensori-motor activity. Abstract, symbolic forms of thinking,reasoning, and meaning-making are not localized in special modules thatdeveloped later, but are emergent from the requirements of sensori-motorregulation. Moreover, in emerging from sensori-motor regulation, abstractsemantic categories are able to reorganize and to direct the flow of sensorimotor behaviour along semantically guided trajectories. This is so both from thepoint of view of behaviour and from the point of view of perception: we cansemantically entrain our own behaviour so that it instantiates a token of a recognizable act-type at the same time that we can construe observed behaviour asan instance of this or that act-type. This is clearly evident in the ways in which thesemantically organized lexicogrammatical categories of, for example, [AgentAction-Affected] and their more specific instantiations (e.g. [Actor-ActionGoal], [Sayer-Verbal Process-Verbiage], and so on) can be seen as deriving fromsensori-motor categorizations at the same time as constituting their moreabstract reorganization as the semantic categories that we use in higher-ordermental functions such as verbal reasoning, planning, and symbolic construal ofour inner and outer environments.

The recursive organization of cerebral organization is reflected in therecursive organization of language itself (see above quotation). This does not


mean that cerebral architecture is hard-wired for universal principles oflinguistic organization. Rather, it reflects the way in which neurological processesare entrained to higher-scalar attractor states such as the meanings that derivefrom the ecosocial level. Specifically, it reflects the way in which the brain isentrained to principles of linguistic organization that are shared by very largenumbers of people on the much slower, larger timescale of the historicalevolution of a language community of many generations. Boundary conditionsare established by higher-order, more schematic categories which set theparameters within which more specific categories and their instantiations inbehaviour are meaningful. Thus,

Patients with focal brain lesions do not violate the rules that governed theirpremorbid performances. They remain guided by preexisting parameters butbecome less specific, slower, and less stable in their responses. (Kinsbourne2001: 243)

Patients with focal brain lesions can be said to recede further 'up' the scale ofinstantiation such that the broader, shallow systemic properties are retained inways which constrain behaviour in some more schematic domain whereas themore delicate, more context-specific properties are impaired, compromised, orlost in ways which adversely affect the patient's ability to make very fine-tunedcontextual adjustments to more specific domains. This suggests that the neuralnetworks for language and those for other domains are organized according tosimilar principles of recursive organization.

In each case that Kinsbourne describes, the patient regresses to moreschematic, hence less specific, categories that regulate behaviour and specify itsmeaningfulness in increasingly broader, shallower ways. These observations ofKinsbourne suggest that the information that is potentially afforded by lowerscalar neurological processes and the reorganization across progressively higherscalar levels of organization is what enables linguistic forms to 'differentiate outofless-specific preconscious precursor states, with the word order implicit in theprecursor state. The brain models and remodels until the utterance is perfectedin its analytical detail' (Kinsbourne 2001: 244). In terms of the Principle ofAlternation (Lemke 1999), we can say that the brain organizes and reorganizestopological pre-semantic states on lower levels of neural organization until thetypological-categorial linguistic form emerges. Rather than a bottom-upassemblage of parts into whole, language in the brain is a result of the crossscalar reorganization across levels of the information potential that is affordedby lower levels as meaningful patterns on higher levels. Kinsbourne expressesthe matter in the following terms:

While no truly apt metaphor for how the brain works comes to mind, 'crystallizing out' seems more fitting than 'assembling together.' Interestingly,brain development proceeds according to similar principles. The newbornhas a full complement of neurons; further development proceeds by selectivecell death and elimination of synaptic connections. The biological chiselprefigures the microgenesis of brain states. (Kinsbourne 2001: 246)

INTRODUCTION

5. Topological and Typological Modes of Semiotic-materialInterdependence

23

Mental states, experiences, feelings, and so on, are generally assumed to haveproperties that are intrinsic to them. But it is more useful to think of suchphenomena as complexly constructed meanings that we define in terms ofpatterned contextual relations. The term 'meaning' is normally reserved for thekinds of patterned relations that we recognize and use in, for example, language.Terms such as mental states and experience refer to particular patterns ofmeaningful relations that our community's habitual ways of thinking and actinghave taught us to regard as phenomenal givens that are the personal property orpossession of a particular individual. However, our ability to construe somethingas a mental state or an experience depends on our making sense of patterns ofrelations whereby we connect some things to other things and not to others(Gendlin 1962). There is a widespread tendency for people to say, for example,that they are conscious of something, that they are having or have had a certainexperience, or that a certain mental state occurred in their mind. In the presentstudy, I shall assume that the given experience, mental state, and so on, is mademeaningful for us by the ways in which we construe it in relation to other experiences, mental states, material phenomena, and so on. That is, we constructcontexts for something through the patterned relations that we construebetween some events, things, inner body states, and so on, and others.Experience of all kinds is based on and depends on patterned contextualizingrelations which select or foreground some patterns rather than others, in theprocess drawing attention to only some aspects of all the possible ways in whichour inner and outer environments are meaningful to us.

Following Lemke (2000a: 194), we can say that selves relate to and interactwith both their inner and their outer environments on the basis of two kinds ofsemiotic-material interdependence. The first kind is topological and is based oncontinuous variation. Lemke (2000a: 194) points out that this topological modeof semiosis is 'primary and characteristic of thermodynamic and biologicalsystems'. Topological semiosis is based on quantitative or analogue differentiation or variation that makes a difference in relation to some system of relations.Topological semiosis is not based on differences or contrasts between discrete(digital) categories.

The second kind of semiotic interdependence is typologicaL It functions on thebasis of discrete categories, e.g. the lexicogrammar and semantics of naturallanguage (Halliday 1988). Typological semiosis creates meaning through thecontrasting of discrete categories and types. A given sign is contrasted with othersigns in ways which are, in turn, construed by still higher-order contextualizingrelations that specify how a given combination of signs is relevant andmeaningful in relation to that context, rather than other possible contexts inwhich it might have occurred and in which its meaning might have beendifferent.

The distinction between topological and typological semiotic interdependencies is a further development of Bateson's (1973c: 342-4) distinction betweenanalogic and digital coding (see also Wilden 1980 [1972]: chapter 7, 1981:


19-21). The former, Bateson argues, is concerned with 'patterns and contingencies of relationship' (1973c: 337, 1973b: 388). The latter solves the problem as to'how to be specific about something other than relationship' (1973c: 337).Bateson (1973b: 387) also argues that human verbal language is not derived 'inany simple way' from the iconic-indexical semiotic modalities of, say, 'kinesics'and 'paralanguage'. In my view, this means that human language, with itstypological-categorial grammatical and semantic categories, did not simplyemerge, bottom-up fashion, from these evolutionarily prior modes. Nor does itmean that language is a unique, one-off result of a language gene that distinguishes us from all other animal species. Both of these accounts, in theirdifferent ways, are reductionist.

The first account is reductionist because it assumes that lower-level enablingconditions such as the motor processes of the body are building blocks whichcan be assembled into human language in bottom-up fashion. The secondaccount is reductionist because it invokes explanations based on innate geneticfactors existing on the lower-scalar level of the organism (not the person). In thisaccount, privileged sub-personal causal agents that originate from within thebiological organism are primary whereas environmental factors play no morethan a subsidiary or 'triggering' role. Language is an emergent intermediatelevel of organization that is the result of the interaction between both higherscalar ecosocial constraints and lower-scalar biological (neurophysiological,genetic) affordances. The fact that the (phylogenetically and ontogenetically)prior topological semiotic modalities that are concerned with relationship arenot transcended by language, but are integrated to and reorganized by it,requires explanation.

In the Darwinist view, lower-level processes bring about effects on higherlevels through a process of bottom-up emergence. In this view, the whole is anaggregate of its component parts. But this view has nothing to say about howhigher-level processes act on and bring about constraining effects on lower-scalarlevels. How can our systems of meaning-making (L+1) and the agents (L) whodeploy them act upon and bring about effects on lower-level (L-l) biologicalprocesses? (see also Finnemann 2000: 281). The old dichotomies, though stillpersistent, between 'biology' and 'culture', or between 'nature' and 'nurture',are both useless and irrelevant. As the previous discussion of Gibson's ecologicaltheory of perception emphasized, both the information afforded by the environment and the neurophysiological characteristics of the biological organismare component levels of a single system comprising many different levels andtheir dynamic interactions in time. What is important is not the reduction of onelevel to the other, but an account in which social semiosis, or meaning-making,is explained in terms that are consistent with what we understand about thebiological basis of semiosis without, however, reducing all of semiosis to biologyper se.

Rather than explanations based on one-way efficient causality and the modelsand metaphors derived from the digital computer, the emphasis will be on aconceptual framework of complex, dynamic, open, adaptive and goal-seekingsystems which share certain fundamental thermodynamic properties with allliving systems, as well as with some abiotic systems such as ponds, eddies, and

INTRODUCTION 25

deep-sea vents whose macroscopic configurations qua dissipative structuresprovide both the thermodynamic potentials and the kinetic substrate for livingsystems to emerge (Salthe 1993: 277-8). Human agents qua trajectories-in-time,as well as the ecosocial semiotic systems they are embedded in, though ondifferent scalar levels, are systems which evolve or individuate (Salthe 1993:147-51). The same can be said of particular semiogenetic trajectories. Evolvingsystems are neither designed nor rule-based. Their organization, their structuralintegrity at any given moment of their existence, is a result of the ways in whichthe relational dynamics of the system as a whole constantly adjust to and selforganize in response to changing contextual parameters on diverse timescales.

Topological semiosis is characteristic of our interactivity with our immediatematerial environment, including the external world, our own bodies, and feltinner states of the body. We can then ask the question as to how our semiotic differentiations - both topological and typological - function in the ways in whichwe construe and construct our relations between our inner and outer states, ourbodies, and the material world with which we interact. For example, the ways inwhich we construct meaningful patterned relations between perceived or feltinner states of the body and our (always constructed) sense of a self with anindividual identity, or the ways in which we use our bodies to engage with and tointeract with our external material environments, entail the use of the body itselfas a material artefact whose topological interactivity with its immediate environment provides the basis for the emergence of typological modalities of semiosisthat greatly extend the possibilities for meaningful interaction between ourbody-brains and the wider, or higher-scalar, ecosocial system in which organismsare embedded and to whose dynamics their body-brains are entrained. In thisview, topological semiosis, which forms the ground for all of our interactionswith our environment, enables typological semiosis to emerge and, at the sametime, the former is never transcended by the latter. In this way, the body-brainitself, along with its extra-somatic artefactual projections in the form of texts,tools, toys, electronic gadgets, and so on, are vectors (Lemke 2000a: 202) for thespread of typological semiosis and its categories through a particular ecosocialsystem.

Both continuous variation (topological) and discrete contrast (typological)can be picked up by the organism and construed as signs of inner and/or outerenvironmental events and experiences. In this way, organisms build up andelaborate in the Innenwelten that are stored in their central nervous systemsmodels, or systems of interpretance, for responding to and acting on the worldin ways that are mediated by topological or typological modes of semiosis (seesection Il, pp. 49-54). It is not difficult to see here that the 'signs' which mostimmediately engage the organism are topological ones, based on the ways inwhich continuous change or variation in material phenomena are picked up andinterpreted as signs of environmental events.

For example, continuous variation in a given haptic or olfactory stimulus ispicked up by the receptor organs and transduced into neural impulses that standin no necessary relation to the environmental event. In contrast, the stimulusinformation about some environmental that was picked up by the receptorsstands in an indexical, therefore, a necessary relation to that event. This


transduction into neural impulses means that the resulting neural impulsesstand in a symbolic relation to the environmental event, i.e. they are translatedinto information stored in neural impulses that constitute a higher-scalar systemof symbolic possibilities, based on informational-semiotic constraints, for interpreting and responding to specific events that are (indexically) picked up by theorganism.

6. Contextualization and Meta-redundaney

Typically, we say that we tasted, smelt, felt, and so on, such and such, as if it werea thing, a phenomenal given that we experience as such. However, this way oftalking overlooks the essentially relational and contextual nature of theexperience. The relational and contextual nature of the patterns of experiencecan be described in terms of the principle of meta-redundancy that Lemke(l984c: esp. 35-9) has developed on the basis of the earlier proposals of GregoryBateson concerning meta-communication and meta-learning (e.g. 1973d:102-11). Redundancy refers to the likelihood or the probability with which twoitems, say, connect or combine to form a larger patterned whole or relation.Consider, for example, the eo-occurrence of language and gesture. Aredundancy relation exists between some combination of linguistic utteranceand gesture when not all possible combinations of words and gestures are likelyor possible when we speak. When such a redundancy relation exists, this meansthat the word and the gesture mutually predict each other's eo-occurrence.There is a better than random chance of their combining or occurring together.Thus, linguistic expression and gesture are redundant with (redound with) eachother in the sense that the occurrence of one predicts the probability of its beingcombined with the other.

Now, language and gesture are very different semiotic modalities, organizedaccording to different principles. It is not the case that each modality is simply adifferent material means for expressing the same kinds of meanings. Languageis predominantly typological-eategorial; it is based on discrete categorial contrastor difference. Gesture, on the other hand, is topological-continuous; it is basedon continuous variation of visual and spatial relations. The two semioticmodalities do not simply express the same meanings by alternative means ofexpression. Instead, they make different meanings on the basis of their differentprinciples of organization. Furthermore, there are no fixed, context-freemeanings which language and gesture have independently of the contexts inwhich they are combined. This means that the two modalities can combine inmany different, situation-specific wayssuch that their combinations produce newpossibilities for meaning in different contexts. By the same token, there are alsolimits on the number and kinds of combinations that actually do occur. Not allpossible combinations are equally probable and some combinations are morelikely or typical than others even though the range of possible combinations ofthe two modalities is obviously very large indeed. It is this flexibility of combination, within limits, which allows for the creation of meanings that would not bepossible on the basis of words or gestures taken separately.

INTRODUCTION 27

The combination of words and gesture depends on the context. The combination of word and gesture forms a patterned relationship in which the wordsand the gesture are redundant with each other. The words and the gestureconstitute a first-order redundancy relation. However, the patterned combination of words and gesture on this particular level depends, in part, on the expectations people have about the contexts in which this particular combination istypically used. The context is redundant with the combination of words andgesture, not with the words or the gesture, taken separately. That is, the contextis redundant with the redundancy between the words and the gesture on the firstlevel. Situations are of a higher logical type than are the combinations of wordsand gestures. For this reason, a situation or context is said to stand in a secondorder or meta-redundancy relation to the first-order redundancy relationbetween words and gestures. Words and gestures combine to form patternedrelationship on the first level. However, this level in itself does not specify thecontext which the patterned relationship on the first level indexes or in someway creates or defines. For this reason, it is necessary to specify the second-orderor meta-redundancy relation of the context in which a particular combination ofwords and gestures might have the meaning that it does. Importantly, there isusually no fixed relationship between the context and the combination of wordsand gestures: the meaning of the latter can vary according to the differentcontextual relations that they participate in or are redundant with.

Following the me ta-redundancy formalism proposed by Bateson (l973d:102-7) and developed further by Lemke (l984c), the different orders ofrelations that are involved in a given contextualizing relation are represented bythe use of the slash, i.e. 'I'. A single slash indicates a first-order redundancyrelation, a double slash a second-order or me ta-redundancy relation, and so on.In the following example, the two orders of relations are therefore representedas follows:

word I gesture I I situational context

Let us now return to the previous discussion of examples involving ourexperience of taste, smell, and so on. It is not difficult to show how the samebasic logic of contextualization applies here as well. For example, the perceptualpick-up of stimulus information about some environmental event is a first-orderredundancy relation along the lines discussed above. In this case, the stimulusinformation that is picked up by the receptor organs in the form of energy thatis produced by the given environmental event is redundant with that event.Thus, stimulus information and environmental event are redundant with eachother. Importantly, redundancy relations are always symmetrical. Thus, stimulusinformation is redundant with some environmental event and, at the same time,the environmental event is redundant with the stimulus information that ispicked up by the perceptual systems of the organism. Likewise, words areredundant with the gestures that eo-pattern with them just as the gestures areredundant with the words. The same logic also applies to the meta-redundancyrelations between the redundancy relations on the first-level and the contextualizing relations on the second level of the given hierarchy of contextualizingrelations. Thus, context, C, is meta-redundant with the redundancy ofwords and


gestures, just as the redundancy of the words and gestures are meta-redundantwith context C.

To return to our example involving the perceptual pick-up of stimulus information, we can postulate the existence of symbolic information in neuralimpulses for interpreting the significance of the first-order redundancy relationbetween stimulus information and environmental event. Once again, higherorder contextualizing relations define the significance of the first-order relationsand, at the same time, the latter create patterned relations that constitute theformer. This is so because contexts are enacted, created, and defined by thepatterned combinations of items on the level of first focus at the same time asthe higher-order context operates as a frame of reference in which the relationson the first level are interpreted, or given meaning for the observer. In thepresent example, the orders of relations involved may be represented as follows:

stimulus information / environmental event / / interpretation: response / / /SI in central nervous system ...

More concretely, we can imagine a scenario such as the following by way of illustration. A monkey encounters a species of venomous snake in the tropical rainforest. In so doing, the monkey, in detecting the movement of the reptilethrough the debris of the jungle floor, picks up stimulus information (visual,auditory, and so on) about the given environmental event, i.e. the presence ofthe snake, its movement, its location relative to the monkey, and so on. Thestimulus information which is picked up provides the monkey with informationabout a material event in its environment. Stimulus information and environmental event constitute a first-order redundancy relation. The monkey'sreaction is one of fear and he withdraws from the scene so as to avoid the dangerposed by the snake. The monkey possesses in his central nervous system a modelas to how to behave with respect to snakes. That is, his central nervous systemconstitutes a higher-scalar system of interpretance which enables the monkey tointerpret the lower-order redundancy relationships in the appropriate way danger! - as well as to adopt an appropriate course of action (flight). In this case,the levels of contextualizing relations involved may be approximated as follows:

stimulus information / presence of snake / / danger: flight / / /SI in central nervous system ...

Thus, the monkey's physiological reaction ('fear') and its behavioural response('flight') are second-order redundancy relations which are redundant with theredundancy between the stimulus information and the environmental event onthe first level. In this example, a third-order relationship shows the presence inthe monkey's central nervous system of a system of interpretance which enablesit to interpret the first-order relation in the appropriate way and therefore toadopt an appropriate course of action.

Now, the symmetrical or two-waynature of (meta)-redundancy relations doesnot assume a determinate or fixed relationship between levels in the metaredundancy hierarchy. Such an assumption would mean that, for example,

INTRODUCTION 29

second-order or meta-redundancy relations are reducible to and, therefore,indistinguishable from the first-order ones which constitute them. In otherwords, this would mean that the description of the second-order relation isexhausted by the description of the first-order one. This would also mean thatthere is no distinction between context on the second level and the patternedcombination of relations on the first level. However, we have seen that a givencombination of, for example, words and gestures on the first level can bespecified by different contexts. The meaning of the first-level combination eovaries according to the higher-order context in relation to which it is interpreted.On the other hand, the assumption of a fixed or determinate relation betweenthe two levels would mean that no such variation between lower-level (linguisticand gestural) forms and higher-level contexts could exist.

The principle of redundancy, on the other hand, demonstrates that therelations between higher and lower levels allow for constrained variability in therelations between levels (Thibault 1991a: 92; Finnemann 2000: 280). Thus, thesame lower-level redundancy relations may participate in a higher-order metaredundancy relation with different contextualizing relations. The same formalrelation can be contextualized in different ways. Secondly, the same contextualrelation, the same meaning, may be realized in different lower-level patternedcombinations oflexicogrammatical, gestural, pictorial, or other formal relations.There is no fixed or wholly determinate relationship across levels in the metaredundancy hierarchy.

The various examples discussed above were all hypothetical. In asense, theyconstitute thought-experiments, though this has been for a purpose. My purposehas been to show that both the interpretative activities of different species(human and monkey) and what we commonly think of as semiosis andperception can be seen to share a similar underlying logic. This does not meanthat the differences between perception and activity are being disregarded. Nordoes it mean that the me ta-redundancy formalism as presented here can revealthe detailed neurophysiological and other architecture of either perception orsemiosis. What I am interested in here are some general principles which theyboth share. In particular, we see that it is the recognition of patterned relationsas occurring in certain contexts rather than others which is the salient point inboth cases.

In relating patterned relations to some contexts and not others, both humansand monkeys are able to attribute significance to these patterns and hence to actaccordingly. Moreover, the patterned relations to which a given individual human or monkey - attributes significance are constrained by higher-orderboundary conditions. In the monkey case, the boundary conditions are thetropical rainforest and the existence of life-threatening species of snakes, inrelation to which the monkey adopts - partly through learning, partly throughgenes - a particular behavioural stance. In the human case, the boundaryconditions are the cultural practices which constrain and regulate human activityin a particular community or social network.

The principle of meta-redundancy refers to a system of contextualizingrelations among different classes of acts, e.g. words and gestures. Firstorder redundancy relations are constrained by second- and higher-order


meta-redundancy relations. The principle of meta-redundancy thus shows howmeanings are made on the basis of relations of relations of relations ..., ofsystems of systems of systems . . ., and so on. Higher-order relations act ascontextual constraints on lower levels by synchronizing the component partsof the lower levels in ways which permit order and patterned arrangements ormeaningful combinations to emerge. That is, the probability of combination ofthe items on a given level is contextually dependent on the higher-orderrelations which constrain its possible patterns of combination. These constraintsimpose limits on the possible patterns of combination on lower levels. In the caseof language, the syntagmatic and paradigmatic patterns of its lexicogrammarresult from the correlating of the previously discrete functional regions that aretypical of protolanguage into a more complexly differentiated and interrelatedwhole. This shows how different functional regions in the grammar of alanguage, or in different semiotic modalities, depend on the ways in which theother components of the whole behave in relation to each other.

The first-order redundancy relation that we considered above between wordand gesture means that the occurrence of the particular words depends on andpredicts the likelihood of its eo-occurrence with a given gesture, and vice versa.The establishment of such a two-way or symmetrical redundancy relationbetween the two items means that they are systematically and internally relatedto each as parts of some larger whole, e.g. an occasion of talk or some part ofthis. To the extent that word and gesture are now internally (contextually)related to each other, or, in other words, to the extent that they are now mutualcontexts for each other's interpretation, we can say that words, say, have beenintegrated to the internal dynamics of gesture, and vice versa.

Meta-redundancy relations are not static and timeless. Instead, they aredynamic, time-bound, and historical relations and processes. A given contextualizing relation at a given moment in time depends on its history on diversetimescales - the history of the unfolding discourse occasion, its intertexts (whatwas said, done, and so on, yesterday, last week, last year, and so on), theindividual histories of the participants, the history of the community and itsmeaning-making practices.

7. A Critique of the Causal View of Brain-mind Relations

Semiotic relations and practices are always materially embodied in physicalmaterial processes and interactions, though not reducible to them. As I pointedout in section 2 (pp. 8-11) we can talk about social semiotic practices in twoways: (I) in terms of their meaning relations to other practices in a given socialformation; and (2) in the language of physics and biology, in terms of thephysical processes and relations which embody them (Lemke 1995b: 106--7).Physical and biological processes can only have the meanings they do in a givensocial and historical formation through the social meaning-making practices ofthat formation. In this section, I shall consider some of the problems that arisewhen social semiosis is reduced to or causally explained in terms of physicalbrain processes.

INTRODUCTION 31

The two domains - the semiotic and the physical - are distinct, though complementary. This complementarity means that the social semiotic can be seenfrom the dialectically dual perspective of these two conceptual domains, whichare not reducible the one to the other. The trouble starts, as Bhaskar (1979: 126)points out, when the social and psychological sciences seek to overcome theirown conceptual inadequacies by reducing the domain of their inquiry to theconceptually more highly developed domains of the biological, chemical andphysical sciences. Such a move, as Bhaskar points out, can only be explanatorilyadequate when the 'to-be-reduced science' is itself conceptually adequate. Giventhe current conceptual and epistemological inadequacies and confusions whichare rife in the social sciences, this means that the resulting reduction explainsvery little. Social meaning making (viz. semiosis) is only possible through theenactment of material processes which acquire meaning as social semioticpractices. The conflation of the two domains misses a number of crucial points:(1) the same social semiotic (c.f. social and psychological) processes may beembodied by 'lower'-level material processes in a plurality of different ways; and(2) there are multiple social semiotic ways of construing material (c.f. neurophysiological and other) states. Typically, such a conflation entails the followingkind of downwards reduction:

(a) Social semiotic relations and practices (social)

~(b) Psychological states and cognitions (mind)

~(c) Neurophysiological relations and processes (biological organism)

The problem with these reductive explanations is that while all social semiosisoccurs in accordance with physical-material processes and laws, social semiosis isnot completely explainable in terms of them. Again, social meaning-making isnot reducible to the intentions and 'inner' cognitive processes of the individualsocial actor. These processes, too, are constructed and construed out of the sameresource systems through which we make and exchange meanings with others insocial situations. The fact that these resource systems have been adapted tospecialized 'inner' uses does not mean that a separate 'psychology' of theindividual is required, seen as explanatorily independent of social semioticrelations and practices. What we call the 'psychological' in Western cultures is amethodological (and political) focus on the individual as an independentdomain of cause and effect. The principles of organization of the individualcertainly draw on and make use of a folk-theoretical repertoire of (individual)beliefs, intentions, wants, and so on, which form the hypostatic local accountswhich actors deploy in order to explain their own and others' behaviour. But thesenotions are really contextually organized and constructed relations and practices,not reducible to such hypostatic local accounts that are founded on an epistemology of mind, seen as an independent and reified domain of cause and effect.

Now, the meta-redundancy relations that constitute the different levels of acontextualization hierarchy are not related to each other by efficient causality(see also Halliday 1992a). As I pointed out above, me ta-redundancy constitutes


a system of constraints in which all levels operate on and constrain all others.The failure to understand the essentially contextual nature of experience,mental states, and so on, is evident in those accounts of the mind-brain relationthat try to account for this in causal terms. For example, Bhaskar, in order touphold the causal relation between 'mind-states' or 'belief and neurophysiological states or 'matter', argues that: '... such causality as holds between beliefsand matter must be conceived, in the absence of any connecting mechanism, astranscategorial' (1979: 132). Earlier in the same book, Bhaskar had pointed outthat this criterion is needed so as to overcome the objections that (1) cause andeffect must be embodied in a common medium or share common properties;and (2) the absence of any such medium constitutes an arbitrary break in thechain of cause and effect (1979: 131). Psychologists and others often use suchcriteria to justify an independent science of psychological or mind states on thegrounds that this constitutes an autonomous domain of cause and effect. Forexample, they may argue that P-states (intentions, wants, etc.) CAUSALLYACTUPON/AFFECT N-states (neurophysiology).

Why is it seen as so important to justify an autonomous psychology? Accordingto Bhaskar (1979: 108), the social meaning of an action - e.g. greeting someone- is recognizeable as being an act of a certain cultural type. This is seen as independent of the occasion-specific intention whereby it is performed by a givensocial actor. This then allows for a causal chain to be reconstructed such that thereasons (beliefs) which lead the actor to want or to intend to perform aparticular action may be postulated. The causal chain runs like this: (1) SOCIALACTION is caused by (2) ACTOR'S INTENTION is caused by (3)BELIEFS/REASONS LEADING TO INTENTION. Two separate dyadic series arethus postulated. But rather than say that I is caused by 2, and that 2 is caused by3, we can say, following Halliday and Lemke, that 1 redounds with theredundancy of 2 and 3, in a given context of culture, let us call it O. Note thatBhaskar wanted to separate this out from the intentional action performed bythe actor. When we put the ecosocial system back into the picture we can seethat: 0 redounds with the redundancy of (1, 2, and 3). For Bhaskar, agencyemerges from the psychological links between actors and the locally definedresources they use to interpret each others' intentions. This meets his criterionof synchronic emergence, which correctly recognizes that the higher-orderprinciples cannot be completely explained in terms of the lower-order ones(Bhaskar 1979: 125). Agency, then, is a psychological phenomenon, notreducible to neurophysiological states. But this causal way of reasoning struggles,as we have seen, with the problem of finding some linking mechanism transcategorial or otherwise - which can relate cause and effect. With the notionof meta-redundancy, the problems of form-substance dualism from which thisderives are irrelevant. Let us now see why.

Earlier, I referred to the complementary dual nature of the relations betweenthe semiotic-discursive and the physical-material domains. Bhaskar attempts toovercome the problem of reductionism with the notions of transcategorialcausality and synchronic emergence. His attempt to do so relies on the ontological stratification of the domains of the social, the psychological, and thebiological. The resulting localization of these domains, along with Bhaskar's

INTRODUCTION 33

(transcendental) realism, means that these domains are seen as further decomposable into the given reality described and the way this is talked about. Thecomplementarity of our dual perspective means, on the other hand, that thereis no essential separation of these domains. Instead, which level of reality isconstrued depends on which level of contextualization and which operationsthis performs on the relevant system of relations. This approach accords withBohr's (1948) principle of complementarity in quantum mechanics: there is nosingle or global perspective which can reveal an 'objective' description of acomplete system. Instead, there are multiple construals and contextualizingprinciples which operate on different levels of the system, thereby giving rise toa plurality of complementary ways of construing this 'reality', without, however,being reducible to any single principle. Thus, we can say that the socialredounds with the redundancy of the psychological with the redundancy of theneurophysiological.

The complementary duality of semiotic-discursive systems and practices andphysical-material relations and processes does not mean (1) that actions arecaused by psychological states (intentions, ete.), which are caused by neurophysiological states, but rather (2) that actions redound with the redundancy ofpsychological states with neurophysiological states. The emergent properties canbe explained not in the language of (transcategorial) causality, but in thecontextual language of me ta-redundancy (chapter 6, section 7, pp. 300-1). Theunits and their relations on the psychological level redound with another set ofunits and relations on the neurophysiological level, so that the units andrelations on the higher-level ecosocial semiotic level redound with theredundancy of the psychological with the neurophysiological. The psychologicallevel of analysis, as McKellar (1990: 337) points out, is really a contextualizationof the functional, open, and adaptive interface between the biological (neuroanatomical, neurophysiological, and neuropsychological relations andprocesses) and the social semiotic relations with which it both eo-evolves andcontextually redounds. Thus defined, there is no autonomous or independentdomain of psychological theory; the psychological domain is always definable inrelation to the two levels of the biological and the social semiotic mentionedabove. This does not alter the fact that in Western culture an ontology of the psychological is constructed in and through the social practices of both our official('scientific') and folk-theoretical (,common-sense') accounts of the individual.But this is a local cultural hypostatization.

Where do notions such as agency, subjectivity, and so on, come from? Thecausal species of reasoning which we have examined here locates this in the(ontological) domain called the psychological, seen as part of a causal chainwhereby intentions, motivations, and so on, are derived, explained, andattributed. The ontological stratification of the domain of the psychological insuch accounts does not distinguish between the accounting procedures and folktheoretical (common sense) explanations which derive from the local hypostaticculture of the informants, and wider explanatory claims of a truly 'scientific'psychology.

In other words, given the fact that informants regularly deploy a repertory ofnotions such as 'self', intentionality, agency, and so on, what is the relation


between the use of these terms in the theory and in the folk-theoretical accountsof the culture it seeks to explain? This is a major, though unaddressed problem,in those psychological, philosophical, and other accounts which are based on thekind of ontological stratification that I discussed above. Such theories use acausal-deductive model of explanation, which is not formally recursive.This means that the explanatory bases and me ta-theoretical assumptions ofsuch theories and explanations go undisclosed and are not subject to metatheoretical scrutiny (see Thibault 1991a, 1999a, 1999b for further discussion).

8. Defining and Extending the Notion of Meaning in Terms of theThree-level Hierarchy and the Specification Hierarchy

The notion of 'meaning' can be thought about in terms of a system comprisingthree levels. Meaning is not to be confused with the notion of information ininformation theory (Shannon and Weaver 1949). The latter notion is concernedwith how all forms of information can be quantified in terms of common unitssuch as bits or bytes (see Bateson 1987 [1951]; Wilden 1980 [1972]: 233-6;Lemke 1995c: 169-70). Information theory is concerned with the statisticalfrequency with which different units and their combinations come about. Information theory is not, therefore, concerned with the interpretation of meaning.Information is independent of particular observer perspectives. When we focuson meaning, on the other hand, we are interested in the differences between thediverse semiotic modalities and the different ways in which their distinctivepatterns of difference - whether topological or typological - are organized so asto form a distinctive kind of semiotic system.

To go back to our previous example (section 6, pp. 26-30), we are concernedwith the different sets of alternatives, the different principles of their combination, the different contexts with which they redound, and so on. Traditionally,semiotics, or the study of signs, has been concerned with the interpretation andclassification of signs (Noth 1990). However, the term 'sign' will not be centralin the present study. The pervasive and uncritical acceptance of the notion of thesign as something which calls to mind something other than what it is has failedto clarify the ontological status of the concept of the sign or the ways in whichthis concept relates to processes of meaning-making and their textual and artefactual products (see Bouissac 1998: 737). The fundamental question in thecontext of the present study has to do with the contextualizing relations that weconstruct and enact in order to recognize pattern and organization in ourexperience and in our interactions with the world beyond the self, i.e. thenonself, including other selves.

In terms of the logic of the three-level hierarchy, interaction between an interpreter and some portion of the world is the focal level. This is the level on whichthe self encounters the nonself. The self qua self-organizing system is a firstnessin Peirce's sense. A firstness has the characteristic of existing in the world and ofhaving particular perspectives on the world by virtue of its position in the world.But firstness does not exist alone; it is always defined in relation to that withwhich it interacts in the world, viz. secondness. However, the interaction between

INTRODUCTION 35

firstness and secondness occurs and is made possible by the existence of ahigher-scalar system which mediates the relations between firstness andsecondness and which makes the interaction between them possible. A higherscalar system of this kind, e.g. a system of interpretance, is an example ofthirdness. The relations between firstness, secondness, and thirdness can bedescribed in terms of the three-level hierarchy system of relations. I shall nowexplain how Peirce's categories can be assimilated to the logic of the three-levelhierarchy in order to provide some very general principles for thinking aboutthe notion of meaning.

The Italian bio-chemist and semiotician Giorgio Prodi has shown that anyinteraction between an interpreter and some change in a given materialphenomenon presupposes some relation of correspondence between thephenomenon observed and the interpreter. Observation, Prodi (1977: 20)points out, is always a participation in the processes of change: the interpreter,in other words, is always altered by the phenomenon which it observes. Thisdiscovery of the mutual interference of observer and observed was first made inthe early years of the twentieth century by advances in the theory of quantummechanics, as developed by N. Bohr and W. Heisenberg. To a given variation insome object, there corresponds a correlated variation in another object, or seriesof objects. In all such cases, these correspondences occur through the modification of a system of relations, i.e. through exchanges of energy. This basic pointis so whether we are talking about a biochemical interaction between twomolecules, or the mother-infant dyads in and through which early infantsemiosis emerges (see below).

Any physical-material perturbation in a given environment constitutes acatastrophic release of a certain quantity of energy. This release of energy mayimpinge upon the interpretative apparatus of some interpreter. The perturbation in the environment that is entailed by a given release of energy brings abouta corresponding change in the interpretative apparatus of the interpreter. Theinterpreter is biologically adapted to construe certain kinds of perturbation in itsenvironment as being in some way significant to its further development andsurvival. Observation, so defined, is not 'outside' or 'external to' the observedevent. Instead, observer and observed jointly participate in the event, and in wayswhich change the observer-cum-interpreter. In other words, the processesinvolved are transactional in character.

Prodi also discusses more complex 'situations of correspondence' in whichthe exchange of energy between, say, A and B is conditioned by the reciprocaladaptation to each other of the two terms in the exchange process. Thus, thematerial correspondence which triggers the exchange of energy presupposessome more specific relation of 'reciprocal adaptation' between A and B. Prodi(1977: 22) emphasizes two important facets of such processes: (1) the release ofenergy is conditioned by the existence of a specific (interpretative) 'key'; and (2)the triggering of the exchange of energy is subordinated to its being in contactwith this key. It is clear here that A and B, taken separately, are formallyequivalent to what Peirce called firstness. The interaction - the reciprocaladaptation - between A and B is a secondness. Moreover, the interpretive 'key'that regulates and mediates this reciprocal adaptation is a thirdness.


Now, the phenomena that we are discussing here are very general in natureand are common to all biological and social systems. This is indifferently sowhether we are talking about two molecules, A and B, in a reciprocal relation toeach other, two amoeba, or the two members of a parent-infant dyad in a humansocial system. The basis of this reciprocal adaptation - the exchange of energyand the establishment of material correspondences between A and B - is thesame in all cases. What is different is the level and degree of the complexityinvolved, as well as the extent of the socio-historical processes, in the case ofhuman, and other, social systems which have further elaborated the dynamics ofthe exchange processes involved and the ways in which these cross-couplephysical-material and semiotic-discursive processes.

The prerequisites of this process are as follows. First, there must be twomaterial objects which interact. In so doing, they are transformable. It is not thecase that two objects happen to interact by chance, as it were. Again, we can notethe principle of the contextual redundancy of the organism-plus-environmentsystem that Bateson pointed out (see section 6, pp. 26-31). The three-levelhierarchy is useful in sorting out the levels of relations involved. Thus, the A andB parties to the interaction are on the focal level L. This level is subordinated tosome higher-order principle of correspondence, or of complementarity - cf.Prodi's interpretive 'key', or some higher-scalar system of interpretance - on levelL+1. Such a principle is a contextual relationship which specifies that in a givencontext-type, A and B select each other as a function of that context. There is, inother words, a contextually governed relation of complementarity between A andB, which, in selecting each other, remain indifferent to other potentially availableinformation in their environment. That is to say, the resulting A-B dyad discriminates and selects contextually relevant information from its environment inspecies-specific ways, which are relevant to its further development and survival.

Secondly, whatever principles of 'order' or 'structure' are derivable from sucha me tastable complex of relations are completely immanent in them. These donot simply reside in either A or B, taken 'separately'. Rather, A and B are contextually organized in relation to each other. This obviates the need for someextrinsic metaphysical principle of order (intention, purpose, or plan) which thestructure simply contains.

Thirdly, the relation of complementarity between A and B is always regulatedby material forms (substrates) and thermodynamic potentials which constitutethe intrinsic dynamics of A and B and which make the exchange processphysically possible. Such material substrates and thermodynamic potentials arethe affordances on level L-l that I discussed in section 3 (pp. 11-18). Thegeneral principles discussed above can be extended to the human case. Toillustrate this, I shall now discuss the dyadic interaction that characterizes jointmother-infant semiosis in the phase that Trevarthen has defined as 'primaryintersubjectivity' (1978, 1987, 1992).

Trevarthen (1987: 184) describes the way in which newborns seek out eyecontact with significant others (parents, caretakers). For Trevarthen, theseactivities constitute the very earliest stages of dialogic interaction. Suchexpressive movements on the part of the newborn 'trigger' reciprocal and corresponding acts on the part of the caretaker, who construes the newborn's

INTRODUCTION 37

movement as having some meaning relevant to the emerging dyad. For this to beso, it means that the caretaker must construe the newborn's movement perhaps the establishing of eye contact - as having a form which regulates theexchange. The form which the caretaker construes in the newborn's movementelicits a complementary action on the part of the caretaker. The relevant formemerges out of the joint, or dialogic, act of meaning, in and through whichforms are construed in the material processes (bodily movements and orientations) which always implicate the release of energy.

This process is always a semiotic one. The two participants in the emergingand developing dyad are dialogically oriented to each other insofar as the dyadis governed by higher-order contextualizing principles which regulate theexchange. The child turns his or her gaze towards the caretaker. This action constitutes a release of energy which impinges on the consciousness of the seniormember of the dyad. In so doing, the caretaker construes this release of energyin ways which bring about complementary changes and responses on the part ofthe caretaker. As embodied beings, A and B are entities whose materiality, alongwith the exchange of energy in and through the interaction of A and B, providethe only basis whereby a given form may be construed. The embodied beings Aand B enter into a structured relation of complementarity which cannot but existwithout a particular spatio-temporal configuration and a particular materialform.

In the case of the early mother-infant dyads studied by Kaye, Trevarthen, andHalliday, the dialogic exchange process means that the two members of the dyadengage in a reciprocal exploration of and a selective attention to the environment in which the dyad is formed. In the first instance, this means the reciprocalexploration and selective attending to each other as embodied beings in thespace-time of the dyad. In this way, as Halliday (l992a) shows, the earliestnotions of 'you' and 'me' emerge. It is only after this stage that the two membersof the dyad jointly explore and attend to the physical-material environment, ofwhich they are always a constituent part, in any case. What is important toemphasize here is that this exploration, too, can only occur through theresources provided by the dyad. Thus, the two members of the dyad, in theprocess of being constituted as 'you' and 'me', are reciprocally oriented to eachother in a process of mutual exploration and interpretation (see also Bates1979).

The child's attention-seeking gaze is an act of exploration. It is both aphysical-material act performed by an embodied being in space-time, as well asbeing a release of energy. Such an act is not random; it is an exploratory actwhich is oriented to getting a significant response from the other member of thedyad. In this sense, it is a 'reading' of reality, because it is oriented towardsselecting a significant response from the other. The newborn explores the realityof his or her Umwelt through the dyad and in so doing he or she also seeks acomplementary response from the other. When that happens - when A's dialogicorientation to B produces a significant complementary structure or orientationin B (a response, in other words) - then a specific change may be said to havebeen induced both in the overall organization of the dyad and in the internalstructural organization of its members. A's dialogic act is a specific intervention


in and an orientation in the space-time of the dyad. While it is true to say, asHalliday does, that the resulting interact has no content in the adult sense, itdoes, nevertheless, constitute a judgement on - a proto-modalization of - thespace-time of the dyad and on the nature of the relationship between the twoparties to the transaction. It both intervenes in and comments on that relation.As an intervention, it plays a significant part in the co-ordination and channelling of matter, energy, and information flows in the dyad. As a commentary,it provides an implicit judgement - a meta-communication - about the relationsbetween the two members of the dyad.

It is the exchange process which both enables and facilitates the progressiveself-organization and individuation of both members of the dyad. The membersof the dyad, as I noted earlier, are embodied social beings. However, this doesnot mean that they reduce to mere physical points in Newtonian space-time.Rather, they are also internally complex beings (Ham'; 1983: 75), capable offurther internal elaboration and self-organization. This point is not in contradiction with the fundamentally dyadic and dialogic nature of the organizationalprinciples involved. The dyad and the exchange processes it enables are fundamental. Without these, there would be no exchange processes, no principles ofself-organization, no 'individual' members. In other words, the processes,relations and structures involved are fundamentally transindividual andultimately socio-cultural in nature. This does not imply an opposition betweenthe biological and the social dimensions of these processes. There is no contradiction here because the biological processes are themselves, at all levels oforganization, attuned to and consistent with the socio-cultural processes towhich they are integrated and entrained. Thus, any equation of the biologicalorganism with the 'individual', and of the social with the 'collective', is falsebecause there is no contradiction, or discontinuity, between the two. Our biochemical, neuroanatomical, and neurophysiological makeup predisposes us toparticipate in complexly interrelated processes of matter, energy, and information exchange at all levels of organization (Prodi 1977; McKellar 1987, 1990).

Now, the last example discussed here most clearly applies to the specificallyhuman systems in which meaning is made, stored, and interpreted. But if wecompare it to the two examples that I have derived from Prodi, we can concludethat meaning, as seen from our human perspective, is a more highly specifiedsubclass of progressively less specified subclasses that stretch all the way backthrough the symbolic capacities of bonobos such as Kanzi and Panbanisha(Savage-Rumbaugh et al. 1998; Benson et al. 2002, 2004; Thibault 2004a) to theproto-language of other mammals such as dolphins, cats, and dogs, right back tothe 'situations of correspondence' between, say, an amoeba and its food source.The notion of meaning is a highly specified category which we normally reservefor the kinds of semiotic transactions with others and the perspectives theseafford that characterize oursemiotically mediated transactions with the non-self(cf. secondness), including our conspecifics.

However, the progressively more general, or less specified, cases mentionedabove show, albeit to differing levels of specification, that our human notion ofmeaning (semiosis), rather than being something which is unique to andcriterial for our perspective, is, in actual fact, a subclass of a very large number

INTRODUCTION 39

of progressively less specified subclasses. In saying this, I am adopting the logicof the specification hierarchy, as defined by Salthe (1993: 52-74; see also Lemke1995a: 93-6; Thibault 2000a: 298-300). Thus, each level in the specificationhierarchy is what Salthe (1993: 7Q-4) calls an integrative level: more-specifiedlevels integrate less-specified levels to their own perspective, as well as beinginformed by the lower-level perspectives which make up the whole hierarchy.This means that meaning, as understood from the specifically human perspective, includes and integrates to its own perspective, progressively vaguer, moregeneral classes which we might refer to, starting with the human case, as workingbackwards, as follows: meaning ... [meaning} ... IImeaning}} ... IIImeaning}}},ete. The unbracketed term 'meaning' here applies to the most specified case, viz.human semiosis: the use of single, double, triple, ete. curly brackets serves todesignate progressively more general, vaguer analogues for this term as weprogress towards the most general, least specified analogues of meaning.Similarly, we might say that the reciprocal adaptation of A and B in the examplediscussed above is a less-specified case of IIldialogue}}}, as seen from the perspective of human serniosis.

9. The Signifying Body: Rethinking the Stratified view of Semiosis in Termsof the Three-level Scalar Hierarchy

The notion of the sign has been central in serruoucs for many centuries.Generally speaking, signs are defined in terms of their stable and unchangingproperties, rather than in relation to the processes in which they participate andthrough which they are defined. That is, signs are defined as stable things whichsimply are. This fact is evidenced in the ways in which signs are classifiedaccording to properties of their substance (acoustic, manual-brachial, visual, andso on). In such a substance-based view, the fundamental unit of analysis is somekind of entity or particle. This view mirrors the concerns of the reductionistproject of the classical view of how the world is structured and came into being.For example, the model of hierarchical constituency in language reflects thespatial metaphor according to which larger-scalar or higher-level units (e.g.sentences, discourses) are de-composed and analysed in terms of interactionsamong smaller-scalar units and subsystems. The spatial modelling of theserelations views them as a hierarchy of relations among units on progressivelysmaller (or larger) scales such as sentence, phrase, word, morpheme, and so on,on the grammatical level. Analogous hierarchies are postulated for the phonological level.

In general, smaller-scalar units are seen as the building blocks out of whichlarger-scalar units are assembled in a bottom-up, aggregative fashion. Processand time are largely absent in this view. Such views of the sign are not concernedwith the dynamics of time-bound processes in and through which 'signs' emergein semiosis. In a dynamical, process-oriented view, meaning-making (serniosis)can be understood as the integration of faster, smaller, more local processes intoslower, larger and more global ones (Salthe 1993: 46; Lemke 2000b: 275;Thibault 2003a: 137-8). The linguistic notion of the stratification can be refor-


mulated in the light of this understanding. But first, a few words concerning theway in which stratification is usually understood in linguistics with respect tospoken language.

The stratified model of the language system takes a given linguistic form ascomprising various interconnected layers of organization. For example, it canbe shown that the phonological organization of / dogz/ is a single syllablecomprised of four phonemes, but that this does not correspond to its grammatical organization in any direct way. Grammatically, it is a morphemic wordcomprising the two morphemes [dog] + [z]. The various layers of organizationphonological and morpho-syntactic in the present example - are called strata,following Hjelmslev (1954), Lamb (1966), Halliday (1979), and Martin (1991).The various strata are related to each other by means of realization relationships(themselves a sub-class of the more general meta-redundancy relations thatoccur among the diverse levels of a given contextualization hierarchy). Realizatory relationships functionally relate the selection and organization of units ona given stratum to the strata 'above' and 'below' it. Realization relationships areinterstratal coding relationships. Each stratum also has its own principles oforganization for the units and structures that occur within that particularstratum. These are referred to as tactic relationships.

In stratal terms, we can define a given unit in functional terms along threedifferent dimensions: (1) its tactic relations with other units on its 'own' level;(2) its realizatory relationships with the stratum 'above'; and (3) its realizatoryrelationships with the stratum 'below'. Realizatory relations are not one-way ordeterministic, but are best seen as a two-way productive dialectic, whereby unitsand relations on a given stratum are constructed and construed in relation toprinciples of organization on some other stratum. Interstratal realizatory relationships are both dynamic and open. For example, the morphosyntacticstructure [[dog] + [z]] realizes the meaning - the semantics - of this compositegrammatical structure at the same time as the given meaning is realized by thegrammatical structure in question. Likewise, the morphosyntactic structure isrealized by a given phonological structure at the same time as the phonologyrealizes the relevant grammatical structure.

Rather than falling back on shop-worn definitions of the sign that have beenrefined and reinterpreted according to the assumptions and requirements ofvarious philosophical and theological discourses since classical antiquity to theeffect that the sign is something which calls to mind something, or that it issomething which stands to someone for something (see Bouissac 1998: 737), wecan reconsider the stratified view of semiosis in terms of the three-level scalarhierarchy view of complex, dynamic open systems (Salthe 1993: 36-46; Lemke1995b: 106-12; Thibault 2000a: 296-8). Thus, a spoken utterance consists ofpatterns of neuromuscular (vocal tract and other) activity whereby the distinctive sound patterns of the particular language are produced. These patterns ofneuromuscular activity take place on a very fast timescale relative to the words,phrases, clauses, and so on, on the next higher level of the proposed hierarchyofrelations (see also Cowley 1998: 547-51).

For example, Lemke points out that the articulation of a typical English voweltakes about one-tenth of a second (Lemke 2000b: 276). Below the level of artic-

INTRODUCTION 41

ulation, there are still smaller, faster timescales such as the activation of theneuronal networks that co-ordinate muscular activity on the timescale of tens ofmilliseconds, and below this there is the timescale of the neurotransmitter flowswhere events have a duration that can be measured in milliseconds (see Lemke2000b: 276). The events that I have referred to here occur on a timescale that isbelow the level of human awareness. Events on these timescales cannot beobserved by humans except by technologically enhanced means. The point isthat events on these very small, fast timescales do not belong to the environmentwhich we inhabit and in which we interact with others. They are not 'cogentmoments' on the ecosocial scale where our own observer perspectives are to befound. At all levels of organization both 'above' and 'below' the focal level of agiven analysis, there is organization of process on different space-time scales,though there is no reduction of processes on higher levels to localized andatomized substances with causal powers on lower levels of organization(Bickhard and Campbell 2000: 330-1). Meaningful pattern and organization onhigher levels is not supervenient on some ultimate, substance-based level ofatomized causal factors. Instead, it is the reorganization across scalar levels ofpattern and organization of lower levels such that what constitutes meaning toour observer perspectives on our scale - seconds, minutes, hours, days, weeks,... years - is pattern or organization of process that is relevant to our own scale.Consider the following example:

L+1: discourse, dialogic exchange

L: dialogic exchange units

L-1: vocal tract articulatory (neuromuscular) activity

Following the logic of the three-level hierarchy view, the middle level of anygiven triplicate of levels is the focal level (i.e. level L). The focal level is the levelon which the processes which represent the focus of interest in a particularanalysis occur. In the present example, level L consists of dialogically coordinated utterances, realized by lexicogrammatical units of such as words,phrases, clauses, and so on. However, the lexicogrammatical patterns that areobservable on level L are themselves made possible by and depend on processeson level L-l. The neuromuscular patterns of activity on level L-1, as Lemke(2000b: 277) points out, constitute affordances for processes on level L. By thesame token, level L processes are constrained by longer-timescale processes onthe next higher level L+1. In the present case, level L+1 represents an emergingpattern of discourse-level activity. The emerging pattern on this level is able toinfluence the probability of patterns and processes on the next level down,thereby acting as a set of boundary conditions or constraints on the lower level.

The three-level hierarchy principle, as applied here to linguistically mediatedsemiosis, illustrates the principle whereby processes on different timescalesbecome cross-coupled in more complexly interdependent systems with manyinteracting variables on many different levels of organization. In the newborn'svocal tract, there are potentially very many degrees of topological freedom such


that, within the constraints afforded by biology, different patterns of neuromuscular activity are possible. Over time, the infant develops an increasingly largerepertoire of possibilities of the co-ordination of vocal-tract muscular activity byneuron impulses. The co-ordination of these possibilities acts as catalysts for stillmore patterns of activation. A situation is then reached in which a relativelyrestricted set of patterns of activity in the form of regular, repeatable vocal-tractgestures emerges (see also section 11, pp. 49-54).

Consequently, other possible patterns that are potentially afforded by the neuromuscular substrate tend to wane as the activation of the substrate is entrainedto the newly emergent level of organization. The newly emergent pattern ofvocal-tract gestural activity is afforded by the topological variety of the neuromuscular substrate, though it is not determined by this. The newly emergentlevel of gestural activity takes place on a longer, slower timescale with respect tothe neuromuscular patterns of activity on the next-lower scale. This means thatpatterns of activation on the lower level are filtered or buffered by the newlyemergent level of organization. Information on the lower-scalar level of neuromuscular patterns can only be meaningful to longer-scalar L+1 processes such asthe dialogic exchange of meanings between mother and infant after beingfiltered by the newly emergent level of the vocal gestures on level L. Thus:

L+1: protolinguistic dialogic exchange

L: vocal gesture repertoire

L-l: neuromuscular patterns of activity

In infant semiosis, the emergence of level-L vocal gestures alters the probabilitylandscape of the lower-level neuromuscular patterns by entraining them to itsown dynamics. By the same token, the emergence of proto-linguistic vocalgestures in the infant on level L itself provides a basis for the emergence of thenew intermediate level of organization of lexicogrammar between the previouslyexisting levels L-l and L+1 of the proto-linguistic stage. In this way, previouslydiscrete functional components - cf. Halliday's (1975) discrete micro-functionsin infant protolanguage - become co-ordinated and reorganized into a morecomplex and internally differentiated system of multiple, overlapping functionalcomponents which greatly enlarge the phase space of the system and hence thevariety of contexts it can access on an increasingly diverse number of space-timescales beyond the here-now interactional event (see also Silverstein 1985).

The two examples discussed above show the timescales that operate above andbelow the focal level of utterance and proto-linguistic gesture, respectively. Eachlevel in the hierarchy represents the dynamical processes and the timescale thatare characteristic of a particular type of event or activity, relative to its scale. Withrespect to focal-level processes on level L, the processes on the next level down,viz. level L-l, provide principles of organization that make possible the patternsof organization on the next higher level L. At the same time, the patterns oforganization on the still next higher level L+1 constitute higher-scalar levels oforganization which provide the context for lower-scalar processes and their

INTRODUCTION 43

interpretation. Higher-scalar processes can thus be seen as deploying lowerscalar affordances in the direction of particular semiotic attractors. Neuromuscular activity is meaningful insofar as it is integrated to and interpreted as partof an utterance, rather than being a mere movement of the vocal tract in its ownright. In its turn, an utterance is integrated to and interpreted as part of anunfolding discourse exchange between the interlocutors in that event.

The principles that I have outlined above draw attention to the ways in whichsemiosis, or social meaning-making activity, violates the principle of the separability or insulation of non-adjacent timescales. This is the principle which Lemkehas formulated as scalar heterogeneity in contrast to the scalar homogeneity ofsystems that are 'described solely in thermodynamic or biological terms' (Lemke2000a: 183). Systems of the latter type are characterized by the way in which theontological isolation of levels in scalar hierarchical systems ensures that'dynamics at different levels are screened off from one another, with descriptivevariables necessarily representing a single level only, ... ' (Salthe 1993: 45). Eachlevel has its own specific characteristics; hence the ontological isolation of thecomponent levels of the system (Salthe 1993: 46). This reflects the principle ofthe 'nontransitivity of effects across levels' (Salthe 1993: 45).

Lemke (2000a: 182) points out that the properties of scale homogeneity, suchthat the ontological isolation of levels is the norm, is problematic in the case ofhuman ecosocial systems, which are characterized by cross-couplings of physicalmaterial processes and interactions and semiotic-discursive ones which mediatehuman activity in the overall ecosocial system. In my view, the structuralistreading of the semiotic notion of stratification has overemphasized the nontransitivity of the relations across different strata. This view had led to theassumption that descriptive variables pertaining to any given stratum - e.g.phonology, morphology, syntax, semantics in traditional accounts - pertain toonly one level. In this account, the variables on any given level constrain those onother, especially adjacent, levels, yet the abstractions of phonological andmorphosyntactic form, on their particular levels, are not connected to bodilyprocesses on other, lower-scalar levels. The structuralist account of the relationbetween phonology and morphosyntax is an example. In this account, phonological distinctions, which are not considered to be meaningful in any way,merelyfunction to constrain and specify morphosyntactic distinctions on their level.

Thus, human cultural systems cannot be adequately described in terms ofscale-homogeneous systems whose regularities can be described in terms of thermodynamic and biological processes per se (see above). The emergence oftypological-categorial modalities of serniosis, Lemke (2000a: 183) points out,quickly leads to a situation in which radically different spatial and temporal scalesintersect in meaning-making activity. In this way, the here-now event of neuromuscular activity in the vocal tract, for example, implicates meanings on manydifferent space-time scales - e.g. the very fast timescales of neuronal processes,the neuromuscular timescale of vocal-tract articulation, the timescale of theutterance qua meaningful event exchanged between interlocutors, the conversational episode, the intertextual links between today's conversation and the oneheld last week with the same person, the historical-biographicallifespan scale ofthe individual, the history of the community in which the individual lives, the


history of the culture, the language itself and the ecosocial system in which thelanguage is embedded (see Lemke 2000b: 275-9). In semiosis, particularly of thetypological-eategorial sort, all of these scales may be implicated in ways that areseamlessly interwoven with the here-now event or occasion of interactionbetween two or more individuals.

The two hypothetical examples discussed above are intended to show howhigher-scalar dynamical processes on their timescale have constraining effects onmuch shorter timescales (Lemke 2000b: 280). For example, the much longer,slower timescale of the emergence of semiotic action formations - i.e. thediscourse genres, activity-structure types - in a community have constrainingeffects on the shorter, faster timescale of the neuromuscular processes of thehuman vocal tract. The human body has signifying potential because it is asemiotic-material participant in dynamical processes on diverse timescales. Forexample, fast small-scale neuromuscular activities of the body on, say, the herenow scale of a given interpersonal exchange between two persons are attractedto and integrated with the socio-cultural systems of meanings and practices thatpertain to processes on much larger timescales, as compared to the immediatecontext of the body's material interactivity with its here-now environment (seeLemke 2000a: 194).

Human bodies qua material artefacts circulate in social networks, participating in some kinds of activities and not others, according to one's social positioning in these networks and their associated activities on the basis of factors suchas sex, age, profession, social class, leisure activities, consumer patterns, lifestyle,and so on. Taken as a whole, the body's participation in various activities in thesenetworks, the movement from one activity to another in the course of shiftingfrom one context to another, amount to a longer-timescale semiotic process, ascompared to any of the individual shorter-scale activities that the individual participates in. This longer-timescale semiotic process will be defined in this book asthe individual's lifespan or historical-biographical trajectory.

The body qua artefact is a material participant in processes - activities - thatafford its semiotic integration to and interpretation by higher-scalar systems ofinterpretance that persist on time scales that go beyond the material body itself.Brains function to co-ordinate, to cross-couple body-brain systems to theecosocial environment by means of the mediating effects of semiotic artefacts,including objects, textual artefacts, tools, technological aids (computers, calculators, slide rules, OHPs, and so on), architectural features, human bodies (bothour own and those of others). It is through such processes of semiotic mediationthat the trajectories-in-time of selves and their dialogic intertwining with thetrajectories of other selves come into view on various scales - semiogenetic,ontogenetic, historical-biographical, and so on.

While I agree with the critique put forward by integrationallinguists such asHarris (l995a) and Cowley et at. (In press) that language tokens are not realmental entities that are 'represented' by neural processes in the brain, I alsoargue for the need to reframe the counter-elaim put forward by Harris and otherintegrationalists to the effect that linguistic forms and meanings are 'secondorder entities arising through the codification (see Harris 1995a) central to bothlay and professional metalinguistic analysis' (Cowley et al. In press). We must be

INTRODUCTION 45

careful here not to think that language forms and meanings are mere appearances with respect to some truer or more real underlying reality such as, forexample, biomechanical activity. In saying this, I am not denying that both layand professional me ta-linguistic analyses construct linguistic 'objects', as indeedSaussure (1971 [1915]: 23-7; 1993: 187-8) well understood.

My point is that language forms, just like neuromuscular activity, have theirown patterns, which are distinctive to their own level of organization, rather thanbeing false to some truer underlying reality. I have suggested that languageforms and meanings are integrated with bodily processes and can be understoodto be the result of these same processes. At the same time, language forms andmeanings have properties and relations that cannot be reduced to or explainedwholly in terms of the lower level. Thus suggests that (1) words emerge fromneuromuscular activity when neuromuscular activity is entrained to certainpatterns rather than others; and (2) the emergent patterns make a difference,i.e. are semiotically salient in some community or social network. Point (2)further suggests that language forms have the potential to make differences thatbear directly on our own experience, as well as on the experience of others. Inthis way, we see that language forms and meanings have the capacity to entrainand direct physical-material processes and flows, including biological ones, byvirtue of the properties and relations that are specific to their own level oforganization.

The fact that the practices - lay and professional - of meta-linguistic codification accord properties to second-order entities such as sentences suggests thatlexicogrammatical forms and their meanings are themselves treated as secondorder artefacts that circulate in social networks. This does not mean that thesesecond-order artefacts float free of or transcend their material embodiments in,say, neuromuscular activity or visual tracings on a treated surface (themselves theextra-somatic products of neuromuscular activity) What it does mean is thatlexicogrammatical forms and their meanings are more abstract semiotic artefactswhich our meta-semiotic practices, in part, de-locate from the full multimodalsystem in which they are deployed and treat as objects of meta-semioticreflection and manipulation in their own right.

This perspective on multimodality in the mature system may be compared tothe early-infant system where the separation into distinct semiotic 'modalities' language, gesture, gaze, movement, and so on - is problematic on account of theinherent topological vagueness and relative indeterminacy of the infant system.Cowley et al. (In press) refer to the infant's 'full-bodied sense making' to capturethe fact that in early-infant semiosis this separation into diverse modalities andtheir combinations and recombinations has not yet occurred. This is so becausethe lower-scalar affordances of the immature system are sub-personal in thesense that they have not been fully contextually integrated to the perspective ofthe agent, as in the mature system (see section 3, pp. 11-18).

In the mature system, on the other hand, diverse semiotic modalities havebecome discursively, or semiotically, determinate. Therefore, it is possible to talkabout in some meta-serniotic discourse - either lay or professional - language,gesture, gaze, movement, and so on as 'distinct' systems. Here, too, this does notmean that the vagueness and indeterminacy of the earlier system have been


transcended or superseded. Rather, the properties and effects of the earlier,immature system have receded further down the scale of implication such thatthe diverse modalities can be distinguished and described according to therequirements of the meta-discourse. Moreover, the increasing determinacy ofthe diverse semiotic modalities also allows for individuating patterns wherebysome (topological) differentiations make a difference concerning individualidentity, personality, and so on. The arguments advanced here further suggestthat the term 'rnultimodality', as applied to the semiotic resource systems of themature system, is a meta-semiotic term that best captures the increased determinacy of specific resources in the mature system and, at the same time, the termpresupposes the increasing capacity to deploy and to (re)combine specificmodalities in ways that display a relative independence and distinctiveness withrespect to the primordial vagueness and indeterminacy of the immature system.

10. Systemic-functional Linguistic Theory: Bringing Together theIntra-organism and Inter-organism Perspectives on Meaning-making

The present study draws on and is informed by the systemic-functional theory oflanguage which has been developed by Michael Halliday and colleagues over thepast forty or so years. A central tenet of this theory is that the internal organization of language - its phonological, graphological and lexicogrammatical form- is shaped by the diverse functions which phonology, graphology and lexicogrammar play in the construction of discourse - spoken and written - in its socialcontexts of use (Halliday 1994 [1985]; Matthiessen 1995). This claim hasentailed a shift away from the sentence-based units that characterized formalgrammars, which were mainly based on assumptions about written languages. Ithas also meant a focus on the fundamentally dialogic nature oflanguage and onthe ways in which discourse contributes to and maintains a sense of the semioticwholeness and unity of particular discourse contexts. In so doing, systemicfunctional theory has helped to redefine what grammar itself is and does.

In this perspective, grammatical patterns and relations and grammaticalcategories are seen as standing in an intrinsic and motivated relation with theactivity-structures in which the former are deployed and embedded. Thus, thedescription of grammar and the description of meaning-making activity are seenas commensurate - the one is responsive to and sensitive to the analytical anddescriptive requirements of the other. The relationship between the internalorganization of language form - e.g. its lexicogrammar - and meaning has ledHalliday and others working within the systemic-functional framework topostulate the existence of a small number of diverse functional regions known asthe metafunctions in order to explain the always fluid, dynamic, and contextualized nature of the ways in which language forms relate to meaning in context.

What does the term 'rnetafunction' mean? Halliday (e.g. 1979) has proposedthat the content stratum of language - viz. its lexicogrammar and semantics - isinternally organized into three or four different kinds of very general functionalregions which are simultaneously interwoven and configured in the internalorganization of this stratum. He has glossed these as the experiential, interper-

INTRODUCTION 47

sonal, textual, and logical dimensions of linguistic meaning, respectively. Therehave also been further developments and extensions of this principle withrespect to semiotic systems other than language (see ]ohnston 1992; Kress andVan Leeuwen 1996; Lemke 1998; Thibault 1994, 2000b, 2001; Van Leeuwen1999).

Experiential meaning interprets the phenomena of the world as categories ofexperience, as configurations of, for example, clause-level process-types (actions,events, states, and so on), the participants that take part in these, and the circumstances that may be attendant upon them. Experiential-meaning relationsare realized in the grammar as particulate or part-whole structures which arebased on the principle of constituency. Interpersonal meaning is concerned withthe grammatical resources for organizing language as interaction (c.f. speechacts, dialogic moves), the expression of attitudinal and evaluative orientations(modality), and the taking-up and negotiating of particular subjective positionsin discourse. Typically, interpersonal meaning is expressed by field-like prosodiesrather than particle-like segments. Textual meaning is concerned with the organization of language into semantically coherent text in relation to its context. It isconcerned with the distribution of information in text, continuity of reference,and lexico-semantic cohesion. Textual meanings tend to be realized by wave-likepeaks of prominence. Logical meaning is concerned with relations of causal andtemporal interdependency between, say, clauses. Logical meanings are realizedby recursive structures which add one element to another so as to build up morecomplex structures.

In the systemic-functional framework, the language system is a network ofvalue-creating relationships. The meta-grammar is a description of the proportionalities, the paradigmatic systems of options, and the value-makingdistinctions in and through which the resource systems (phonology andlexicogrammar) for constructing texts are organized. Theory, then, is concernedwith how these realize, construct, and construe social reality. The grammar is aresource - a meaning potential - for making meanings in and through texts,which are produced and assembled in meaning-making activity. The grammar issemantically oriented and internally organized along multifunctionallines, as inHalliday's (l978a; 1979) metafunctional hypothesis. The relevant epistemologyis one of 'complexity': lexicogrammatical forms realize a complex interplay ofmultiple semantic factors, aspects of which may be selectively foregrounded inthe real-time process of meaning-making, so as to construe some dimension ofmaterial-phenomenal reality, which is being attended to in the process of socialinteraction. The multiple strands of meaning potential in linguistic form eooccur in a subtle and indeterminate relationship of 'complementarity', whichcan never be reduced to a single or determinate meaning on account of thedynamic and open nature of the interacting functional semantic subsystemsinvolved. Systemic-functional linguistic theory provides useful guidelines as tohow the semiotic principle of value as first developed by Saussure (1993) canbe reconnected with an account of meaning-making activity (see section 11,pp. 49-54).

By and large, systemic-functional theory has been seen as a socially basedsemantic theory of lexicogrammar and its role in the construction of discourse


(see Halliday's 1978a notion oflanguage as a form of social semiotic; see also hisdistinction between the 'intra-organism' and 'inter-organism' perspectives onthe study oflanguage in Halliday 1978: 10; see also Wells 1999: 10-11). Biologicaldiscourses have not played a prominent role in the development of systemicfunctional theory (see, however, Halliday 1995; McKellar 1987, 1990; Lemke2000a; Thibault 2000a). This may be due to the conviction - perfectly correct that social semiotic phenomena are not reducible to biological phenomena.However, more recent developments in the theory of complex dynamic opensystems show the importance of developing a theory of social semiosis in whichthe socio-cultural and the biological domains of inquiry are brought into a newdialogue with each other (see also Wilson 1998). It is now becoming possible tomake a start in the process of building the theoretical bridges between the 'intraorganism' and the 'inter-organism' perspectives on meaning-making activitysuch that there is no contradiction or dichotomy between the two.

I define my own starting point in the present study as follows. The purpose oflanguage and other semiotic modalities is to guide and co-ordinate our interactions with the non-self and to integrate us with our ecosocial environment acrossspace-time scales that go beyond the here-now scale of the biological organism'smaterial interactivity with its immediate physical environment. Given thispremise, it seems logical to say that the intrinsic organization of language hasevolved in the species (phylogenesis) and develops in the individual (ontogenesis) so that it cross-couples both with the biological architecture of the bodybrain system and with our ecosocial semiotic environment in ways that closelyrelate to the kinds of social activities that humans perform and the meaningsthey make in and through these activities. The functional and contextual basisof systemic-functional theory will prove to be an ideal conceptual and analyticaltool for developing these links.

Rather than the frozen 'syntactic islands' that are studied by formal theoriesof language, the emphasis here is on discourse as a form of activity that is richlyconstrained by both society and culture, on the one hand, and by the languagesystem qua meaning-making resource and by the emerging constraints of aparticular discourse context, on the other (Halliday 1978a; Beaugrande 1997:48). Rejecting the view that a universal grammar enables language users torepresent the world through relations of 'truthful' correspondence between thesyntactic forms of a language and an objectified external world, the importancegiven in systemic-functional theory to the interpersonal and textual dimensionsof meaning in discursive activity, in addition to the experiential dimension,draws attention to the role of discourse in mediating and enacting body-worldrelations, rather than seeing language as a symbolic-referential means forrevealing and transmitting disembodied 'inner' thoughts, intentions, andsymbol tokens. The metafunctional basis of meaning-making activity and thesemiotic resources that are eo-deployed in discourse contexts therefore suggesta number of parameters in terms of which both body-environment transactionsand body-brain functions are co-ordinated and organized without in any waysuggesting that either body-environment transactions or body-brain functionsare explanatorily or causally reducible to principles of semiotic organization orformal structure per se.

INTRODUCTION 49

However, the emphasis suggested here does require us to take the intrinsicorganization of semiotic activity seriously not as an end itself, but as a guide tothe ways in which semiosis enables body-brains contextually to integrate withtheir ecosocial environments across potentially very many different space-timescales (see chapter 5). In this perspective, body-brain organization can be conceptualized in terms of the way it affords possibilities for the contextual integration of persons into meaning-making activity, rather than in terms of putativeinternal mental states and competences that 'cause' semiotic activity. The alternative is to remain in the old mentalist paradigm according to which languageand other semiotic modalities are, above all, formal windows on internal mentalstates, rather than modes of embodied meaning-making activity-in-time whichdirectly contribute to the self-organization of neural processes at the same timeas the latter - neural processes - directly participate, on their own scalar level, insemiotic processes, rather than lying behind these as the efficient causes ofsemiotic activity. In this view, discourse, irrespective of the specific semioticmodality or combination of modalities, is a dynamic, time-bound interfacebetween socio-cultural and material processes and constraints in the world andthe body-brain processes and functions that afford the production and receptionof vocalizations, gestures, and so on.

Moreover, the dialogical basis of all forms of meaning-making, including thevery earliest infant-parent dyads, suggests that meaning-making is inseparablefrom the forms of semiosis that are specialized to the inner realm and usuallydescribed in terms of a separate language of 'thought' and 'cognition'. AsVygotsky has shown, interaction with others both precedes and guides andshapes the transition to the forms of 'inner' dialogue that we conventionally call'thought'.

11. Reconnecting the Semiotic Concept of Value to the Body-brain Systemand to Meaning-making Activity

Saussure (e.g. 1993: 357-66) theorized the language system (la langue) as asystem of differences (see Thibault 2004b). The value of a given term in thelanguage system is defined in terms of its differential relationships with the otherterms with which it contrasts in that system. The language system is a (decontextualized) system of differences, abstracted from actual uses of this systemin contexts. Saussure's theory was by and large eo-opted by structuralism. In thestructuralist view, there is no real place for time or for change. Structures areunchanging and immanent potentialities that are not linked to real-time activity,or to developmental and evolutionary processes on their timescales. They areseen to pre-exist their manifestations in texts and social practices, rather thanbeing emergent properties of self-organizing processes on various timescales.The alternative view that I put forward here seeks to connect the concept ofvalue with the theory of complex self-organizing dynamic open systems. Timeand activity are central to this account. In this account, the language system isnot a pre-existing code, but is emergent from the real-time meaning-makingactivities that individuals engage in.


On the cultural level, a given language community displays considerablediversity and variability in its patterns of meaning-making both on the individuallevel as well as on the level of particular social groups and the heteroglossicrelations of alliance, cooperation, competition, conflict, co-optation, and so on,among these. This diversity is manifested in the range of social meaning-makingpractices and their distinctive discourse voices that constitute the system of socialheteroglossia of a given community. A language system qua system of differentialvalues is an n-dimensional semantic space (Halliday 1991; 1992a: 30). Such aspace is a purely hypothetical region encompassing a very much wider range ofpossibilities than those that are typically or actually used. The system of socialheteroglossia refers to the range of typical discourse voices and their associatedevaluative viewpoints, their relations of alliance and conflict with other voices inthe same community. Typically, the system of social heteroglossia activates onlysome of the semantic regions of the n-dimensional semantic state space becauseonly some possible combinations from among the very many theoreticallypossible ways actually get made and recognized in a given community. However,a language system is always also embedded in a higher-scalar ecosocial semioticenvironment, which it in part constitutes. Changes in the ecosocial system - e.g.changes in social institutions, social practices, heteroglossic alignments ofdiscourse voices, matter-energy flows, and so on - can lead to changes in theattractor spaces of the language system of a given community.

Similar observations can be made concerning the individual level. In theinfant's semiotic development, the primordial richly interconnected network ofneurons is a very large state space which gets channelled along more specificpathways and more specific functional differentiations in the course of its development. As the infant engages in meaning-making activity with others, this richinitial vagueness will, in time, shift to functionally more specific semanticregions, depending on the type and the frequency of the various discourse experiences that the infant participates in. Each individual has a unique trajectorythrough this semiotic state space because individuals can occupy a range ofdiverse semantic state spaces as well as adopt unique pathways for getting fromone to another. By the same token, individuals also converge on similar solutionsto the kinds of problems that they are required to resolve in their discursiveinteractions with others. This convergence can be explained in terms of theoperations of constraints working on various levels: (1) the constraints of thelanguage system itself; (2) the constraints of the discourse genre or social-activitytype which is perceived to be in operation; (3) the constraints of the particularsituation or context as it emerges in real-time activity; (4) the constraints of thespecific meanings that participants are negotiating or think they are negotiating;and (5) the constraints of individual experience and one's personal historicalbiographical trajectory. Rather than saying that a given grouping of individuals'shares' an already pre-existing language system, it is more logical to say thatindividuals are constrained by factors operating on various interrelated scalarlevels of organization - e.g. phylogenetic, ontogenetic, and semiogenetic - suchthat they converge on solutions which are construed as being the same or similarand, at the same time, there is considerable diversity and variability in the waysin which they may arrive at these solutions.

INTRODUCTION 51

The differences that characterize a given language system are the semioticallysalient differences that are recognized and used by the members of a givenlanguage community. It will be clear that I do not in any way assume that thesystem of differences is identical or homogeneous for all the individuals whomake use of its resources. It is not therefore a uniformly 'shared' system. As anabstraction, it refers to what some theoretical viewpoint constructs as the regularand typical patterns of language that are recognized and used by a given set oflanguage users, however large or small this set may be. The system of differencesrefers to the possible kinds of meanings that language users can make in thevarious types of social contexts and social relationships which language is used toenact and mediate. This system of possible kinds of meanings can then beconnected to the various ways in which meanings are actually made, in whichcontexts, and by which social participants.

The system of differences that constitutes a given language system is not valuefree in a very non-trivial sense. The semiotically salient differences that arerecognized in a given language community are a system of typological-categorialdistinctions whereby the users of the resources of that system can orient to,categorize, evaluate, act in, and interpret the world in and through their participation in discourse. Rather than seeing the brain as a sort of digital computerwhich operates on and manipulates abstract symbol strings inside individualheads, the notion of value, as discussed here, shows how we can connect biologyand society in a conceptually unified way through the notion of activity. Both theprimitive, biologically intrinsic, pre-semiotic value biases in the newborn, asdiscussed by Edelman (1992) and Thelen and Smith (1994), and the full-fledgedsystems of ecosocial semiotic values characteristic of semiotic systems such aslanguage, gesture, facial expressions, depiction, and so on, afford ways oforienting to, interpreting, and acting in one's material and social environments.They do so in ways which are always dependent on activity in real-time. It is inthis way that prior, more primitive value biases afford access to more complexdiscriminations in and through the time-bound activities that these valuesmotivate infants to engage in. This is where value connects with activity.

The activities in which the infant participates enable the child's engagementwith other persons, sounds, faces, texts, tools, artefacts, symbols, pictures, and soon. These activities loop from the world to the child's intrinsic body-braindynamics in ways which shape and reorganize them. There are no symbols,words, signs, texts, and so on, in the individual's head. Rather, symbols, words,signs, and texts are dually semiotic-material artefacts that exist in the externalworld - the ecosocial environment - by virtue of somatic or extra-somaticprocesses of production and manipulation. It is their dual status as material andsemiotic artefacts which affords their integration both to the here-now scale ofbody-brain activity and to other processes, activities, happenings, and so on, onother space-time scales, beyond the here-now one of the text as material artefact.In this way, semiotic-material artefacts such as texts are extra-somatic resourcesthat enable the participants in meaning-making activity to co-ordinate andintegrate their contributions to that activity. Texts qua semiotic-material artefactsare functional components of a meaning-making (semiogenetic) trajectory thatextends from the organism's central nervous system through its bodily (sensori-


motor) activity and into its ecosocial environment and then loops back again tothe organism. Meaning-making is distributed among all the components of thistrajectory-in-time. Participants in meaning-making activity must lock into anddraw on the resources afforded by texts in order to co-ordinate their activitiesacross a diversity of space-time scales.

When Saussure points out, in his famous discussion of linguistic value, that'the purely conceptual mass of our ideas, the mass separated from the languagesystem represents a sort of unformed cloud' (1993: 362), he is showing how asystem of typological-categorial differences - i.e. the terms and their values in agiven language system - emerge from and, in turn, reorganize and give morespecified meaning to the topological-eontinuous substrate ofwhat he calls 'ideas'(see Thibault 1997a: 164-73). The principle of value therefore has the potentialto contextualize the phenomena of human experience in and through thesystem of semiotically salient differences that characterize a particular languagesystem. However, Saussure did not take the further step of showing how signscontrast with each other in ways which also depend on the discourse contexts inwhich they function in the making and negotiating of meanings in human life.Saussure's observation is not unlike those made by Edelman and Kinsbournethat the primitive state of the brain in the infant is a richly interconnectednetwork ofneurons that differentiates or 'crystallizes' into a more clearly definedset of functional distinctions through processes of activation and entrainment ofneural networks into more highly specified functional sub-regions by virtue ofthe initial value-biases that set these processes in motion in the first place.

Edelman showed that the values placed on early infant behaviour lead tosome activities being preferred rather than others. It is these biologicallyintrinsic values which, in the first instance, motivate infants to attend to theirsurroundings, to cry when some need has to be satisfied, to seek warmth, food,and affection, to orient to and discriminate faces, to lock into and to engage withadult meaning systems, and so on. Thus, babies attend to the faces, the sounds,the movements of adult caretakers, in ways that enable them to integrate theiractivities with those of more senior others. They do so, in the first instance, onthe basis of their sensori-motor capacities for interacting with and integratingwith their immediate material environment. The resulting dialogic closure(BratenI992; Thibault 2000a) of the human dyads so formed leads, in time, tothe transformation of the infant's body-brain system and the building up ofincreasingly more specified and differentiated semiotic takes on the world.

]ohnson and Morton (1991) have postulated that the initial value orientationthat infants have to faces is based on an initial discrimination between a numberof contrasting 'blobs', corresponding to the location of the mouth and eyes(chapter 4, section 12, pp. 209-12; see also Thelen and Smith 1994: 315). Thisprimitive distinction is based, initially, on topological-continuous variation,rather than on the typological-categorial distinctions of the kind that are foundin the lexicogrammar and semantics of natural languages. The discriminationsthat the newborn begins to make are motivated in the first place by a value onthe infant's self-organizing activity such that the infant's intrinsic dynamics arebiased towards preferring to orient to his or her caretaker's face. This preferential bias occurs in synergy with the dynamics of other interacting subsystems such

INTRODUCTION 53

as, for example, the need - itself a value - to lock into and obtain informationfrom one's caretakers. The initial value bias discriminates between 'orienting tofaces' versus 'not orienting to faces'. In time, this leads to the further discrimination between the 'blobs' corresponding to the eyes and the 'blob' corresponding to the mouth, and so on. These early primitive differentiations themselvesbecome the topological substrate for increasingly more complex and highlyspecified differentiations whereby the faces of others serve not only to indexindividual identity, but function as a complex semiotic system of facial expressions involving many interacting subsystems, including, for example, theposition and movement of the eyebrows, the eyes, the mouth, the lips, the nose,and so on. Each of these subsystems, along with the possible interactions amongthem, as well as the interactions with other semiotic systems such as language,means that the face is a semiotic-material artefact. It is capable of making acomplex set of semiotically salient distinctions that can be interpreted, both inrelation to other semiotic modalities and in relation to the overall situation inwhich they occur, as having particular meanings. Adults smile and pull faces atinfants in ways that treat the baby as a full participant in the interactions betweenthem. They do so partly on the basis of the infant's own repertoire of behavioursfor eliciting responses from adults. Babies are thus entrained to the idea thatfaces are an important source of information about the others' meanings,including their intentions and feelings.

The complex systems of differentiations that can be made by the differentsub-regions of the face, both separately and in relation to each other, aresemiotic values - i.e. semiotically salient distinctions - which can be used both tomodulate and to orient to activity and to one's interpersonal engagements withothers. The initial value bias to orient to faces is biologically driven and isfounded on the intrinsic biological dynamics of the newborn. In time, however,the increasingly complex range of discriminations that can be made by the facein partnership with the child's increasing attunement to the potential significance of these in discursive interaction entrains the system to a more highlyspecified system of differentiations that has social semiotic significance in aparticular community.

The processes described here for the face are also relevant to the protolinguistic sounds that infants produce in their first nine months or so. In time,the infant's vocalizations are entrained to the articulatory practices that subtendthe phonological system of a given language system. At the same time, they learnto integrate the phonological units and structures of the language with its lexicogrammatical and semantic structures in contextually relevant ways. Theinfant's vocal-tract activity is a complex topological space that is capable of discriminating, initially, a far wider range of phonetic distinctions than those thatare recognized by the system of phonological distinctions that are salient in agiven speech community. Through processes of mutual activation and entrainment over time, the child's body-brain, and more particularly its vocal-tractactivity, is shaped so as to recognize and use a more limited set such that otherpossible distinctions wane and fall into disuse.

The very primitive value biases that function to initiate and guide the infant'searliest activities along some pathways rather than others can therefore be seen


as the vaguer, less specified precursors of the more highly differentiated andmore specified systems of distinctions that are characteristic of full-fledgedsemiotic systems such as language. Following the logic of specification hierarchythinking (section 8, pp. 34-9), this does not mean that the earlier, vaguer, moregeneral stages are transcended as later, more specified ones emerge along theindividual's ontogenetic trajectory. Instead, the earlier stages are integrated toand reorganized by the emergence of the later stages. The earlier stages can besaid to prepare the emergence of the later stages insofar as the earlier stagesconstitute the substrate from which the later stages arise as newly emergent intermediate levels of organization. Thus, the presence of the more specified laterlevel implicates the presence of the less specified earlier level, but not the reverse(Salthe 1993: 64). The earliest, most primitive value-biases motivate the infant tolock into and engage with the semiotic and material resources that are affordedby both caretakers and the physical milieu such that their body-brains areentrained along particular developmental trajectories.

The semiotic principle of value thus has the potential to contextualize and toentrain both the body-brain and the world to its categories in and through thesystem of semiotically salient differences that characterize a particular languagesystem. The concept of value, when reconnected with activity in the waysuggested here, is therefore an important component in an ecosocial semiotictheory of how body-brain systems are contextually integrated with their ecosocialenvironments.

12. A Thumbnail Sketch of the Book

The book is divided into three Parts. Part I consists of the general theoreticaloverview that was presented above in the present chapter. Part II focuses on theexpression stratum - the phonetics and phonology - of spoken language inorder to explore the principle of scalar heterogeneity in semiotic systems.Chapter 2 takes as its point of departure Hjelmslev's account of the stratifiednature of language as a semiotic system. In this chapter, Hjelmslev's insightsconcerning the ways in which the expression stratum interfaces with andconstrues the 'phonetico-physiological sphere of movement' (1961 [1943]: 54)are explored. With reference to Lernke's (2000a) notion of 'scale heterogeneity'or 'scale mixing' in ecosocial semiotic systems, Hjelmslev's seminal insights aretheoretically reconstituted in order to shed light on the ways in which the articulatory processes of the vocal tract in vocalization are cross-coupled with higherscalar systems of meaning relations and practices on the ecosocial level.

Chapter 3 begins by considering the ways in which action and meaningemerge from the time-dependent cross-eoupling of a diversity of systems ondifferent space-time scales. The Principle of Alternation (Lemke 1999) isintroduced and its implications for how the continuous topological variety of thesounds produced at level L-I of the perceptual-motor activities involved in theproduction and perceptual pick-up of speech sounds is reorganized and interpreted as discrete phonological categories at level L+I, corresponding to thephonological system of a language. In chapter 3, I also explore the iconic,

INTRODUCTION 55

indexical, and symbolic dimensions of both the expression and content strata oflanguage. The cross-scalar nature of both expression and content is discussed interms of a nested specification hierarchy consisting of iconic, indexical, andsymbolic layers of organization. Edelman's (1992) account of the brain as aselective recognition system which recategorizes sensori-motor routines asconceptual ones is related to the cross-coupling of expression and content. Inturn, the above considerations are related to the ways in which individuals areintegrated to their semiotic trajectories by virtue of the ways in which diversetimescales are integrated across expression and content.

The three chapters in Part III address different aspects of how brain processesand functions relate to meaning-making on multiple timescales. In chapter 4, Ibegin by rejecting the view that consciousness can be explained in terms oflower-scalar neural processes per se. I critically discuss Terrence Deacon (1998[1997]) on the relationship between subjective experience and brain processesand begin to develop an alternative to the 'theory of mind' that is invoked byDeacon (see also Shanker 1996). I propose that self-organization entails anincrease in a system's overall semiotic complexity and specification as a result ofthe collection of the products produced by its own cascading. This proposalprepares the ground for the central argument of this chapter, viz. consciousnessis a structured system of meanings involving all levels of the specificationhierarchy of iconic, indexical, and symbolic modalities of semiosis. It is in thisway that we organize our experience around a notion of the self along atrajectory. The metafunctional character of semiosis is then related to the waysin which the hemispherical organization of the brain itself suggests ways in whichthe structure and dynamics of meaning-making are consistent with our neuroanatomical architecture without, however, being reduced to this. A furtherimportant development in this chapter concerns the entropic character ofmeaning. Using Salthe's (1993) theory of infodynamics as my starting point, Isuggest that the meaning which is stored along the individual agent's historicalbiographical trajectory increases as the trajectory develops and individuates. Theimplications of this are further discussed in terms of developmental issues.

Chapter 5 takes up the issue of the metafunctional organization of semiosis bysuggesting that lexicogrammar is embodied in dynamical attractors or constrained pathways through the multidimensional semantic space that constituteshigher-order consciousness. The question is posed as to how we can talk aboutthe temporal and spatial grounding of consciousness, its richly patterned connectivity that implicates networks of relations across many diverse space-timescales, and the sense of being part ofand related to a wider social and perceptualfield of relations, which includes others. Some further developmental aspects ofthis question are discussed in connection with the perceptual-motor phase ofcategorization in the infant's earliest sensori-motor engagements with the worldin the work of Thelen and Smith (1994) and in Halliday's (1975, 1992a, 1993)account of a child's transition from protolanguage to language. These discussions prepare the ground for an extended discussion of the metafunctionalshape of consciousness in relation to Darnasio's (1999) theory of the relationshipbetween the experiencing self and the object of consciousness. The dialogicalbasis of consciousness is also briefly discussed in relation to some connections


which I postulate between recent work on mirror neurons and interpersonalmeaning. The chapter concludes by extending these observations to innerspeech and its relationship to higher-order consciousness.

Chapter 6 begins by emphasizing the unity and continuity of the physicalbiological and the social-eultural dimensions of our existence. This initialemphasis forms the backdrop for an extended discussion of Flohr's (1991)theory of phenomenal awareness. Flohr is interested in overcoming theprevailing dualism between the physical and phenomenal dimensions of brainstates. In my view, Flohr's description of 'activity-dependent self-organization ofneural nets' shows that contextualizing relations are relevant on the diversescales of neural organization that give rise to phenomenal awareness. Theprinciple of me ta-redundancy operates here as well. I then show that thePrinciple of Alternation provides a further crucial perspective on the brain asmeaning-making organ. Thus, the transformation from discrete, local neuralfirings to global configurations of neural networks on higher scales of brainactivity and the transformation of such global configurations to objects ofconscious experience within the brain constitute a semiotic transformation, suchthat conscious experience emerges in the perspective of the self.

Part 11

2 Sensori-motor Activity, Movement, andSocial Meaning-making: Rethinkingthe Expression Plane of Semiosis

I. Some Early Signposts from Saussure and Hjelmslev: The ExpressionPlane as Embodied Articulatory Movement

Saussure's notion of the 'signifier' and Hjelmslev's concept of the 'expressionstratum' of semiosis have often been mistakenly interpreted as if they referred toa carrier, vehicle, or medium that is distinct from the level of meaning. Phonological or graphological forms, rather than contributing to the overall process ofmeaning-making, are simply seen as the means for communicating a meaningwhich is separate from them. Thus, form and meaning are dichotomized. In thismistaken view, phonological and graphological forms are not considered to playan active and constitutive role in the process of meaning-making. In someaccounts, they simply serve as the means for differentiating meaningful distinctions in the morphology and syntax of the language in question; in others, theyare no more than the means of transmission of a meaning which is separate fromthem. In both versions, form and meaning are separated. Furthermore, suchaccounts, in virtue of a strict dichotomization of phonological form andphonetic substance, fail to account for the ways in which material (e.g. bodily)and semiotic processes and their dynamics necessarily intersect in semiosis.

In my view, both Saussure and Hjelmslev developed the beginnings of a verydifferent view, one which has to do with the ways in which material, bodilyprocesses and semiotic processes cross-couple and intersect in the dynamics ofsocial meaning-making. In so doing, the two kinds of processes may be said tomutually determine the overall dynamics of the system. In this perspective, a verydifferent status is given to the expression plane of semiosis. Rather than postulating the separation of 'form' and 'meaning', each on distinct levels, there areconstant and simultaneous transactions across all levels and in all directions from sensori-motor activity to perception to actions performed in the environment of the self. We act on and explore the world, we orient to others in andthrough articulatory movements of our bodies, which bring about specificmatter-energy effects and consequences in the world, including our interactivepartners. In turn, the vocal-tract gestures that we articulate, the sounds thatresult from these, and their cross-modal couplings with visual or other events inthe environment of the participants to discourse events produce reentrantmappings onto the very bodily gestures that engendered these cross-modalperceptions-cum-actions in the first place.


Saussure's (1971 [1915]: 156--7;see also Thibault 1997a: 166--71) discussion ofthe way in which language emerges between the two 'amorphous' or topologicalcontinuous substrates of 'sound' and 'thought' can be seen as an early attempt toexplain how language and other semiotic resource systems must reduce andentrain the many degrees of freedom - the vague and indeterminate possibilities- of both the bodily processes and dynamics involved in the articulation andproduction of speech sounds and the stimulus information which the organismpicks up in its environment (Gibson 1986 [1979]: 55-7). Saussure put it like this:

The characteristic role of the language system [la langue] vis-a-vis thought isnot to create a material phonic means for the expression of ideas, but to serveas the intermediary between thought and sound, so that their union necessarily leads to the reciprocal delimitation of units. (Saussure 1971 [1915]: 156;my translation)

B

In the first case, a given language system provides the phonological (or other)resources whereby the initially vague or indeterminate potential of the muscles,joints, and so on involved in vocal-tract or other motor functions are entrainedand categorized as co-ordinated and patterned bodily movements for thepurposes of dialogic interaction. This is the task of the stratum of the signifier(Saussure) or the expression plane (Hjelmslev) of semiosis. Saussure makes itquite clear that, in his view, the signifier is not simply a material carrier of a moreabstract and extracorporeal signified.

In the second case, the language system provides the lexicogrammatical andsemantic resources in and through which individuals construe the phenomenaof experience as a more determinate, though never fixed or unchanging, systemof semiotic categories. However, these are not two separate activities. Nor arethey on two distinct levels of abstraction. Saussure shows that language formentails a 'reciprocal delimitation of units' in both 'sound' and 'thought.' Inacting as the intermediary or the interface between the two, language crosscouples and synchronizes the bodily (articulatory) dynamics of the individualwith those of other individuals as well as with their shared Umwelt of perceptualinformation, tools and other environmental affordances. I am using Cibson'sterm perceptual information rather than the more physiologically based conceptof a perceptual stimulus because the former refers to environmental affordanceswhich act as material causes relative to the environment of the individual. A

RETHINKING THE EXPRESSION PLANE OF SEMIOSIS 61

perceptual stimulus, on the other hand, is an efficient cause insofar as the application of energy will stimulate a receptor. However, a perceptual stimulus doesnot specify any information about the source of the stimulus. As a material causerelevant to the organism's behaviour, perceptual information is a relevant orsalient aspect of the environment with which the individual interacts inmeaningful ways.

In this Gibsonian framework, it is possible to re-interpret Saussure's distinction between 'sound' and 'thought' in the following terms. The semioticentraining of the body's many degrees of freedom, whereby the indeterminateflux of sound is re-construed as articulate speech, is a way of putting theindividual into contact with relevant aspects of its ecosocial environment,including his or her conspecifics. Articulate speech and other semioticallyentrained motor functions or gestures of all kinds constitute bodily activitywhereby socially shareable ways of interacting and integrating one's bodilyactivity with others are projected into the environment. At the same time, individuals pick up relevant perceptual information by means of their perceptualsystems and then semiotically elaborate or re-construe this in their Innenweltaccording to the categories of their stored langue interieure: In the first instance,this information provides indexical information about certain relevant aspects ofthe environment, as perceived by the individual. However, the emergence of alanguage system between 'sound' and 'thought' means that the indexical relationship of necessity between environmental events and their perceptual pick-upis severed. That is, their semiotic re-eonstrual in and through the resources of ashared language system is based on the fact that the individual's stored langueinterieure provides a means for both producing and interpreting such environmental events as symbolic possibilities, rather than as indexical necessities(Salthe 1993: 176). The existence of this capacity presupposes, of course, theability to recognize that other organisms also share the same system of symbolicpossibilities whereby environmental events are interpreted and acted on as signs,rather than as indices of physical events per se in the here-now relation betweenthe organism and its immediate physical environment. In this way, the manydegrees of topological freedom of the phenomena of experience - cf. Saussure's'thought' - are themselves entrained to the semiotic categories of a givenecosocial system.

Thus, the sensori-motor functions of the body must be harnessed and coordinated so as to cross-eouple with and function in relation to the extra-somaticworld of perceptual information. By the same token, the semiotic activity ofconstruing the world as meaningful 'content' must be cross-eoupled to andmatched up with the dynamics of the agent's bodily activity. If this were not thecase, then there would be no way in which body-brains could be co-ordinatedand synchronized through jointly made interaction procedures for the purposeof social meaning-making. Furthermore, this co-ordination of body-brains showsthat motor activity, perception, and meaning-making are all of a piece - they areall heterogeneous dimensions of the one overall self-organizing system in andthrough which meanings are made in ecosocial space-time. As Saussure andHjlemslev understood, a language system entails the progressive differentiationand entraining of both sensori-motor activity and its products ('sound') and the


phenomena of experience ('thought') to the categories of the expression andcontent strata of a given language system (see Thibault I998a).

It makes little sense, then, to think in terms of the phonological realization perse of a particular linguistic meaning, seen as existing on a different level ofconceptual or symbolic abstraction. Instead, there is a rich, on-line connectivitybetween bodily articulation, the acoustic information which is projected into theenvironment, and its correlations with objects, events, and so on, in othersensory modalities such as the visual, the spatial, movement, and so on. This richconnectivity suggests that a primary function of the signifier or expressionstratum of semiosis is the physical-material integration of multimodal perceptualevents into a more global meaning which constrains lower-levels. In other words,the expression stratum both (1) indexes its connections with other sensorimotor and perceptual events in other modalities and (2) affords possibilities ofits symbolic construal in virtue of its mediation by, for example, the lexicogrammar and semantics of language. In such a view, we may hypothesize that, forexample, audible vocal-tract gestures, facial movements, and manual-brachialgestures are connected by re-entry such that the visual, auditory, and movementsystems, while independently mapped in the brain (Edelman 1989: 64-90), arealso mutually connected so that higher-order or me ta-level relations amongthem emerge. These higher-order relations, however, do not simply face one way- to the body of the articulator. They are both proprioceptive and exteroceptive- they also face outwards to the world of the non-self. That is, emergent cross-modal connections among sensori-motor activities are directed to and focusedon perceptual events in the purview of the participants in some occasion of interaction. This is where lexicogrammatical mediation comes in: in mediating thecross-modal integration of sensori-motor activity and its indexical connectionswith specific events in the world, the symbolic possibilities of lexicogrammarmassively expand and elaborate in culturally specific and shareable ways themeaning-making potential of sensori-motor classification, and in ways that gobeyond the here-and-now of perceptual categorization per se (Edelman 1989:187).

Salthe points out that models of complex systems 'change the rate at whichthe processes we are modelling occur so as to bring them into line with our ownobservation rates' (1993: 48). Insofar as language and other semiotic systemsmodel and construe the world, we can say that such systems are smaller-scalemodels of the higher-scalar ecosocial system that we are embedded in. Themodels of the world afforded by language and other semiotic modalities belongto our own scale, i.e. the world in which we live and in relation to which we shareperspectives with others. Since the reference scale includes our own embodiedsensori-motor activities, this suggests, for example, that the phonologicalentraining of vocal-tract gestures as a semiotically formed expression substanceis a means whereby the many degrees of freedom or difference - the vague andindeterminate possibilities - of the bodily processes and dynamics involved inthe articulation and production of speech sounds constitute a means of bringingthe phenomenon which is modelled into line with our own embodied observation rates. Thus, factors such as intonation, rhythm, tempo, and syllabificationon the expression stratum entrain vocal-tract dynamics at the same time as they


construe the lexicogrammatical units and structures on the higher-level contentstratum. It is the latter stratum which, in turn, interfaces with the world so as toconstrue the phenomena of experience in and through the semantic categoriesprovided by the language system. The content stratum thus construes thephenomena of experience with which it interfaces as meaning. Halliday (I 992a)points out that for meaning to be made there is necessarily a second interface viz. the expression stratum - whereby meaning as construed on the contentstratum is transduced into bodily processes of articulation. In Halliday's ownwords, 'it is the transduction of the phenomenal back into the phenomenal viathese two interfaces of content and expression' (1992a: 22). The two 'outer'faces of content substance and expression substance embody what Halliday callsthe 'contradiction between the material and the conscious' (1992a: 22). Therelationship of realization between the expression and content strata thus showshow our semiotic models of the phenomena of experience are self-referentiallylinked to the observational rates and dynamics of our own bodies.

The two outer faces of this relationship both work to complete or integratetheir respective lower-level phenomena. The expression stratum thus works tocomplete the physiological activity of the vocal tract by entraining and integrating it into the meaningful categories of the phonological system of a givenlanguage. Likewise, the content stratum integrates and completes thephenomena of experience by placing them within some system of interpretancesuch as a given language system so as to give them a meaning. Generallyspeaking, the stratal organization of language is represented as a series of levels,or strata, which are re cursively related to each other. However, this way of conceptualizing stratification does not tell us how the system of relations betweenstrata is self-reflexively connected to the embodied agents who deploy the systemin discursive interaction. In other words, where is the language user situated withrespect to the (potentially) infinite regress of levels that the recursive nature ofthe relations between levels involves? Given that stratification is always relative tosome system of interpretance, an agent can be said to constitute an interruptionor break in the recursive relations between strata (see Salthe 1993: 51). That is,the agent is located between or interfaces with both the expression and contentstrata, as shown in Table 2.1. The location of the agent between these twointerfaces thus constitutes a break or interruption in the stratified hierarchy ofrecursive relations.

The agent's interfacing with the two strata of expression and content has twomain consequences. First, it provides a means whereby the agent is attuned tothe meaning-making resources of the higher-scalar ecosocial semiotic system.That is, the agent is sensitive to the contextual constraints which emanate fromthis level. Secondly, the agent has internal dynamics which buffer the systemfrom disturbances which would otherwise compromise, destroy or overwhelm it.In this sense, an agent is said to be robust (Salthe 1993: 51; Juarrero 1999:249-50). Thus, the agent maintains its structural integrity and its agentive determinativeness - i.e. its embodied presence and agentive effects - in relation to thesystem of relations in which it is embedded. This system of relations is itselfdetermined in and through the agent's interactions with other agents who sharethe same perspectives and viewpoints in some system of interpretance. Agency is


Table 2.1: The robustness of the agent and its interfacing with the expression andcontent strata of language

Content Purport Vague topological-continuous; many degrees ofdifference

SOCIAL CONTEXT OF SITUATIONContent Substance(construal ofphenomena ofexperience in 'outer'and 'inner' domains

Content Form

Field Tenor

DISCOURSE SEMANTICS

metafunctional interface

LEXICOGRAMMAR(clause to morpheme)

Mode

Interface with the Domain of the Phenomena of Experience

~SELF AS EMBODIED AGENT with its own internal dynamics

~Interface with the Domain of the Body-brain

Expression Form PHONOLOGY

tone groupfootsyllablephoneme

GRAPHOLOGY

paragraphorthographic sentencesub-sentenceorthographic wordgrapheme

Expression Substance AUDITORY ACT

auditory perceptionvocal tract articulation

VISUAL-GRAPHIC ACT

visual perceptionhand-arm-joint-eyekinaesthesis

Expression Purport Neuroanatomical, neurophysiological andneuropsychological capacities of organism as initiatingconditions (material affordances) of semiosis


progressively revealed during the historical emergence of the individual in termsof the way the agent interrupts and interfaces with both the expression andcontent strata in semiosis. Agents individuate and have effects in meaningmaking activity by virtue of the ways in which both their bodies and the ecosocialenvironment are mediated by the expression and content strata. Table 2.1models the robustness of the agent in relation to the two strata of expression andcontent.

Phonologically entrained and habituated vocal-tract gestures enact specifictrajectories in the ecosocial environment of speaker and listener. This is a resultof the history of the use of such gestures both on the scale of the individuallanguage user and that of an entire speech community. It is because of thishistory that there has been reentrant mapping of specific gestures with specificperceptions and categorizations of the world. These gestures are, then, morethan just body movements per se. As a result of the history of their use, they createexpectations which are based on an understanding of the correlations betweenthese and other events in the world. The continuous experience of using phonologically habituated vocal-tract gestures in relation to some classes of perceptionand action rather than others leads to the emergence of stable attractors. Thus,the same patterns of vocal-tract activity used by different speakers on manydifferent occasions will lead to the same kinds of consequences in the world approximately the same kind of awareness and categorization of objects, events,and so on; the same kind of interactional routines; the same kinds of integrationof cross-modal correlations into our unified understanding of the resultingwhole as a socially stabilized meaning. There are no stored internal representations of external events in the world which are then matched with the formerand interpreted. Rather, patterns of socially entrained bodily activity are stableattractors for the multimodal integration of events and actions, includinginternal ones in 'thought', as emergent, context-specific occasions of socialmeaning-making.

This requires us to take into account what Lemke (2000a: 181) has discussedas 'scale heterogeneity or scale-mixing' in dynamic systems. In other words, the verydifferent scales of what Hjelmslev revealingly calls the 'phonetico-physiologicalsphere of movement' (1961 [1943]: 54; my emphasis) in the vocal tract of thespeaker and higher-scalar ecosocial systems of meaning relations and practicesare strongly cross-coupled to each other. The cross-coupling of the two cutsacross and reveals the inadequacy of those views which artificially separate thebiological and social-cultural dimensions of meaning-making activity. Rather,meaning-making activity emerges from the interaction in time of a wide varietyof diverse elements on different space-time scales. The cross-coupling ofexpression and content is what makes this scalar heterogeneity - the continuityof time scales - possible in meaning-making activity. Importantly, Hjelmslev, likeSaussure before him, also notes the fact that the expression plane of semiosisfunctions in ways which are precisely analogous to the content plane. Typically,there has been a division of labour between those sub-disciplines - phoneticsand phonology - which concentrate on the expression plane of spoken languageand those which focus on lexicogrammatical and semantic relations on thecontent plane - semantics, grammar. Hjelmslev's approach grasps the necessary


unity of the two in an overall theory of meaning-making. However, this approachhas not been widely assimilated into mainstream thinking about language, whichcontinues to operate the divisions and consequent fragmentations of understanding mentioned above.

How do Hjelmslev's proposals help us develop a new way of thinking aboutthe role of the body in social meaning-making? Having defined content-formand its relation to content-substance, Hjelmslev, following in the tracks alreadylaid down by Saussure (see above), then proceeds to investigate the preciselyanalogous functions which expression-form (e.g. phonology) has in relation toexpression-substance (e.g. phonetics), as follows:

Just as, for example, the color zone or the morpheme zones are subdivideddifferently in different languages in that each language has its own number ofcolor words, its own number of numbers, its own number of tenses, etc., so wecan also disclose, by subtraction from a comparison oflanguages, zones in thephonetic sphere which are subdivided differently in different languages. Wecan, for example, think of a phonetico-physiological sphere of movement,which can of course be represented as spatialized in different dimensions, andwhich can be presented as an unanalyzed but analyzable continuum - forexample on the basis ofJespersen's system of 'antalphabetic' formulze. In suchan amorphous zone are arbitrarily included in different languages a differentnumber of figurze (phonemes) since the boundaries are laid down indifferent places within the continuum. An example is the continuum made bythe median profile of the roof of the mouth, from the pharynx to the lips. Infamiliar languages this zone is usually divided into three areas, a back k-area,a middle z-area, and a front ?area. If we consider only the stops, however,Eskimo and Lettish, among others, distinguish two k-areas, whose lines ofdivision do not coincide in the two languages. Eskimo places the boundarybetween a uvular and a velar area, Lettish between a velar and a velo-palatalarea. (Hjelmslev 1961 [1943]: 54-5)

It is clear on a close reading that Hjelmslev's 'unanalyzed but analyzablecontinuum' refers to the 'phonetico-physiological sphere of movement' of theorgans of speech in the vocal tract. This 'expression-purport' - cf. Saussure's'sound' - is defined in terms of 'spatialized' or topological parameters in whichdifference is perceived in quantitative or analogue terms as, for example, moreor less degrees of intensity rather than in terms of discrete, digital categories.Therefore, the epithet 'amorphous' must be taken to refer to the analogue ortopological-continuous character of this 'unanalyzed' continuum before theentraining effects of a specific system of typological-eategorical semiotic possibilities. It is the typological-eategorical distinctions of expression-form - e.g. thephonology of a given language - which add new emergent properties to thetopological parameters of the phonetico-physiological sphere of movement by'laying down', as Hjelmslev puts it, 'different places within the continuum' asphonologically salient configurations of vocal-tract movements or gestures. ForHjelmslev, the language system does not exist sui generis. Rather, it exists insofaras it is a resource for acting on and intervening in physical-material and social


reality. Ordinarily, this is taken to refer to the world which we perceive throughour senses and which is selectively transduced by the symbolic categories ofcontent-form as a content-substance. However, as Hjelmslev shows, this affirmation applies equally to both the content and the expression planes of linguisticorganization. Hjelmslev explains this last, most fundamental point as follows:

That a sign is a sign for something means that the content-form of a sign cansubsume that something as a content-substance. Just as we felt before a needto use the word purport, not simply of the content, but also of the expression,so here again, in the interest of clarity, despite the time-honoured conceptswhose shortcomings now become increasingly evident, we feel a desire toinvert the sign-orientation: actually we should be able to say with precisely thesame right that a sign is a sign for an expression-substance. The soundsequence [nu] itself, as a unique phenomenon, pronounced hic et nunc, is anentity of expression-substance which, by virtue of the sign and only by virtuethereof, is ordered to an expression-form and classified under it together withvarious other entities of expression-substance (other possible pronunciations,by other persons or on other occasions, of the same sign).

The sign is, then - paradoxical as it may seem - a sign for a contentsubstance and a sign for an expression-substance. It is in this sense that thesign can be said to be a sign for something. On the other hand, we see nojustification for calling the sign a sign merely for the content-substance, or(what nobody has thought of, to be sure) merely for the expression-substance.The sign is a two-sided entity, with a Janus-like perspective in two directions,and with effect in two respects: 'outwards' toward the expression-substanceand 'inwards' toward the content-substance. (Hjelmslev 1961 [1943]: 57-8)

If content-form functions semiotically to construe the phenomena of the world- real, imagined, hypothetical, it does not matter - as instantiations of thecategories which belong to content-form, i.e. as a content-substance, then, asHjelmslev argues, exactly the same reasoning also applies to the phoneticophysiological movements of the vocal tract. These, too, are entrained,channelled, and categorized as a language-specific expression-substanceaccording to the typological-eategorical parameters of some expression-form(phonology, graphology, ete.). In the final analysis, the language system is notdetached from the physical-material reality to which it 'refers' on both sides ofits internal organization. Rather, it is, through a long process of eo-evolution,derived from it. This is so both from the point of view of the way in whichexpression-form interfaces with the kinetic processes of the body and its technological extensions, and from the point of view of the way in which contentform interfaces with the perceived phenomena of the world we live in. Ratherthan say, for example, that the kinetic or articulatory movements of the vocaltract and the system of phonological categories are two distinct scalar levels, wecan say that there is a high degree of eo-evolved heterogeneity or intersectionbetween them. Bodily processes are mediated and entrained by the possibilitiesafforded by the system itself. In other words, bodily processes are shaped by andadapted to a given ecosocial system both systemically and instantially.


2. Energy Exchange and the Complementarity of Interacting Body-brains

The relationship between vocal-tract gestures and the phonology of a givenlanguage is not a private or individual matter which refers uniquely to the bodyof the single speaker. Instead, it is a collective and ecosocial one. The emergenceof a semiotically formed expression-substance means that the gestural potentialof the body becomes standardized and, hence, reproducible across manydifferent occasions and social agents. It also means that possible bodily gesturesmay be hypothesized and their correspondence with actual bodily states maythen be corroborated. From the evolutionary perspective, the first possibilityarises as a result of the second: the standardization of bodily movements andgestures occurred on the basis of a perspective which goes beyond theindividual's body per se. Thus, the emergence of a socially shared expressionsubstance is the semiotic-material resource which (1) integrates a multiplicity ofindividual bodies into the social body of shared bodily dispositions, movements,etc. by construing and entraining these articulatory movements in standardizedways; and (2) affords the possibility of dialogic interaction among them. Thesetwo possibilities arise from the fact that semiotically formed expression-substanceemerges from and is tied to the body of the user at the same time as it is systemically harnessed and channelled so as to co-ordinate intentionally directeddialogic interaction with others.

However, this intentionality of the 'individual' body does not in itselfconstitute the meaning-making act. The intentionality of the meaning-makingbody expresses the selective focusing of the system's resources on the specific actof meaning-making; yet, the systemic possibilities for meaning-making exist independently of the specific act. That is, at high-scalar levels of ecosocial organization at the same time that they speak through the individual act and give formto it. This 'giving form' to bodily kinetics as semiotically formed expressionsubstance is always intersubjective rather than subjective in character. It is alwaysformed with a view to (1) mutually synchronizing and co-ordinating bodies forthe purposes of dialogic interaction; and (2) selectively fixing the attention ofthese body-brains on some aspect of the world which is semiotically construedand exchanged in the act of meaning-making. It follows that gestural activityalways 'faces two ways': (l) to the body of the articulator insofar as it is the articulation of subjective, embodied feeling states, and so on; and (2) to the body ofthe interlocutor - the other - insofar as the gesturer's own body is constituted asan 'object' which affords possible dialogic interaction with and for the other.

This 'facing two ways' of human gestural activity rests on a very generalrelation of complementarity between the parties to the interaction. That is, thereciprocal and dialogic orientation of the two interactants is based, in the firstinstance, on the fact that bodily movements release energy into the environmentof the organism. This entails a change in both the gesturer and a corresponding,though by no means identical, change in the interlocutor. In this way, the participants to some meaning-making act impact upon and bring about change ineach other. The release of energy by one or the other of participants to the interaction is mediated by the constraining effects of a system of interpretance whichconnects the participants both to each other as well as to the world which they


share, i.e. to relevant aspects of their environment - their Umwelten - and theirselective internalizations of this as their respective Innenwelten. In such a view, theindividual's lang;ue interieure, as Saussure called it, is rightfully restored as constituting possibilities for action and interaction with others relative to a sharedecosocial environment. Thus, signs are assembled in the Innenwelt from theresources of one's stored langue interieure and then projected into the environment as possibilities for social interaction with others. The reciprocal anddialogic orientation of the participants is, then, subordinated to the forming ofan emergent, higher-order metastable complex such as, at the very least, aninteractive dyad. This occurs on the basis of standardized and reproduciblepatterns of bodily movement whereby the reciprocal orientation of the participants emerges and is distinguishable from the analog flux of Saussure's presemiotic 'amorphous nebula' (1971 [1915]: 155-6). That is, semioticallyentrained body movement transforms the coming together of body-brains ininteraction into a process of the selective and adaptive modification of the onein relation to the other. In the midst of the indeterminate - indifferent? - fluxof things in general, the body-brains reciprocally engaged in meaning-makingselectively orient to each other for the purposes of enacting their jointly mademeaning-making activity. Bodily movement is, then, at the basis of the matterenergy exchanges which make semiosis possible. Insofar as such movementconforms to a semiotically formed 'expression-substance', as Hjelmslev wouldput it, this means that the semiotically formed and entrained movements ofparticipant A are perceived as being complementary to those of participant B.The newly emergent semiotic properties consequently alter and entrain theprior topological dynamics of the interacting biological organisms along sociallyorganized lines.

3. The Stratified Model of Semiosis: The Problem of ConceptualAbstractness and Scalar Homogeneity

Typically, the stratum of expression-form is considered to be 'below' that ofcontent-form. In other words, lexicogrammar (morphosyntax) and semanticsare assumed to be strata which are 'above' that of the 'lower' level of expressionform (phonology, graphology, etc.), as shown in Table 2.2.

According to proponents of the stratified view of semiosis, each stratum hasminimal units which are specific to that stratum. These combine to form larger

Table 2.2: Stratified model of lang;uage, showing vertical hierarchy of different levels ofabstraction

SemanticsCONTENT FORM

Lexicogrammar

Phonology; graphology EXPRESSION FORM


structures which are specific to that stratum. The minimal unit of phonology isthe phoneme; that of lexicogrammar the morpheme; and the sememe in thecase of the semantic stratum. The relationship between strata is seen as a two-wayrealizatory one. This means, for example, that the sequence of phonemes /k<et/realizes the lexicogrammatical word 'cat', which in turn realizes the semanticunit [THING; ANIMATE; FELINE]. This encoding perspective may be reversedin the sense that the semantic unit [THING; ANIMATE; FELINE] is realized bythe lexicogrammatical unit 'cat', which is, in its turn, realized by the sequence ofphonemes /k<et/. That is, the relationship of realization is a two-way and symmetrical one at all levels. The notion of stratification designates a system ofnested levels (strata) that continue in both directions from some focal level. Theunits and the relations these enter into at any given level are parts which arenested in larger wholes which are usually described in terms of a hierarchy ofnested constituent units. Further, the relations across levels, or strata, are seen asindirect in the sense that the relations on anyone level do not directly affectthose on other levels. For example, there is not for the most part a direct relationship between phonology and lexicogrammar. In this way, the integrity andthe scalar homogeneity of the units and their relations on any given scalar levelare preserved in the prevailing structuralist reading of the ways in which thevarious strata are related to each other (SaIthe 1993: 45; Lemke 2000a). In myview, the structuralist attribution of scalar homogeneity to what are, in actualfact, radically scalar heterogeneous relations among the diverse levels ofrelations involved can be explained as follows. Typically, the stratified characterof linguistic semiosis is treated as a formal system in its own right, rather thanbeing seen as the means whereby body-brain and ecosocial system are crosscoupled in ways which enable the continuity and the seamlessness of the diversetimescales that are implicated in the concept of scalar heterogeneity.

In such a view, the tendency is to describe the 'lower' stratum of theexpression plane as a medium for the 'expression' or the 'encoding' of ameaning or content on the higher strata of lexicogrammar and semantics. Inpart, this arises from the tendency to describe the relations among the variousstrata as a series of vertically organized levels ranging from 'higher' to 'lower' orfrom more abstract to less abstract. For example, systemic-functional linguistictheory distinguishes three symbolic orders of abstraction - viz. in descendingorder of abstraction, semantics, lexicogrammar, and phonology/graphology - inthis sense with respect to language (see for example Davidse 1992: 108 forfurther discussion of the systemic-functional perspective; see also Whitson 1997,esp. 125-8 for a critique of'conceptual abstractness'). Thus, semantics is realizedby lexicogrammar and lexicogrammar is realized by phonology/graphology. Inthis view, phonology and graphology are seen as a resource for realizing moreabstract lexicogrammatical forms as spoken sounds or written notation. Theproblem with this view is that the more abstract lexicogrammar and semanticsare still seen as transcendent with respect to the embodied 'lower'-level activitiesof vocal-tract or other articulatory movement. Yet this view does not, properlyspeaking, show how the expression plane of phonology or graphology also lookstwo ways. It does not tell us how phonology, say, specifies and construes whatHjelmslev had called the 'phonetico-physiological movements' of the speaker's


vocal-tract gestures. What, then, is the relationship of phonology to bodilydynamics? I shall now discuss this question in the following paragraph. Aswe shall see, the alternative allows us to see that meaning-making emergesdirectly from embodied activities. It is not a second-order happenstance orepiphenomenon.

If, on the other hand, the expression plane is related directly to bodilymovement and activity, it is possible to conceive of this relationship in a quitedifferent way. In this alternative view, A's vocal-tract or other gestural activity issensori-motor activity which is meaningful for B in the sense that it brings abouta complementary change in B. That is, A's body movements constitute a specificmaterial action which intervenes in and exchanges energy with its environment,of which A and B are constituent parts (chapter 1, section 8, pp. 34-9). Further,this bodily action and its corresponding exchange of energy are in some senseorganized and oriented by A's relationship to B (i.e., the non-self, the environment, the other). This means that patterns of vocal-tract and other gesturalactivity qua expression plane depend on their context-specific, moment-tomoment mappings to the external world. It is in the first instance this exchangeof energy which establishes the possibility of a reciprocal, dialogical contactbetween A and B. Moreover, this energy exchange is itself socially organized anddirected in ways which allow participants to access and to orient to specificsocially defined situations.

In this view, form is not something which is separate from matter. Instead,semiotic forms supervene in and organize physical-material, including bodily,states and dynamics for the purposes of social meaning-making. When we saythat the sensori-motor movements and actions of A and B are organized inrelation to each other, this means that these bodily movements are structuredand formed in relation to each other, though always in virtue of some higherorder system ofinterpretance (cf. Peircean thirdness), C. It is C which providesA and B with a shared system of resources for jointly contextualized acts ofmeaning-making. Thus, form is not something which carries a particularcontent. I would say, instead, that the recognition of form arises as a consequence of specific matter-energy exchanges between A and B. This means thatform is itself a material state as it is interpreted by another material state inrelation to some higher-order principle, C, which links A and B to each other.Thus, meaning-making is the process whereby A and B, through their eodeployment of material and semiotic resources, texts, environmental affordances, and so on, act on and bring about socially relevant change both in eachother and in their Umwelt.

Vocal-tract and other forms of gestural activity are a means of acting on andexploring the Umwelt, including the other persons with whom one interacts.These embodied and socially entrained patterns of movement bring aboutchanges in the world of the non-self (the other) that we can perceive. It is ourperception of the consequences of these material transactions with the Umweltand the reentrant mapping (Edelman 1989) of these perceptions onto thematerial activity itself which allow semiotic values to be assigned to classes ofpatterned bodily movement. It is important to emphasize here that such changemay refer to the reorganization of neural maps in the brain, the selective


attending to some object, event, etc. in the material purview of A and B, makingsome perceptual experience or categorization semiotically salient, a change ofembodied feeling state, sensori-motor activity, and so on. This does not meansthat A and B change in identical ways. The point is that meaning-making is notlocalizeable in the textual records and products of such exchanges, but must beexplained in terms of the dynamics ofwhat A and B do to themselves and to eachother in and through such externalized resources. Texts perseare not the sameas meaning-making. I would say, instead, that they are the socially organizedproducts and records of our meaning-making activity (Lemke 1984a: 78-80;Thibault 1991a: chapters 2-4, 1994).

4. Blackboxing the Sensori-motor Dimension: Language Seen as Modalityindependent Centralized Processing Mechanism

Many linguists' conceptions of language rarely make contact with the neurophysiological, sensori-motor and actional dimension of language as eoarticulated bodily movement or gesture in ecosocial space-time. Many accountsof language in linguists, philosophy, and psychology are based on the notion thatlanguage primarily functions to alter the thinking or the consciousness perseofthe other. There is correspondingly less attention given to the ways in whichlanguage functions dialogically to organize and co-ordinate the sensori-motorand feeling states of the bodies of those who engage in social semiosis. Typically,the sensori-motor dimension has been seen in a somewhat restricted way as theprovince of the usually highly technical subdisciplines of phonetics andphonology.

Over the past decade, there have been interesting developments which haveseen a new convergence of the insights of the two subdisciplines. Nevertheless,linguists have been slow to rethink the question as to the waysin which meaningsmap onto the sensori-motor and actional dimensions of bodily movement orarticulation (see, however, Armstrong et al. 1995). In part, this is due to thetendency to view language in mechanistic and formal terms as a bottom-upassemblage of parts into larger wholes. That is, language is viewed as a kind ofmachine in which the parts are treated as regularities which have to be explainedin relation to deterministic rules, seen as efficient and/or formal causes. Thissort of reductionist analysis leads to black boxes in which language 'behaviour',motor activity, and so on, are explained in terms of central programs in whicheither (1) all the spatio-temporal details of articulation are already contained inthe program or (2) there is a referent signal or goal-state in terms of which thecurrent behaviour of the system is compared and appropriate instructions areissued. The first of these corresponds to the AI model of computer programming; the second to the cybernetic model. The search for black boxes leads tothe postulation of internal mechanisms per se for the generation of linguisticbehaviour or for the vocal-tract activity which produces articulate speech sounds.In other words, the sole burden of this causal-explanatory framework is placedon lower-scalar mechanisms which are seen as internal to the mind or the brainof the individual organism. There is little or no accounting for the higher-scalar


levels of, for example, the ecosocial system in which such behaviour takes placeand in relation to which it has the meanings it does.

This view may also lead to the conclusion that the 'formal' structure of alanguage is not required to change 'when "translated" from one modality toanother', as Marshall (1980: 277) puts it. In Marshall's view, syntax is formal andautonomous in accordance with the Chomskyan tradition. Thus, Marshall arguesthat the '(peripheral) medium imposes no constraints upon the (central)syntactic structure of messages' (1980: 276). The latter is autonomous withrespect to the former and is in no way shaped by it. In other words, sound andvisual images are the physical medium of language, but that understandinglanguage has little or nothing to do with understanding the physical characteristics of the peripheral activities which realized it. Marshall's claims are premisedon the view that there is a language faculty which is constrained by universalprinciples and parameters which are independent of sensori-motor modality ofexecution and reception. Even allowing for the observed fact that language andgesture are co-ordinated when people interact with each other, Marshallconcludes that this is due to 'some central mechanism that keeps mouth andbody working together to the same communicative ends' (1980: 284). Marshall'sargument is a species of downwards reductionism to lower-scalar causalmechanisms which are blackboxed and located in the mind or brain of theindividual. One problem with such a view is that it fails to explain that modalityspecific interactional dynamics are shaped by the activity as it unfolds inecosocial space-time and that particular semiotic modalities have eo-evolvedthrough their joint use in the making and interpreting of particular socialoccasions of meaning-making by embodied social agents.

In effect, Marshall's argument is representative of that species of viewaccording to which that which needs to be explained in socio-cultural activity islocated at the level of the individual biological organism. Thus, meaning-makingactivity is a mere secondary happenstance - cf. performance - which does notsignificantly impinge on the structure and functioning of semiotic forms. Inother words, there are no significant constraints from higher-scalar levels thanthat of the individual body-brain complex. There is nothing which might suggesthow ecosocial factors have contributed to the emergence of meaning-makingactivity and its structure and function as an intermediate scalar level which linksthe individual body-brain to the ecosocial environment of which the individualis a constituent and functioning part. The notion of a central processing andgoverning mechanism for the co-ordination of 'talk' and 'gesture' assumes aCartesian-Euclidean-Newtonian model of the world. That is, a unique andabsolute space-time in an objective world which is represented in the brainindependently of human action and movement.

Yet, sensori-motor processes are oriented to action in task-specific ways. Thecombining of various sensori-motor modalities of action is not controlled by asingle principle of the Cartesian-Euclidean-Newtonian type. Instead, diversesensori-motor modalities may be variably cross-coupled in the production of amovement which is directed towards a specific goal. The various sensori-motorsubsystems and their neural substrates are local with respect to their integrationinto some more global corporeal schema. Rather than universal and absolute


frames of reference which are independent of bodily experience, this suggeststhat these subsystems may be differentially integrated according to such factorsas the dynamics of the specific occasion of meaning-making and the semioticmodalities used, the memories of previous interactions which are accumulatedin the individual body in the unfolding of a specific historical-biographicaltrajectory, and socio-cultural habitus and dispositions resulting from class,ethnic, gender, generational, and other factors.

In my view, the notion of a central controlling mechanism simply serves as analibi for not explaining the real complexity of the body-brain complex wheneverwe engage in social meaning-making with others. Rather than the static andreductive model of a central controlling mechanism, we shall need to explore thecomplex, adaptive, and open nature of the relations between sensori-motormodalities, psychological intentions, and social meanings. We shall need to focuson the dynamic and changing relations between the body-brain and its ecosocialenvironments in the processes of meaning-making. As a first step, let us reflect onthe distinction that Marshall makes between 'talk' and 'gesture', as discussedabove. This distinction results from the attempt to understand higher cognitivefunctions by privileging an autonomous language faculty as the seat of humanreasoning. It has its origins in the Cartesian view of language as a uniquely humanfaculty of reasoning which separates us from the other animals (see Shanker 1998for an impressive critique). In recent decades, this has lead to a view of the brainas a central processor and governing principle. The brain so conceived is basedon the computation of symbol strings in which cognition is seen both as transcending its neuroanatomical substrate and as being the sole property of theindividual. The resulting separation of mind and brain and of individual fromecosocial environment serves to recapitulate the Cartesian dualistic separation ofmind and body: I think therefore I float free of my material body!

What Marshall's distinction between 'talk' and 'gesture' fails to heed is thatboth are sensori-motor activities of the body. Both are based on the commandand control of bodily movements and the inertia and other forces which areopposed to these in the media (e.g. air, water) and surfaces (the ground)through and/or over which we move (Berthoz 1997: 9). The degrees of freedomof the two movements are reciprocally constrained together over the entire bodybrain complex acting in its ecosocial context. In other words, the body-braincomplex encounters 'friction' or resistance in the world with which it interacts(see Volosinov 1983: 116; Salthe 1993: 97-9; Thibault 1995: 81-3 for furtherdiscussion of the notion of semiotic friction). This friction or resistance, in orderto be overcome, or otherwise productively harnessed, requires that work physical and/or semiotic - be done. Thus, talk and gesture are not independentfaculties or bodily actions which are co-ordinated by some central mechanism. Itis not the case that the person talks and gestures. Instead, he or she engages ina single overall activity of the body-brain-in-its-ecosocial-environment. Thismeans that the body-brain complex reconstructs and anticipates this brain-bodyenvironment-movement relation as a single overall action through theharnessing and co-ordination of many kinaesthetic receptors (Berthoz 1997: 11).These derive simultaneously from the sensory receptors associated with vocaltract activity and hand-arm-joint-body kinaesthesis.


The coherence I have just referred to derives from the embodied sense ofcongruence between the various sources of sensori-motor information and theaction schema which are stored in memory, and which enable the body-brain topredict the consequences of its actions in relation to the environment. Moreover,this sense of congruence is not simply a moment-to-moment feeling state, but isconstituted by the history of the body's transactions with its environment bothon the scale of the historical-biographical time of the individual and the yetmore extended time-space scale of the accumulated history of such interactionsin the life of a given social body. The many degrees of freedom of movement ofthe body's sensori-motor systems are regulated and constrained so as to producerelatively few degrees of freedom in the form of co-ordinated actions orsequences of actions. Such actions are always in response to friction or resistance- the secondness - that are incurred whenever the organism is required to focusand harness its available energy in response to some environmental event whichoffers resistance to the organism's completion of some goal or purpose (Salthe1993: 99). Thus, gestures harness available energy in some situation by means ofthe regulation of degrees of freedom of movement of the skeletomuscularsystems in the service of some intention, need or purpose. All movement is, then,oriented to some generalized principle of otherness - cf. the non-self - which isthe purposeful outcome of the body work - semiotic and material - that iscarried out.

5. The Intentional Character of 'Inner' and 'Outer' Sensori-motor Activity:Towards a Unified Account

The fact that internalized motor images of movement may be activated irrespective of whether the movement is physically executed or not shows that the braincan simulate and, hence, predict a movement without actually performing it(Berthoz 1997: 36). It is the motor dimension of linguistic activity which acts onand produces the peripheral (e.g. auditory and articulatory) sensations. Further,this motor dimension is intentionally directed or focused. This means thatlanguage activity is not primarily or only a matter of the transduction ofperipheral perceptual and articulatory events by the brain. Rather, internalizedmotor routines act on and influence auditory and perceptual sensations. I thinkthis is what Langacker means when he argues that both the auditory and articulatory facets of speech sounds are 'conceptual':

Even the articulatory facets of speech sounds are properly regarded asconceptual, in the broad sense in which I understand this term. Consider thesegment [i]. From the perceptual standpoint, speakers can deal with thissound in either of two ways: they can actually hear the sound as a stimulusdriven perceptual event, or they can simply imagine hearing it, i.e. they canactivate an auditory image of it (as in silent verbal thought). Moreover, theauditory image is plausibly taken as primary, in the sense that it is used tocategorize acoustic input as an instance of this particular sound. Exactlyanalogous observations can be made about the articulatory representation of


[i]. A speaker can actually implement the articulatory routine and producethe sound, or he can simply imagine implementing it, i.e. he can mentally runthrough the motor routine without this mental activity being translated intomuscular gestures. Once again the cognitive representation is primary, in thesense that it directs the motor sequence but can also occur autonomously.(LangackerI987: 78-9)

This position reverses the view according to which our perception of sensorystimuli is primary for the organization of action. Instead, it recognizes the verycentral role of internalized action schemata in the organization of perception.From this point of view, internalized schemata for movement and language - cf.Saussure's langue interieure - are but two specific cases of the notion of actionschema. Auditory, articulatory and other sensory images are neural events whichmay occur without the sensory stimulation of a receptor organ. When this is so,they are no less instantial to their schemata than are those which are evoked bythe peripheral stimulation of receptor organs. The same basic principle alsoapplies to articulation.

We can for this reason speak, as does Langacker, of mentally 'directing andexecuting the act of throwing a ball or articulating a certain speech sound'(1987: 112) without the execution of a motor response such that the internalizedmotor image of these events is not translated into muscular (gestural) activity.Even in the case of inner speech, there is internally simulated cross-modallinking of the movement of the vocal organs with their linguistic meaning as wellas between the external actions of the agent and objects and events, and so onthat are perceived by visual, haptic, auditory or other means in the stimulusarray. In my view, these observations show that gestural activity of all kinds cannotbe reductively considered as mere responses to external stimulations. Instead,they are internalized action schemata which are a function of the goals and plansof the agent who actively engages with his/her environment. Rather thanresponding to environmental stimuli, (internalized) gestural (movement)schemata - cf. auditory, motor, and other sensory images - are a means ofactively constituting and orienting to the ecosocial space-time of human interaction. This space-time is actively constituted by the gestural and other bodilymovements of social agents as a function of their own projects.

The above observations call into question the distinction between the sensoryand the motor: an internalized sensory - e.g. auditory or motor - image may beused to elicit either a peripheral response or internal neural activity which is notperipherally connected. The same neural structures are used for 'imagined'gestures as well as for peripherally executed ones. In this way, sensory imagesshow how the brain has a fundamentally predictive function. Sensory imagesshow how the brain takes up and enacts orientations in the world even beforethe body carries out the designated movement. Mental imagery is implicated inthe possibility of not executing a certain gesture. In this way, it provides schematafor orienting the body in the world and for the internal simulation ofmovements, bodily orientations as a means of modulating bodily movements inresponse to social meanings. However, the notion of sensory imagery does notpresuppose a dichotomy between 'autonomous' and 'peripheral' or between


'imaginary' and 'executed' movement. The one always implicates the other;execution implies internal simulation (Berthoz 1997: 232). To perceive one'sbody in the world is to imagine the actions which implicate its use.

6. The Symbolic Possibilities of Bodily Movement

Salthe (1993: 97) discusses how the secondness of friction, from the point ofviewof the production of a specific phenomenon - thought, in his example - may bethe energy required 'just to keep the neurons alive, because that is onlyindirectly relevant (because at a lower-scalar level) to the product of interest(here, ideas).' Friction requires the productive focusing of energy to obtain theright relations between means and end. By the same token, the energyconsumed is greater than that which is productively focused on for theattainment of the desired goal. The regulation of degrees of movement inthe production of vocal and other gestural activity means that only some of theenergy consumed by the brain-body is construed as semiotically salient activityrelative to an observer. Such an observer, from our social semiotic perspective, isphysically, biologically and semiotically constrained. It is in this way that we candistinguish the meaningful from the meaningless in the body's articulation of aparticular (semiotic) project. Thus, some movements (e.g. vocal-tract gestures)are more strongly cross-coupled to the symbolic possibilities of lexicogrammarwhereas others are more weakly cross-eoupled. Relative to an observer perspective, the expression plane construes, entrains, and motivates the neurophysiologyof body movement as symbolic possibilities for social meaning-making without,however, transcending the materiality of the body.

It is important to insist on the symbolic possibilities of bodily movements.Symbolic meaning is not inherent in the movement or gesture per se. Rather,social agents harness the meaning-making possibilities of the body in the serviceof their own projects. Yet, the body is not a passive medium through whichmeanings are transmitted to others. The body does not only transmit a force orenergy to another body, though it also does this (section 7, pp. 78--81). Thiswould reduce meaning-making to indexical necessity. Symbolic possibility meansthat others take up and transform the articulatory labour of my body and reworkit according to their own projects. In the process, the symbolic possibilities ofbodies are continually modified by others who take up and adapt these possibilities for their own purposes. Somatic resources for meaning-making, as distinctfrom extrasomatic resources, maximally foreground difference, negotiation, andinstability (Iedema 1997: 11, 108ff., 222). They are the raw materials out of whichsocial agents fashion the techniques whereby bodies are harnessed andentrained for the purpose of creating and perhaps sustaining associations withother bodies. In this way, as Latour has shown, the social body is a contingent andconstantly remade result of the harnessing of such somatic resources for theforming of associations with others (1986: 277). Such associations may be asfleeting as a single dyadic exchange between two individuals or as seeminglyenduring as the millions of associations which sustain the social body itself onmuch larger space-time scales.


The semiotic phenomenon of language is not reducible to some privilegedbio-physical or mental parameter of explanation. This view fails to explain thathuman language is an ecosocial semiotic phenomenon which transcends anymodality-specific physical manifestation such as speech sounds, written notation,Braille, or manual-brachial signing. By the same token, I pointed out in section 4above, pp. 72-5, this does not mean that these specific modalities reduce to aquestion of the physical medium of language without also contributing to thearchitecture and functioning of language. At the same time, language is notreducible to physical modality per se because any such reduction would fail toaccount for the higher levels of linguistic organization, viz. its lexicogrammar, itssemantics, and its discourse-level principles of organization. As Merleau-Pontypoints out, without the (human) intention of 'orienting oneself in relation to thepossible, to the mediate, and not in relation to a limited milieu; ... ' (1983 [1942]:176), there would be no phenomenon oflanguage. There can be no reduction tothe bio-physical principles of vocal-tract or other articulation without losing sightof the factors addressed by Merleau-Ponty. Furthermore, the fact of the diverse,modality-specific physical manifestations of language means that language can bereduced to no single physical principle of organization. Nor, on the other hand,can language be reduced to some disembodied, modality-independent centralcontrolling mechanism. This latter view assumes that language is somehow aboveembodied human experience and action in the world.

Instead, I posit that linguistic and other meaning-making activity cannot beseen as unified either by some mono-modal principle - e.g. spoken vs. written,visual vs. verbal, etc. - or by domain-specific controlling mechanisms. Rather, itemerges from the interaction of heterogeneous elements in the real-time ofhuman activity. That is, there is no single, homogeneous or unified thing calledlinguistic, visual or any other form of semiosis. As a form of human intentionalactivity, language makes both the reality we live in and itself in and through thedynamic and interacting interplay of many factors and resources - perceptual,actional, material, semiotic. Without this form of intentional action, there wouldbe no language. In this perspective, language functions as a higher-orderconstraint or boundary condition which allows symmetry-breaking relative to anobserver perspective to occur. The mediating and categorical functions oflanguage discussed by Merleau-Ponty integrate the environment into theInnenwelt of the organism.

7. Articulatory Movement Seen as Actional Semiotic, Not PhysicalBehaviour

What if both language and movement may be viewed as having both an articulatory and a semantic stratum of organization, i.e. an expression and a contentstratum? Typically, language and movement are not seen as being commensuratein this way. Yet movement, like language, may also be treated as involving arealizatory relationship between the semantic and bio-kinematic strata or levelsof organization. Until recently, the theory of movement has remained bereft ofa grammatical and semantic description. Recently, Radan Martinec (1998,2000)


has re-explored this problem using the analytical tools and insights of systemicfunctional linguistics. This lifts movement out of the realm of mere physicalbehaviour and shows that the formal organization of movement is, grammatically and semantically, probably just as rich and complex as that found inlanguage.

For example, Martinec shows that movement has a rich experiential semanticsand grammar describable in terms of configurations of a semantic Process, theParticipant(s) in the process, and the attendant Circumstance(s). In this way,movement is an actional semiotic rather than physical behaviour per se. Like thetransitivity system of language, movement is revealed to be describable as asystem of semiotically salient differences in which categories of movement aredistinguished as the various act-types, event-types, state-types which operate invarious contexts of human action. In other words, movement, like language,operates in specific contexts for the fulfilment of specific individual and socialpurposes and intentions. Human movement is dually a semiotic-discursive andphysical-material actional resource for organizing and carrying out ourresponses to specific problems and situations in the human Umwelt. Further,movement as actional semiotic is functionally organized in ways which relate itto its environments. The cross-coupling of movement and environment mutuallyconstrain each other.

Seen 'from below', the spatio-temporal articulation of motor activities on theexpression plane is not controlled by a single central governing or causalprinciple. Articulation is not homogeneous in this sense. Instead, articulation isheterogeneous: the bio-kinematic activities which realize a given action aredistributed over many interacting subsystems. This is true of both movementas-action and movement-as-vocalization or vocal-tract activity in speaking. Theprinciple of distributed (not centralized) control means that many differentdegrees of freedom in Bernstein's (1967) sense may be harnessed and eodeployed to achieve a common, overall purpose. By the same token, the samedegree of freedom may be used on different occasions and in different ways toachieve different, local purposes or intentions. There is, in other words, asynergistic entraining of many different subsystems, seen as independentvariables. No single subsystem (e.g. muscle linkage) on a lower-scalar level suchas the articulatory or expression level of movement can bring about change onsome higher, semantic level of organization. If movement, like language,involves higher-order boundary conditions or constraints of a social semioticnature, this means that the lower-level articulatory stratum and the subsystemsthat synergistically operate on this level are amplified in ways that are functionalaccording to the semantics of social interaction rather than to physical orbiological principles per se.

Different principles of muscle linkage comprise an equivalence class notbecause they all have the same degree of freedom, but because they realize thesame actional semantics. This means that the physical movement potential ofthe body is constrained by factors which are not uniquely physical andbiological (Berthoz 1997: 231-2). To be sure, the actional semiotic of humanmovement emerges from biological systems just as these in turn emergefrom physical ones. Yet, there is also what Salthe (1993: 213) refers to as the


supervenience of, say, the specifically social semiotic level of human action andintention. There would, in other words, be no significant effects on biomechanical processes deriving from the social semiotic level. This may beexplained with reference to a discussion in Armstrong et al. (1995: 223)concerning the origin and evolution of language. These authors divide theoriesof the origin and evolution of language into two broad categories. In the firstcategory, language is seen as 'evolving slowly out of precursors in the communication systems common to the primates' (Armstrong et at. 1995: 223). Theview that language derived from prior, gestural forms of communication in theprimates belongs in this category. In the second category, a radical break ispostulated between language and other communication systems. This is theneo-Cartesian view of, for example, Chomsky (e.g. 1965, 1976) and followers.Armstrong et al. discuss this phenomenon in terms of historical emergence.This is consistent with the Darwinian perspective that these researchers assume.Alternatively, I shall borrow and adapt for my purposes Salthe's (1993: 214-15)notion of developmental emergence as a means of bridging the gap betweenhigher-order social semiotic boundary conditions and the bio-physical processesthese constrain and entrain.

The emergence of language and other social semiotic resource systemsrepresents, from the evolutionary point of view, a new level of organizationwhich is intermediate between the body-brain complex of the biologicalindividual and the precursor ecosocial system that this is a part of. It is theexpression plane which interfaces between the sensori-motor potential of thebody-brain complex and the higher-scalar level of the ecosystem. It is possible toconceive of a proto-semiotic evolutionary phase when the relations between ourancestral precursors were mediated by no more than proximate bodily activities.With the emergence of the intermediate level of social semiosis, the body-braincomplex is increasingly regulated by its integration into the higher-scalar level ofthe social semiotic system. Its capacity to engage with distal environmental eventson diverse space-time scales beyond the here-now scale is correspondinglyenhanced. The newly emergent expression plane of semiosis integrates withsensori-motor processes of the body-brain complex. In the process, it bothchannels and directs these in ways which go beyond the requirements of neurophysiological functions per se at the same time as these are connected to astill-higher-scalar level - the content plane - whereby the lower level ofbio-kinematics of the body as construed and directed by the expression planecan be interpreted in relation to a still wider discourse of social meanings andpractices. This shows that there was no prior existing code or system of signswhich constituted the intermediate level of organization between body-brain andenvironment. The system is itself a historically evolving artefact which is, in time,constructed as signs are made and standardized in and through the microencounters between embodied individuals in specific niches of their potentiallyexpanding Umwelt. The latter is the accumulated result of that which is transformed from pre-semiotic indifference into semiotically salient and ecosociallyrelevant difference. The emergence of a system of signs is the history of thestabilization of the many signifying acts which in time come to be mediatedthrough the emergent thirdness of the system.


It does not matter whether sensori-matter activity is simulated or imagined, asin inner speech, or peripherally executed as in audible speech sounds. The sameneural structures underpin both forms of sensori-motor activity. Inner speech isstill internally simulated motor and hence gestural activity even though thisactivity is not peripherally executed. Thus, the brain may 'imagine' or simulateinner speech precisely because the motor activity may be inhibited at differentlevels without, however, suppressing the closed loops in which such activity issimulated (Berthoz 1997: 230). The point is that inner speech is still gesturalactivity which the 'speaker' performs even though the motor systems responsiblefor the peripheral execution of the skeletomuscular systems of vocal-tract-lipmouth-face-skin kinaesthesis are inhibited.

8. Inner and Outer Body States and Social Semiosis

The issues discussed in the previous section raise questions about the 'etc.'function, the unspoken, the background, and so on, which have to do with allthose material and bodily processes that defy categorical specification in asemiotic system of typological-categorial differences per se. As I indicated inchapter 1 (section 5, pp. 23-6), any such limitation of our inquiry to thetypological-categorial is a restriction on the notion of semiosis which needs to belifted so that we can take on board a much wider set of phenomenal states andprocesses and their cross-couplings with the semiotic. How does one talk aboutthe 'inner' visual experience of reading a poem, for instance? Some hints areprovided by the earlier work on synaesthesia, but we can probably go further.The kind of experience I am referring to would be a non-linguistic order ofcontextualization 'above' the denotative level of the linguistic text - a type ofnon-linguistic 'connotative' semiotic. Needless to say, more work needs to bedone on how this is established. Is the evocation of visual imagery when onereads the poem an attempt to create or invent a multimodal coherence forintegrating in (internalized) space-time the experience of reading the text? Thatis, is it a higher-order ('connotative') contextualization which functions to finda solution to the problems of local coherence posed by the text? In this way, thevisual imagery which is internally stimulated and experienced in the silentreading of a poem would be seen as internally stimulated perceptions at veryabstract cortical levels whose function is to provide solutions to sensori-motorincongruities and incoherences - the 'inexplicable swarm and equivocalgeneration of motions in our own brains', as Coleridge (1967: 77) expresses itwhich the text generates. The work of Shepard (1984) is suggestive in thisregard. I would say that problems of this kind only foreground what is in fact amore general phenomenon whenever we read a text of any kind.

Reading is an activity which requires the space-time integration of the readerwith the text. It is an ecosocial activity even though (1) physical movement isslight, being limited, perhaps, to the scanning by the eyes of the written notationon the page and the turning of pages by means of a synergy of hand-arm-jointeye kinaesthesis; (2) it is often 'silent' and contemplative insofar as one does notread aloud either to oneself or to others; and (3) much of the reading activity


takes place on the basis of internalized sensori-motor activity and resultingimagery. But silent reading is no less a multimodal and embodied ecosystemicactivity than is, for example, the activity of talking aloud. From the point of viewof internalized sensori-motor activity, we know from modern neurophysiologyand neuropsychology that a cooperative synergy of sensori-motor modalities isharnessed for the purposes of the space-time integration of participants in someecosocial act of meaning-making. These entail a multiplicity of complementarycorporeal schemata which may be combined and associated in many contextspecific ways for the purposes of such integration. There are, I suggest, nonlinguistic modalities - visual, for example - of 'inner' semiosis which are relevanthere. They are, I think, specialized deployments of the same 'outer' resources weuse in talking, depiction, movement, and so on. Again we see the principle,discussed above, that simulated inner movement cross-couples with otherperceptual and semiotic modalities and helps to guide the emergence of innersemiotic activity.

There are, then, different levels of organization - neural, sensori-motor, biokinematic, expression form - each with its own scale of physical size and rate ofchange. Each in turn is both regulated by the level(s) above it and at the sametime regulating the levels below it. In this perspective, there is no abstract centralprogramme or mechanism which must then be translated into muscularmovement, and so on. Instead, the four levels of organization mentioned aboveare organized into larger-scale 'synergies' (Bernstein 1967) which are mutuallyregulating at all levels. As Salthe (1993: 48-9) puts it, the higher-scalar levels inpart 'complete' or 'integrate' the lower ones by conferring on them a localmeaning which they would not have on their own, Thus, neural activity isintegrated with sensori-motor activity which is integrated with skeletomuscularkinaesthesis, which is integrated with the semiotic categories (phonological,graphological, etc.) of some expression form.

The processes of integration described here reflect the interdependencebetween the topological-continuous properties of the lower-level bio-physicalprocesses and the typological-categorical properties of the expression plane inrelation to these. The latter are based on discrete contrasts. Lemke (2000a: 203)points out that the typologizing properties of semiosis tend to spread or cascadethroughout the world, invading the world with semiotic categories, particularlywhen harnessed to some artefact which enhances their use. However, no suchprocess can ever completely exhaust the meaning-making potential of thetopological substrate, i.e. the bio-physical dynamics of the body in human socialinteraction.

As I shall argue below, the metafunctional hypothesis provides a way of talkingabout the various dimensions along which both the typological-categorical andthe topological-continuous are integrated with each other from the point ofviewof the expression plane. Thus, the bio-physical kinematics of the body are notread or interpreted in exclusively typological-categorial terms. The body-in-itselfis noumenal or pre-semiotic until it is construed as meaningful by some interpreter. This means that there is a relationship of complementarity between thisbio-physical semiotic potential and the interpreter who reads it. In itself thispotential is, so to speak, indifferent and depends on a socio-cultural interpreter


and system of interpretation which can harness and direct its thermodynamicmatter-energy flows in socially relevant ways.

9. The Semiotic Mediation and Entraining of Embodied Bio-kinematicPotential

As I said earlier, the expression plane is also integrated with the content plane ofsemiosis. In the case of language, this means its lexicogrammar and semantics.An expression without a content is a contradiction in terms and no sign-makingcan take place in the absence of one or the other of these levels of semiosis. Thecontent stratum is a higher-order level which further completes or integrates theexpression stratum. The lexicogrammatical mediation and integration of, say,speech sounds means that the bio-kinematic interface is partially integrated witha system of interpretance which links embodied socially shareable meanings tothe wider ecosocial system and its interpretation. By the same token, the lowerlevel interface between the body and the expression plane is not simplycompleted by the higher-level content stratum, but is also a model of it. Thishelps us to see more clearly that the expression plane is itself metafunctionallyorganized (see section 12, pp. 90-4).

From the point of view of articulation, we have seen how the bio-kinematicsof the body are brought into play in order to realize a given act-type. In thepresent perspective, this may, indifferently, refer to vocal-tract activity in the actof speaking or limb, muscular and other systems which I use in order to carry anobject across the room. In both cases, the proximate bodily activities of respiration, muscular and other activity in the vocal tract, visible lip and other facialmovements (in the case of speaking) and the interplay of anatomical (muscles,nerves, limbs, bones, joints) and time-bound constraints that selectively modifydegrees of freedom of movement according to a trajectory in the case oflocomotion provide lower-level constraints - 'from below' - which are thesystem's initiating conditions (Salthe 1993: 214). It is the interaction of the twosets of constraints - time-dependent or dynamical and time-independent which allows the system to anticipate further possibilities and, hence, to act indeterminate contexts.

Thus, the lower scalar level of the body-brain bears the corporeal traces - the. memories - of previous social interaction (Threadgold 1997: 97-103). Thesememories serve both to maintain the overall stability of the system in the face ofexternal fluctuations and to provide internal models of the body and its relationsto its Umwelt (Berthoz 1997: 7). In this sense, individual organisms have arepertoire of movement schemata which are activated and deployed accordingto the dynamical requirements of the movement as it takes place in time andspace. Salthe makes the further point that such models or schemata '... arelower-level to the organism as a whole, and certainly their dynamics are vastlyfaster than the environmental dynamics they could be relevant to, therebyallowing anticipation of environmental changes' (1993: 215). Rather than sayingthat movement is physical, whereas language is cognitive or psychological, wemay see how both movement and language are forms of actional semiotic which


directly mediate and entrain the lower-level bio-kinematic systems which aretheir expression planes.

However, the expression plane is not an objective or ostensive domain in thereductively Newtonian sense. Instead, it specifies the bio-kinematic potential ofbodily synergies to be harnessed and channelled in specific directions for thepurposes of social meaning-making. This means, in other words, that the neurophysiological and motor activities that are variously involved in both movementand speaking are constrained by higher-order boundary conditions, acting as asystem of context-sensitive constraints on the activities of lower levels. In thisway, a potentially very large number of degrees of freedom of movementare channelled into a relatively small repertoire of gestures which exhibitstructural stability from one occasion to another in spite of many situation andperformance-specific variations in their use. Furthermore, rather than postulating more and more layers of formal constraints to explain this as in much phonological theory, it is possible to describe a much more direct relationship betweenmotor activity and its mediation by lexicogrammar and semantics. This bringsabout the possibility of a new rapprochement between the bio-physical and thesocial semiotic dimensions of language and other modalities of social semiosis,Thus, bio-physical systems are dynamic, open and adaptive systems that engagein constant exchanges of matter and energy with their environments. They arealso comprised of nonlinearly related internal subsystems that may stochasticallyfluctuate. This leads to a transformation to some new global structure when theinternal dynamics of the system are coupled to, for example, a critical change inenergy flow. The emergence of a new global structure marks a correspondingreduction in the degrees of freedom of the system's internal dynamics. Further,the corresponding increase in stored information means that the informationalcross-coupling potential of the many subsystems comprising its motor activitypotential are reduced.

According to homeokinetic theory (Kelso et al. 1986), this transformationfrom a high-energy source to a low-energy sink functions to stabilize the systemas an ensemble of nonlinear oscillators. This means that the main organizingprinciple underlying motor activity is that of rhythmical periodicity. Motoractivity and the articulatory forms that emerge from this are shaped by the biokinetics ofweakly coupled oscillators of varying periodicities, Furthermore, vocaland other forms of articulatory activity are describable in terms of nested cyclesof rhythmic periodicities on many different timescales which mutually modeland entrain each other.

In the case of speech, it is the breath group or the tone group (Trager andSmith 1951; Halliday 1967a; Laver 1970: 68-70; Lieberman 1984: 118-20;Edelman 1989: 174-5; Coulthard 1992) which is the primary unit responsible forthe regulation of the energy which is stored, released, and dissipated inphonation. The breath-group theory proposes that the articulatory cycle isorganized by grouping around six or seven syllables on average into completeexpirations which are further characterized by having pauses at their boundaries(Laver 1970: 68). As Halliday and Coulthard, among others, have shown, thesetone-group or breath-group boundaries, which are organized in terms ofa respiratory cycle, are functional in the organization of spoken discourse into


units of information. That is, bodily processes are both integrated into andmediated by lexicogrammatical and discourse semantic factors on the contentstratum (see also sections 13 and 14, pp. 94-100). The following short analysisserves to illustrate the structure of the tone group. Following Coulthard (1992),tone units' boundaries are marked by a double slash; the tonic segment is inupper case:

Dion: / / what's he SPILLT it / / alREADy?

/ / look at the MESS / / PAUL / / you're gonna have to CLEAN UP / / inHERE / / it's a PIG STY / /

Paul: / / it is RATHer a PIG STY / /

/Dion: / YES / / WELL / / it's YOUR ANimal / /

The breath group is the major phonological unit which carries both intonationand rhythmic periodicity. Lieberman points out that rhythmic periodicity'follows from the fact that we can generate a steady subglottal air pressurefunction throughout most of the expiration' (Lieberman 1984: 119). It isnecessary, as Lieberman (1984: 119) further points out, that this subglottal airpressure be stabilized in order to avoid 'uncontrolled variations in perceivedpitch and amplitude'. Neonates are unable to achieve this degree of controlwhen they cry because of their inability to control the fundamental frequency ofphonation during the cry. At the age of around three months children begin tolearn to control 'their subglottal air function throughout the duration of theexpiratory phase' (Lieberman 1984: 119). This occurs as the infant becomesmore attuned to the intonational and rhythmic regularities in the ambientspeech of parents and other caretakers.

For example, Vihman and de Boysson-Bardies (1994; see also Vihman 1991)have shown that phonological structure is an emergent effect of the dynamicinteraction of many different factors on different scales - the neuromotor andperceptual systems, individual experience with language, and the ambientlinguistic influences of caretakers and others. The child acquires new habits asthe stored information in this newly emergent level increases. In this way, thechild's motor activity is entrained to and becomes more specialized to therequirements of a given phonological system.

But I do not see this process as one of learning abstract and formal phonological rules. Rather, I see it as one principle among others for constructingthose global and potentially shareable principles of coherence which allow us todialogically co-ordinate our bodily activities with those of others. Thus, a phonological system is a system of categorial, rhythmic and other regularities in thespeech practices of some community which, as seen from the expression planepoint of view, synchronize both peripheral and imagined speech motor activitieswith those of the other or the non-self - real or imagined - in the Umwelt. It is ameans of harnessing friction to productive semiotic ends. This occurs against abodily background of those primal vagaries and indexical necessities that are not


assimilable to the symbolic purpose to hand. In this way, meaningful bodilymovement - vocal-tract activity, manual-brachial gestures, etc. - always entails aconstant interplay of the regular and the surprising. The latter interferes withthe predictability and stability of the former, in the process contributing to oursense of pleasure in movement based on the interplay of these two elements.

However, this process of integrating the vagaries of the child's immaturemotor activity to the supervenient level of the adult phonological system doesnot mean that prior, non-phonological dimensions of articulation are altogethertranscended (Salthe 1993: 213). In phonation, the breath group functions as amotor schema whereby the brain simulates and predicts vocal-tract activitybefore execution takes place. It is in this sense that Laver (1970: 68) writes of'anticipatory adjustments in articulation'. The basic schema thus comprises anarticulatory movement sometimes bounded by pauses, and in which there is oneprominent syllable characterized by a major change in pitch in intonation.Typically, there is also a fall in fundamental frequency and amplitude at the endof the breath group, though this tendency may contrast with a rising pattern atthe end of the breath group (Lieberrnan 1984: 130).

The breath group is an articulatory prosody. It is a prime candidate for whatCarol Fowler (1986: 5), following in the tradition of Gibson's ecological theoryof event perception, calls an environmental event. Such events are sources ofenvironmental information which can stimulate the perceptual systems of theperceiver. They have 'affordances' or possibilities of interaction with theperceiver. In the case of speech, the environmental event is the vocal-tractgestural activity of the speaker.

10. Metafunctional Diversity on the Expression Plane

In this section, I shall explore the possibility that the expression plane ofsemiosis, no less than the content plane, is organized metafunctionally. Themetafunctional explanation oflanguage form was briefly outlined in chapter 1,section 10, pp. 46--9.

In my view, there are analogues of experiential, interpersonal, textual, andlogical organization and meaning on the expression plane. That is, corresponding principles of metafunctional diversification operate on the expressionstratum in relation to the topological-eontinuous substrate of the articulatorydynamics of the body which these entrain and construe as expression-substance.I am treating the expression plane as a class of physical-material events which areconstrained by the sensori-motor and perceptual activities of interactants as wellas being motivated by and mediated by higher-order social semiotic factors,expression-form (phonology, etc.) in the first instance. Rather than the abstractformalisms of much phonological theory, with its emphasis on levels of formalrules and the psychological reality of these, a metafunctional description of theexpression stratum of semiosis must account for the ways in which the expressionstratum interfaces with and construes bio-kinernatic phenomena of the humanbody in interaction with other bodies on various intersecting scales of spacetime. Thus, it would need to account for phenomena such as the following: (1)


the spatial displacement along a vector-in-time of the articulators in order toachieve specific co-ordinate structures or gestures and the phonemes, syllables,etc. that these articulate; (2) the ways in which modulated rates of change (inpitch, rhythm, movement, and so on) 'index' feeling states, physical sensations,and so on of the body; (3) the ways in which sensori-motor activities, respiration,muscular activity, limb movement, etc. are harnessed and entrained as body workwhich is directed towards the making of socially shared signifying acts in contextually relevant and coherent ways; (4) the ways in which such bodily events arerecognized as having social semiotic significance by others.

There are a number of distinctive ways in which the expression plane constitutes the interface between body and ecosocial environment. I shall develop thispoint in the next section. I shall return to the question of the metafunctionalorganization of the expression plane in section 12.

11. The Expression Plane is the Interface between Body and EcosocialEnvironment

The notion of the expression plane of semiosis as the interface between the bodyand its environment obviates the traditional split between an objective physicaldimension and a subjective sensori-motor dimension. The fact that theexpression plane 'faces both ways' - i.e. towards the body and towards the environment - itself provides the solution to this problem. Thus, from the point ofview of both articulation and perception/reception, the expression plane is anecological event which links or cross-couples the body-brain complex to its environment (Thibault 1997a: 158-60). There is, then, no objective input to theorganism which is proportional to its behavioural output. Instead, informationin the environment of the organism is always relative to the organism. This helpsto break down the antinomy between expression-substance (e.g. phonetics) andexpression form (e.g. phonology) (see Petitot-Cocorda 1985: 97-102). As aninterface between body and environment, the expression plane plays a muchmore constructive role than that of mere information transfer between anorganism and its outside. In actual fact, it is a more highly specified (socialsemiotic) elaboration of the biological principle that organisms selectively andactively attend to their environments on the basis of their own internal, biophysical organization. As an interface between body-brain and environment, theexpression plane construes a relation of 'correspondence' between the two. It isa concrete demonstration of the principle that the organism corresponds 'functionally and morphologically' (Prodi 1987: 2; see also Bateson 1973a: 285-91;Gibson 1986 [1979]: 8) to its surroundings. In the process, the organismmodifies the environment in virtue of its active and dialogic engagement with it.It also modifies its own internal organization. Rather than a transfer of information from organism to environment (articulation) or from environment toorganism (perception), there is a continual process of 'translation'. Forexample, a smell or odour in the environment will, when detected by thereceptor cells in the nose, provoke a chemical change in these same receptorcells of the organism.


The construal of an expression-substance in and through its correlativeexpression form means that an ecological event such as an articulatory act andits reception is always a translation into symbolic signs of the indices of environmental events. There is no necessary or causal relationship between the physicalstimuli and the internal changes these bring about in the organism. For thisreason, the relationship between the two is one of symbolic transduction ratherthan information transfer (Salthe 1993: 17&-7). The environmental informationafforded by the articulatory event function as indexes of that event. However, theexistence of shared models of the world or systems of interpretance which arestored in the central nervous system of the organism constitutes the resourcewhereby interactants use and construe these environmental indices for higherorder symbolic purposes which are not tied to indexical necessities. Thesesymbolic resources are stored in the body-brain complex as schemata which notonly interpret information from the environment (perception) but alsofunctions as the models for predicting future actions and their consequences inthe environment as well as acting in it. In part, they do so on the basis of thestored memories of the past consequences of such actions. In this sense, theschemata regulate the sensori-motor activities which project indexical information back into the environment where, however, its lack of any necessaryconnection to particular external events means it can be interpreted symbolicallyby suitably equipped agents. This is so in the case of the speaker's vocal-tractactivity, which, even while providing important indexical information about thespeaker's bodily states, is not interpreted as being uniquely tied to these onaccount of its further possibilities for symbolic reconstrual. It is always crosscoupled with other modalities of perception-action as well as other scales ofecosocial space-time, along with the meanings that are afforded by these diversescalar levels of organization.

The expression plane is, then, an interface which is characterized by thecontinual exchange which is activated by bodily processes such as respiration,bio-kinematic movements of articulators of various kinds, their modulations, onthe one hand, and the environmental information these produce for suitablyequipped perceivers, as well as their possibilities for symbolic reconstrual in waysrelevant to the ecosocial context in which they occur. The articulatory act quaecosocial event specifies its informational structure to the perceiver; it does nothave to be enriched or re-elaborated as in the inferential model of perception(Gibson 1986 [1979]: 251-3). However, this does not fully explain the symbolictransduction of this physical event into a social semiotic one. For this, it seems tome that we need some criteria of the values which selectively guide and constrainbehaviour. The perception of vocal-tract and other forms of gestural activity doesnot take place in a value-free vacuum. Rather, criteria of value specify that someevents are more important, more relevant, more interesting, and so on, thanothers. Edelman (1989: 98-100) has shown the importance of biologicallyin-built criteria of value which selectively channel the organism's activities alongsome pathways rather than others in the interests of its survival and further development. That is, values enable the organism to make relevant functionaldiscriminations in its environment and to act accordingly. In such a perspective,the functional cross-couplings of classes of vocal tract and other articulatory


events with their social semiotic values is only a further elaboration of this samebasic principle. Values - biological and semiotic - together selectively channel,entrain, and motivate meaning-making activity in certain preferred directionsrather than others.

In recasting the expression plane as an ecosocial event, it is not difficult topostulate that articulatory act and perception are functionally complementary.There is information which is specific to both. This follows from the fact thatsuch events 'face two ways', viz. to the body of the articulator, as well as to theenvironment in which they are perceived, including, of course, the perceiver.The information specifying the articulator and the information specifying theperceiver are not opposed. Instead, they are simply two facets of the one overallphenomenon. In this sense, they are complementary. There is no need toresort to two separate descriptive languages to talk about the two as would bethe case in the various forms of neo-Cartesian mind-body and subject-objectdualisms.

Carol Fowler (1986) shows in her discussion of phonetic articulation as distalevents that are directly perceived that the perception of such an event entails theapprehending of the invariant structure of phonetic gestures as the vocal tractundergoes spatio-temporal change during articulation. The further questionthat arises concerns how the spatia-temporal eo-articulation of vocal-tract articulators and the acoustic information that results give rise to specific ecosocialsemiotic effects. What are the me ta-redundancy relations (chapter 1, section 6,pp. 26--30) among these ecosocial meanings, the sensori-motor action potentialof the body, and their eo-articulation on some particular occasion of interaction?This last question suggests the need to find a unified conceptual framework forexplaining the relationship between bodily processes and the social semioticsignificance and purpose of these.

AIl forms of gestural articulation have spatio-temporal properties. Theyexhibit properties of both temporal succession and spatial adjacency. A gesture,seen as an ecosocial event, exhibits a change in adjacency structure - viz. thespatial overlap of articulators - as well as the successive ordering of gestures intime. How, then, in the face of such spatio-temporal variability, are structurallystable forms recognized? What about seemingly 'static', object-like sign systemssuch as Writing? The predominantly spatial character of writing suggests minimalor no change of adjacency structure in time. However, the bodily activity ofscanning and orienting to the marks on the page or other treated surfacenecessarily requires movement as an integrating activity (see also Harris 1995b:46). In my view, the extra-somatic character of the treated surface on whichwritten notation is inscribed only conceals the event modulation which takesplace when a reader engages with the written text. Writing is not an exception tothis general point, but, perhaps, a limiting case. I would suggest that bodilymovement in one form or another is the fundamental organizing principlewhereby interacting bodies are integrated into both semiotic action performances and their textual records and products. The sensori-motor activities of thearticulator are constrained to informational invariants-in-change by theperceiver's ability to respond to these. Both articulators and perceivers mustlearn to distinguish between those forms of event modulation (change) that


bring about change of meaning and those that do not. In my view, a metafunctional account of the expression plane of such events (section 10) can help us tosolve this problem. I shall now return to this question in section 16 below.

12. The Metafunctional Basis of Vocal-tract Articulatory Activity

In the case of speech sounds, the experiential (or conceptual) dimension refersto the typological-categorical phonological distinctions made by phonemes andtheir combination according to phonotactic principles into larger phonologicalunits such as the syllable and so on. This dimension of speech sounds digitalizesthe acoustic stream in terms of discrete phoneme categories. Thus, the discreteor segmental character of these phonological categories shows how thisdimension of the speech sound is concerned with particulate or constituent-likephonological structures which can be described as a ranked hierarchy of smallerand larger-scale constituents. However, the assumption that phonological unitsare static and context-free mental entities which are divorced from articulatoryand acoustic properties of eo-articulated speech is rejected here (Thibault1997b: 4-5).

Alternatively, Browman and Goldstein (1991, 1995) base their gestural theoryof phonetic articulation on the notion of co-ordinative structures, as developedby Kelso et al. (1986) in movement theory. Browman and Goldstein argue thatarticulatory gestures are defined dynamically as 'the formation (and release) ofa constriction within the vocal tract through the movement of (a) a particular setof articulators, (b) towards a particular constriction location, (c) with a specificdegree of construction, and (d) in a characteristic, dynamically describedmanner' (1991: 315). From the experiential perspective, phonological gesturesconstrue and categorize actual vocal-tract phenomena which take place in thespace-time of articulation. That is, they construe phonetic events which may beanalysed into three main functional components, viz. (1) the dynamicalmovement in space-time of the articulators, (2) the particular articulators whichparticipate in this movement, and (3) the circumstances - cf. the degree andmanner of this movement - which are attendant on this dynamical movementand its participants. Typically, the speaker and the listener experience this eventas a single organized whole. However, the gestural schema which the speakerdeploys organizes the vocal tract in terms of both the articulatory processeswhich go on in space-time and the constituent parts of this process.

Armstrong et al. (1995: 14-15) also show that the manual-brachial gesturesused in sign language realize clause-level semantic structures such as [AgentAction-Patient] (cf. [Actor-Process-Goal] in systemic-functional linguistictheory) directly. That is, the combination of hand-and-arm-movement directlyrealizes the semantic structure in question by performing it. Thus, a hand mayperform the semantic function of agent which moves towards a patient. Likewise,the hand movement itself realizes the verbal process or the action. Suchsemantic functions are, of course, on the content stratum of organization.However, they further comment in connection with speech that:


It is impossible to see the same interrelation of phonology, syntax, andsemantics in spoken language, not only because most of the gesturesproducing speech are invisible, but also because of the difference between theperceptual and productive modes for speaking and signing. (Armstrong et al1995: 15)

In my view, the point is, rather, that there are complementary principles of(metafunctional) organization on the expression plane. In the first instance,these do not construe the semantic structures referred to by Armstrong et al.Instead, they construe analogous and parallel structures in the expression-substance of the speaker's vocal-tract and other articulatory activity. The furtherquestion as to how these relate to the content plane, which is the focus in thediscussion in Armstrong et al., will be taken up later. In the case of the manual-brachial articulatory gestures discussed by Armstrong et al., the movements aresequential in space-time, as seen from the lower scale of the expression plane,whereas the lexicogrammatical and semantic functions that map onto theseare multilayered. This contrasts with the vocal-tract gestures of speech, which areco-articulated or overlapping, rather than serial, from the point of view ofthe expression plane, whereas the morphosyntactic structures realized on thecontent plane preserve a fundamentally sequential structure (Studdert-Kennedyand Lane 1980: 35-6).

Individual gestures may be seen as typologically similar to phonologicalsegments such as phonemes. However, and unlike segment-based theories ofphonology, 'there are no a priori constraints on intergestural organization withinthe gestural framework' (Browman and Goldstein 1991: 319). The notion ofco-articulation refers to the fact that vocal-tract gestures exhibit considerablespatio-temporal overlap. There are, then, differences of both kind and degree ofoverlap and these differences may be phonologically salient in distinguishingone gesture from another. To explain the role of gestural overlap in the articu-lation of vocal-tract gestures, Browman and Goldstein have developed the notionof a gestural score, which is reproduced in Figure 2.1.

Figure 2.1: Example of gestural scores for /pcen/ ('pan') and /been/ ('ban'); borrowedfrom Browman and Goldstein (1995: 189)


Gestures, in the view of these researchers, are not necessarily confined todiscrete phonological segments, seen as the natural place of residence ofparticular constellations of gestures. Rather, an 'entire constellation of gestures'may be 'phenologically' salient. The point is that such gestural constellations arecomprised of various parts, all ofwhich function together, in synergy, to producethe entire gestural constellation. The presence or absence of particular gesturesthus specifies the smallest-scale ways in which the vocal tract is modified in theproduction of particular phonological categories. Gestures, it is important topoint out, refer to higher-order semantic construals ofthe actual spatio-temporalmovements of the vocal-tract articulators (Browman and Goldstein 1991: 323).They are, then, schematic to the actual movements of the articulators. Thus, aspeaker's knowledge of a word also includes 'a specification of its gestures andtheir organization' (1991: 324). Thus, much of the gestural plasticity andvariation which is evidenced in casual conversation arises through theconcurrent production of overlapping gestures, rather than the referencing ofdiscrete phonetic rules (1991: 324). The schematic character of gestures meansthat they exhibit structural invariants irrespective of variations which may occurduring any particular instance of their vocal-tract execution. By the same token,the gestural score emphasizes that gestures also exhibit spatio-temporalinvariants. That is, either the spatial or the temporal dimension of the gesturalscore may be modulated in ways that are invariant. In other words, gesturalscores have both structural and transformational properties which are schematicto specific instances.

The interpersonal dimension of speech sounds is prosodic or field-like. Itis concerned with the ways in which prosodic contours deform the acousticarticulatory shape of the utterance as a means of communicating the speaker'sattitudinal or affective stance or relationship to the utterance or to one'sinterlocutor. Intonation is the main resource which has been systematicallystudied by phonological theory. Intonation is prosodic because its articulatoryand acoustic effects are not localized in any specific (phonological) constituent.Instead, intonational prosodies have as their articulatory and acoustic scope theentire breath group or tone group. From the perspective of the expressionplane, articulatory prosodies deform or modulate the phonological shape sensastrictu of the specific phoneme selections and their combinations. Bolinger(1985) has written of the 'iconicity' of intonation in this sense. That is,intonation, in Bolinger's account, directly registers changing feeling states in thebody of the speaker (see section 14, pp. 98-100).

The textual dimension of speech sounds is wave-like or periodic. It has to dowith the nested rhythmic periodicities that delimit the beginning and the end of,say, the breath group. This periodicity is temporal. Thus, the rhythmicmodulation of respiration in the breath group in terms of a beginning-end cyclemarks the boundaries of the modulation itself. Such punctuations of thetemporal order of succession may provide a means of temporally co-ordinating'the reentrant cortical areas mediating the phonological, syntactic, and semanticlevels of speech' (Edelman 1989: 179), as well as synchronizing interactants intothe spatio-temporal rhythms of the speech event at the level of the timing andthe synchronization of turn-taking.


These rhythmic periodicities have their basis in the respiration cycle and are,for this reason, based on specific rhythmic patterns which are engendered bybodily processes and functions. Rhythm may also be modulated and theprincipal modulating factor would appear to be velocity or change of rate of, say,breathing in the case of speech, acceleration along a movement vector of a handor an arm in the case of gesture, and walking, and so on, in the case oflocomotion. Change of rate applies to some dimension of the spatio-temporalparameters of articulation of the perceptual event in question. Here, I amassuming that the expression plane of semiosis is always a perceptual event insome sensori-motor modality. This is so from both the execution and receptionpoints of view. In other words, it is the expression plane of semiosis which crosscouples the semiotic act to the perceptual and motor activities of the participantsto the act in question. Semiosis is always embodied in this sense. No act ofsemiosis floats free from or transcends this somatic dimension of its executionand reception because it is the expression plane which serves to integrateembodied participants into the meaning-making act itself. This is so in the sensethat the expression plane is the interface which cross-couples the sensori-motoractivities of the body (firstness) to the environment or non-self (secondness) inand through the higher-order systemic resources of some semiotic resourcesystem (thirdness) (see chapter I, sections 3, 8, pp. ll-18, 34-9).

Finally, the logical metafunction is based on the principle ofinterdependency.Its structuring principle is that of recursion. This follows from the spatiotemporal character of articulation. Probably the simplest form of interdependency is that of succession in time. In speech production, the articulation ofphonetic segments is characterized as 'overlapping sets of coordinated gestures,where each set of coordinated gestures refers to a phonetic segment' (Fowler1986: ll). This view obviates the .requirement that a priori planning or controlprocedures existing in an ontologically distinct mental realm act as the formalcauses of articulation. Instead, planning occurs as part of the on-line andcontext-sensitive production of phonetically structured articulations.

The breath group is an example of an event simplex. However, eventcomplexes may also be built up from such simplexes through processes ofnesting, addition, co-ordinating, subordinating, repeating, and so on. Forexample, rhythmic periodicities of different temporal duration may be nestedthe one within the other (Martinec 2000). Again, a movement sequence may benested inside a still larger sequence, and so on. The eo-articulation of articulators in the vocal tract in the performance of a given gestural score involves theco-ordination ofvarious articulators which are independent of each other. Otherevent complexes may involve the sequencing of one event after another. Forexample, one flips the power switch before the computer boots up ready for use.The question that arises in all cases is how and to what extent the informationwhich specifies the event is preserved under such conditions of temporal andcausal transformation. Lindblom (1991: 19) emphasizes, with his principle of'sufficient contrast', the plasticity of phonetic gestures rather than fixed criterialproperties, which are based on the contrary principle of 'maximal contrast'. AsLindblom points out, the latter notion has dominated in the discussion of the


perceptual (phonological) salience of phonetic gestures. That is, time-bound'production contrasts tend to counterbalance demands for perceptual contrast'(Lindblom 1991: 19). Rather than criterial properties, there is a context-specificinteraction of many different factors - perceptual, articulatory, etc. - whichmeans that the ecosocial event is an emergent property of the time-boundecosocial context. They are not structured by fixed a priori perceptual or otherproperties, but constitute what Lindblom calls a 'variable and adaptive means'(1991: 21) for the purposes of ecosocial interaction.

Vocal-tract activity, as I pointed out above, is a form of bodily movement. Itinvolves the harnessing and control of specific skeletomuscular and other bodilysystems so that the acoustic information we produce is appropriately andproductively deployed as we act in and upon our Umwelt. This emphasis onembodied, physical movement highlights the fact that vocal-tract activity is also aforceful physical-material interaction with our environment, including thosewith whom we interact. The metafunctional analysis in this section suggests howfactors such as the dynamical movement of articulators, intonation, rhythm, andother features are all specific dimensions of the ways in which vocal-tract activityis a form of forceful acting on environmental friction or secondness. Abstract,symbolic notions of language too often lose sight of this fundamental insight.Social agents do not simply engage in abstract symbolic 'thinking' or 'cognition'whenever they talk. They also deploy their bodies and its extra-somatic projections as a means of engaging in forceful interactions with the world of thenon-self. They must modulate and control these bodily forces in ways which arenot separable from the lexicogrammatical mediation ofvocal-tract activity. In thenext section, I shall discuss an example in order to illustrate this principle.

13. Subjectivity, Agency, and the Prosodic Realization of InterpersonalMeaning

The interpersonal metafunction is a resource for intervening in, acting upon,and interacting in the world. Its mode of expression is not segmental, butprosodic (Halliday 1979; Matthiessen 1990). Interpersonal prosodies are notconfined to features such as pitch contour on the phonological stratum, but alsooccur on the lexicogrammar and discourse semantic strata. There are also visualgraphological prosodies. Matthiessen (1990: 21), drawing on a distinction firstmade by Robins (1957), identifies two types of prosody: (1) prosodies realizedcontinuously; (2) prosodies realized at boundaries. The first type refers tofeatures whose domain extends beyond the point of realization of the feature, tocover the whole or a part of the whole structure. At the rank of syllable on thephonological rank scale, examples include stress, pitch, and length. The secondtype includes features which delimit some structure by marking what goesbefore, and what comes after. In so doing, such features prosodically extend overthe structure they delimit.

Prosodies are similar to Pike's 'hyper-unit' or field view of sequence: '... a"sequence" of two units is treated as a single hyperunit with no essential requirement of internal specific segmentation or specific peak identification' (1967:


553). Criteria of segmentation and peak identification correspond to Halliday'sexperiential and textual metafunctions, respectively. Halliday (1979) andMatthiessen (1990: 22) argue that prosodies, which characterize units such asthe clause as a whole, act like a continuous motif which colours this whole. I fullyagree with this characterization, but I should like to develop here the questionof interpersonal prosodies in specific ways, which are relevant to the centralquestions of this chapter. Let us consider the following Italian example, as shownin Table 2.3. This exchange was part of a larger sequence during which themother had repeatedly attempted to convince the six-year-old child to get up inorder to get ready for school.

The clause initial Vocative element selects for tone 1 (falling). The tone grouprealizing the pitch contour over the clause in the first part of the exchangeselects for tone 2 (rising), which is the typical tone for interrogatives. The child'sresponse selects tone 1. The falling-rising tone of the first speaker (the mother)operates as a single semantic unit, extending over the whole clause; it expressesthe contradictory nature of the interpersonal semantics, i.e. compulsion and adegree of uncertainty as to the modal investment in the exchange (,should I bemaking the child get up, or should the child be acting off her own bat?').

The example cited above also illustrates what Matthiessen (1990: 23) calls a'3rd order or semantic prosody': Ilaria ti vuoi alzare. The grammaticallysegmental elements in bold type realize a semantic prosody on the discoursestratum. Thus, the proper noun Ilaria, the second-person reflexive pronoun ti,and the morphemic realization of the semantic agent tu, which is fused with theverb vuoi in the lexicogrammar, realize a semantic prosody of this participant'ssubjective 'presence' in the discourse. In this case, it is a directional vector, whoseenergy extends from the speaking agent (the mother) to the 'making present' ofthe child as a (reluctantly, on this occasion!) volitional agent in the mother'sdiscourse. Thus, we see how the first-order (phonological) and third-order(semantic) prosodies are harnessed in the vectorial movement of energy whichoccurs, as the mother seeks to transfer the agency to the child, or enjoins thechild to take up this agency on her own account. In Italian, it should also be noted,this use of the reflexive, where the verb process concerns the whole of the Agent,rather than a part of the Agent (e.g. a part of the body), means that the agency isself-engendered; it originates, semantically speaking, 'from within' the Agent.

The co-referentiality of reflexive pronoun (Goal + Experiencer) and subjectpronoun (Agent) in such cases suggests a self-monitoring of this 'inwardly'attributed volition, which the English I get up does not realize. In this semanticenvironment, the reflexive pronouns in Italian are a systemic reactance of asemantic cryptotype of such self-monitored and self-engendered processes. InEnglish, the periphrastic I make myselfget up better glosses the Italian mi alzo. InTable 2.3, the reflexive pronoun conflates the transitivity roles of Experiencerand Goal. I have borrowed the first of these terms from Longacre (1983: 155).Experiencer is similar to Halliday's Medium (Halliday 1994 [1985]: 161-75), andis the participant which is centrally involved in the experience realized by theverb process. In the case of the Italian reflexives, I prefer this interpretationbecause it better glosses the semantics of the self-reflexive or self-monitoringrelation between the participant as both Agent and Experiencer of its own


Table2.3: Metafunctional analysis ofexchange unit: phonological, lexicogrammaticaland discourse semantic strata; tonic segments in upper case

Tonic 1 Pretonic Tonic 2 Tonic

ILARIA ALZARE

vuoi

ti NO

ILARIA YOU WANT TO GET UP NO

Vocative Comple- Subject + Residue Minor

ment Finite: clause:Modal: NegativeVolition Polarity

Goal- Agent Process:

Experienccr Material

Action

Theme Rheme

Given New

Starting Target Situation (agency imputed to orSituation transferred to addressee by addresser)

(agencynot yetimputed)

Directive/Exhortation to Act Non-compliancewithDirective

Speaker.: Mother Boundary Speaker;marking Six-year-old

change of child

speakingturn in

exchange


action. At the same time, the transitivity role of Goal (cf. Longacre's Patient) isconflated with that of Experiencer in clauses of this type. The Goal (or Patient)is the participant that undergoes some kind of change as a result of undergoingthe process. It seems necessary to make this three-way distinction in order todescribe the transitivity relations in such clauses.

Matthiessen (1990: 7) also characterizes prosodies in terms of the directionality of, say, the pitch movement. I would go further. This movement is thedynamic energy which is exchanged, and, in part, experienced as a movingforce, whenever a linguistic act is uttered. The giving form to a linguistic act alsoentails a dynamic relationship between this energy and the other strands ofmeaning which are configured in the linguistic form (see Cremonini andFrasnedi 1986: 23). Interpersonal prosodies are just one parameter of linguisticmeaning that shows the scalar heterogeneity of the bodily and semiotic processesinvolved. This would not be so if meaning were no more than an abstract formcontent relation. Meaning-making has an undeniable material and sensualquality, which, I think, can be explained by the relations between the energy thatis released in the production of the linguistic act as well as by its materiality. Thisenergy, as Cremonini and Frasnedi (1986) point out, has 'directional vectors',which both extend beyond the utterance and into its environment and feed backinto it, producing thereby a resonating density of significance. These authorsalso point out that this energy, in order to be effective, and not merelyredundant and 'empty', must be brought under control so that it be effectivelyharnessed and synchronized with the other dimensions of an utterance'smeaning. This is the task of rhetoric; rhetoric has the task of both giving directionality to and anchoring this energy in the dialectic with the other strands ofthe utterance's meaning potential. Lack of such directionality and control wouldseem to suggest emotional or other disturbances, pathology, or disorientation inthe communicative ecosystemic environment of agents.

It is this energy which contributes to that 'making present' of an embodiedsubjectivity that I mentioned earlier. It is the interface between expression andcontent whereby meaning comes into contact with the materiality of the body,and its neuroanatomical, neurophysiological, and neuropsychological levels oforganization. In so doing, the energy of the utterance gives both presence andforce to an immanent subjectivity/agency through this eo-articulation of thephenomenal-material and social semiotic domains. It is in and through thedialectically dual and complementary nature of the relation between the twodomains that body rhythms, physical pain, and so on, exchange energy with thesemiotic, and in so doing they give presence and energy to the social semioticvoicing of our subjectivity and agency (Bakhtin 1990 [1924]: 292). The energyand vectorial directionality of interpersonal prosodies endow our meaningmaking with the presence of a temporal event and a spatializing value. This isone concrete illustration of the principle that the matter-energy which isharnessed and channelled in meaning-making constitutes a forceful interactionwith the other.

Human meaning-making activity is a hybrid phenomenon: we do notunderstand and respond to it solely on the basis of the semiotic relations that weconstrue between the different parts of the activity in relation to the systems of


paradigmatic alternatives that are relevant to the contextualization of theunfolding text or discourse event. We also understand and respond to it on thebasis of the physical-material relations and processes with which these semioticrelations cross-couple and whose flows they entrain not only in the local contextof this here-now event, but also in relation to physical-material flows andprocesses on other spatia-temporal scales beyond the immediate here-nowcontext. To verbally abuse another person is more than just a 'choice' of oneparadigmatic option among other possible ways ('praise', 'encourage', 'comfort',and so on), more than just a component functioning in some culturally recognizable activity-structure (syntagm) such as 'intimidation', 'threat', 'harassment',and so on, more than just an index of some particular contextually relevant valueor social situation-type (gender relations, employer-employee relations, and soon) which is invoked as relevant to the meaning of the verbal abuse. It is also toinduce or bring about clusters of physical symptoms (e.g. racing heart, tensemuscles, trembling, upset stomach, and so on) which individuals associate incomplex ways with emotions such as fear, guilt, sadness, and so on, and in wayswhich, in turn, index social situations of, say,parent--child or gender domination,teacher-pupil control, and so on. In other words, our social semiotic practicesalso induce bodily sensations - both pleasurable and painful - in others, and inways which contextually redound with higher-order emotional states that arelinked to one's positioning within particular social activity-structures or discoursegenres, as well as to the accumulated bodily memories that the self collects alonghis or her life trajectory as a result of the various contingencies which constitutehis or her historical emergence as an individual.

The cross-coupling of the material and the semiotic, or, in other words, theconstant cross-scalar intermingling of these two facets of semiosis, shows how theindividual, in using the impersonal meaning-making resources available to himor her by virtue of his/her belonging to a given ecosocial system, manages todistinguish him- or herself from his/her surroundings in the specific act ofmeaning-making. In using the standardized resources of the system, theindividual is inevitably differentiated from his or her surroundings. He or sheinevitably indexes the specific states of his/her brain-body in a given time andplace at the same time as the behaviour of the individual is cross-linked tohigher-scalar, hence impersonal, levels of organization in the ecosocialenvironment.

14. Vocal-tract Gestures and Grammar: Symptom and/orSupervenience?

Bolinger views intonation as iconic or symptomatic of emotions and theirdegrees of intensity. In other words, intonation is directly 'presentative' ofembodied feeling states of the speaker (1985: 98). Bolinger explicitly arguesagainst the 'grammaticalization' of intonation, claiming as his starting point that'high pitch symptomizes a condition of high tension in the organism, low pitchthe opposite' (1985: 99). In this account, intonation and its meaning has itsorigin and its basis in 'a gestural complex whose primitive and still surviving


function is - however elaborated and refined - the signalling of emotions andtheir degrees of intensity' (1985: 98). I agree with this claim entirely as far as therelationship between speech sounds and the speaker's embodied feeling states isconcerned. Nevertheless, I do not think that Bolinger provides an entirely satisfactory account of the ways in which intonation is also integrated with andmediated by lexicogrammar. Intonation is not simply a 'primitive' gesturalresidue of speech which, phylogenetically speaking, preceded language andthen merged with it. That is why the term 'paralinguistic' is, in the final analysis,misleading: intonation does not simply accompany language, but is integratedwith it.

This would not in itself explain the teleological function of intonation in, say,the interpersonal dimension of meaning. Intonation is often the crucial distinguishing factor in construing what kind of dialogically co-ordinated move thespeaker is uttering (Halliday 1967a: 21-2). The biological differentiation whichBolinger sees as fundamental to intonation itself undergoes significant reorganization as the social semiotic dimension of intonation emerges more clearly.Lukacs most insightfully explains this general process in the following terms:

The division of labour is originally based on the biological differentiation ofthe persons who form the human group. The receding of the natural barrieras a consequence of the social being's becoming ever more strictly and purelysocial is revealed, above all, in the fact that this principle of differentiationwhich is biological in origin, assumes in itself ever more numerous momentsof sociality and ends up acquiring in itself a first order importance, downgrading to secondary status the biological moments. (Lukacs 1981: 138; mytranslation)

In this quotation, Lukacs addresses the question as to the ways in which a socialsystem emerges from a biological one. His argument expresses the logic of thespecification hierarchy of integrative levels (chapter 1, section 8, pp. 34-9).Bolinger's perspective clearly does not deny the emergence of intonation frombiological systems in the individual organism. He argues against those who seeintonation 'as part of the abstract code' (1985: 99) on this very basis. What hedoes not explain is the effect on intonation of its integration with the contentstratum of lexicogrammar and semantics in language. If intonation had simplyremained 'symptomatic' of the feeling states of the speaker, then there would beno significant traces of its reorganization as a consequence of its integration withthe linguistic semiotic. If, on the other hand, the division of social labour entails,as Lukacs argues, the increasing specialization of action for social purposes, thenthis would suggest that with its progressive integration into the linguisticsemiotic, intonation itself becomes more specified as a semiotic resource whichis able to realize teleological acts the aim of which is to bring about in others 'thewill to carry out determinate teleological positions' Lukacs (1981: 139).Certainly, the crucial role that intonation plays in many languages for distinguishing various categories of dialogically co-ordinated move would seem to bearthis out.

As Lukacs further points out, such acts, if they are to prove effective, require


that other persons in the human group recognize their significance and teleological intent. That is, there are higher-order boundary conditions of a socialnature which provide the conditions whereby the otherwise purely symptomaticsignals from the individual's body are amplified and adaptively modified associally shared resources for making meaning. The social power of a given bodylies in the ways in which others take up and further amplify and modify in sociallyentrained ways and on many different scales the effects of the bio-kinematicenergy, rhythms, and movements of 'powerful' individuals.

This does not mean that the prior 'symptomatic' layer of bodily significancehas been superseded or transcended. Rather, intonation both retains importanttraces of its earlier ancestral inheritance and at the same time has undergone significant restructuring and modification in the course of its integration into themore highly specified linguistic semiotic. Only in this way can we explain howintonation, from the systemic point of view, has itself emerged as a highlyspecified system ofvalue-producing, or semiotically salient, differences in a givenlanguage system. How otherwise might we explain the complex distinctions inmeaning that intonation makes in relation to grammatical, facial, kinesic,musical, and other semiotic modalities?

15. The Intentional Character of Articulatory Activity

There has been considerable discussion in recent years as to whether the objectof perception in speech sounds corresponds to vocal-tract articulatory gesturesor to the acoustic signal which results from this. This discussion hinges on thefurther question as to whether vocal-tract activity is merely the surface manifestation of pre-programmed and invariant control structures for phoneticgestures, as in the motor theory of Liberman and Mattingly (1985), or is itselflinguistically motivated gesture. In the motor theory, the control structures areabstract invariant properties that control articulatory movements. It is thesecontrol structures, rather than the movements themselves, that correspond tothe speaker's intentions (Liberman and Mattingly 1985: 23). The controlstructures, which are said to be encoded as abstract neural structures in thebrain, fail to explain that speakers' intentions cannot be seen as a priori abstractions of on-line vocal-tract activity. Rather, intentionally directed sensori-motoractivity always takes place in a dynamic and constantly changing context whichitself shapes and modulates the speaker's activity. Intentions are a function ofmany interacting contextual factors rather than an abstract cause of these. Motoracts qua intentional acts are understood, as Merleau-Ponty put it, 'in referenceto the aims oflife' (1983 [1942]: 163).

The structure of the acoustic array is the result of vibratory resonancepropagated from a mechanical source. Thus, the source of the acoustic signal the distal event in the environment - is strongly cross-coupled with the energywhich is propagated. It is this which provides the listener with information aboutthe distal event. The listener samples change in the acoustic array which providehim or her with information about the distal event. This event is the source ofthe disturbance in the ambient acoustic array and of any changes in its dynamics.


In other words, the acoustic array which the listener samples specifies the environmental event - e.g. the speaker's vocal-tract gestures - to the perceiver. In thisway, the listener uses this information to orient to the speaker for the purposesof dialogic interaction. It is not the physical properties of the acoustic signal persewhich are relevant to the listener. Rather, the linguistically motivated characterof the speaker's vocal-tract gestures - Le. the physical properties of the distalevent itself - is itself revealed to the listener in and through the informationwhich is structured in the acoustic array. This means that the perceiver directlyaccesses the gestural properties of the distal event of speaking through theinformation structured in the array.

Fowler and Rosenblum (1991: 36) cite evidence on the perception ofintonation (among other factors) to show that the listener's perception ofintonation peak corresponds not to a perception of the 'objective' or 'absolute'rate of opening and closing of the vocal folds during the production of theintonational melody, but to those modulations of the rate of opening and closingbrought about by gestures which the speaker intentionally uses to modify theintonational melody. The point is that listeners selectively sample the acousticvariables in the acoustic array as a guide to the linguistically organized and semiotically motivated gestures of the speaker. There would also appear to be a valuecomponent in this process (Edelman 1989: 152). The selective nature of thisorientation to the array suggests that the listener's response to these variables isan adaptive one. In establishing specific links between categories of acousticvariables and their linguistically motivated gestural values, the activities ofspeaker and listener are reciprocally modified and co-ordinated.

For proponents of the motor theory, the abstract control structures which arehypothesized correspond not to the speaker's articulatory movements but to hisor her intentions (see above). This leads to the 'encoding' view of the relationship between the control structures and the eo-articulation of phonetic segmentsin articulation (see Fowler and Rosenblum 1991: 47 for critical discussion). Theview of articulation-as-encoding-of-phonetic-segments assumes that speechsounds are segmental. On the other hand, the boundaries between discrete(phonetic) segments are distorted by their encoding into the continuous andoverlapping movements of the articulators in real-time vocal-tract activity. In thegestural view oflinguistically motivated vocal-tract activity, on the other hand, eoarticulation is seen as the functional cross-coupling of articulators which aretemporally heterogeneous. Rather than the presumed homogeneity of a serialordering of phonetic segments which is then distorted by its encoding in eoarticulation, there is a complex layering of gestures that sometimes overlap andsometimes do not in time (Fowler and Rosenblum 1991: 47). The lowest scalarlevel of phonetic organization, in this view, is not comprised of abstract phoneticprimitives. Instead, it is comprised of gestural prosodies which are modulated byhigher-scalar levels in the ecosocial semiotic system that speaker and listenershare. This is implicit in the following criticism of the motor theory of phoneticgestures made by Fowler and Rosenblum:

Do listeners need an innate vocal tract synthesizer to recognize acousticreflections of phonetic gestures? Although it might seem to help, it cannot be


necessary, because there is no analogous way to explain how observersrecognize most distal events from their optical reflections. Somehow theacoustic and optical reflections of a source must identify the source on theirown. (Fowler and Rosenblum 1991: 48)

Fowler and Rosenblum draw attention to the ways in which the motor theoryblack boxes the question of how the vocal-tract gestures identified in the distalevent functions on the higher-scalar level of ecosocial organization (see alsosection 4, pp. 72-5). The speaker's intentions can only be recognized as functioning in and through phonetic gestures because these very small-scalephenomena directly implicate much larger-scale processes on the ecosociallevel(see Lemke 1997 for critical discussion of this important point). This is what thenotion of an ecological event in Cibson's (1986 [1979]: 93-110) sense in factentails. The black boxing that I referred to above occurs, for example, when aninnate vocal-tract synthesizer is postulated, as in the motor theory of phoneticgestures, as the explanation for the ways in which interactants recognize andinterpret gestural activity in distal events such as the speaker's vocal-tract activity.Thus, the black boxing of the higher levels in the form of a presumed vocal-tractsynthesizer reduces this interplay of scalar levels to a single level of preprogrammed central-processing units, seen as hard-wired modules in the humanbrain. Consequently, it fails to show how these smaller-scalar events in the vocaltracts of speakers are directly implicated in larger-scalar dynamics which mayinclude the individual's own individuating trajectory, the speech practices of thecommunity, the shared interactional practices of particular speaking dyads, andso on.

For a theory of vocal-tract gestural activity, this has a number of implications.First, the question of the speaker's production of phonetic gestures and thelistener's perception of these in the distal event are seen as complementarydimensions of a single larger-scalar ecological event that obviates the need forany reductively bottom-up causal explanations from within the individualorganism. The ecosocial character of the event means that both speaker andlistener perspectives are necessarily implicated in it. Secondly, the mediation ofphonetic gestures by their higher-order lexicogrammatical and semanticmeaning relations which construe the larger-scalar phenomena of the world inwhich speakers and listeners live means that there is a high degree of direct,unmediated and fully motivated relations between the smaller-scalar dynamics ofthe former and the meanings these transparently have for interactants. That is,higher-scalar semantic meanings and the large-scalar historical and culturalrealities these index are directly implicated in lower-scalar phonetic gestures forboth speakers and listeners. Indeed, Studdert-Kennedy and Lane (1980: 36-8)advance arguments to show that both phonetic gestures in speech and manualbrachial gestures in sign language exhibit serial properties of motor executionthat are directly implicated in the realization of morphemic distinctions. This isso in spite of the parallel processes of eo-articulation which are also evidenced.


16. Embodying the Metafunctions: The Example of Vocal-tractArticulatory Activity

My focus on the metafunctional dimension of the expression stratum highlightsthe embodied, physical-material basis of meaning-making. The vocal tract is atopological-continuous dynamical system which admits of complex attractors, asdiscussed by Kauffman (1993: 177). Thus, the metafunctional organization ofthe expression plane provides a framework for describing the multiple andvariable ways in which trajectories within the topological substrate of the vocaltract may converge on a set of states over the space-time span of the articulatoryact. Is it possible to describe the four metafunctions as trajectories which flowthrough particular basins of attraction in the limit cycle around which themovements of the articulators flow during articulation? That is, are the metafunctions comparable to sets ofvariables which exhibit their distinctive repetitiveoscillations (particulate, wave-like, field-like, recursive) around the limit cycle ofthe system?

If the answers to the questions posed in the previous paragraph is 'yes', thenthis would mean that the four metafunctions represent basins of attraction towhich different functional couplings of articulators converge in the overall articulatory act. I am making this suggestion on the basis of evidence provided byFowler and Rosenblum (1991: 48-9) that listeners 'are sensitive to the distinctgestural sources' that affect fundamental frequency in different (metafunctionally motivated or entrained) ways.

Consider the following research findings that are discussed by Fowler andRosenblum in support of the above cited claim:

What guides the listener's factoring of converging effects of Fa [fundamentalfrequency, ~JT]? Presumably, it is the configuration of acoustic products ofthe several gestures that have effects, among others, on Fa. Intonationalpeaks are local changes in an FO contour that are effected by means that, toa first approximation, only affect Fa; they are produced, largely, by contraction and relaxation of muscles that stretch or shorten the vocal folds ... Incontrast, declination is a global change in FO that, excepting the initial peakin a sentence, tracks the decline in subglottal pressure ... Subglottal pressureaffects not only FO, but amplitude as well, and several researchers havenoticed that amplitude declines in parallel with Fa, and resets when FO resetsat major syntactic boundaries (e.g. Breckenridge, 1977; Maeda, 1976). Theparallel decline in amplitude and Fa constitutes information that pinpointsthe mechanism behind the Fa decline - gradual lung deflation, incompletelyoffset by expiratory-muscle activity. That mechanism is distinct from themechanism by which intonational peaks are produced. Evidence that listenerspull apart the two effects of Fa (Pierrehumbert, 1979; Silverman, 1987)suggests that they are sensitive to the distinct gestural sources of these effectson Fa. (Fowler and Rosenblum 1991: 48-9)

The 'distinct gestural sources' of these various effects on Fa echoes the distinctgestural effects of the relatively discrete functional couplings of articulators that


give rise to vowels and consonants in eo-articulation. Thus, vowels are producedby slow movements that are finalized towards global configurations of the vocaltract; consonants are movements on a faster timescale that effect local restrictions of the vocal tract. It is the interplay of these two heterogeneous timescaleswhich allows for the phenomenon of eo-articulation, Rather than a corrupted ordegenerate encoding of static and abstract phonetic features, eo-articulationshows that the topological state space of the vocal tract is, in eo-articulation,divided into what Kauffman calls 'disjoint basins of attraction' (1993: 176-7). Inthis way, the greater number of degrees of freedom of the vowel-state functionalcouplings of articulators need be harnessed only initially such that only a smallernumber of parameters need be harnessed subsequently. The more global,slower-scale movements required by vowels constitute the initial state of atrajectory to which the smaller, faster-scale movements of consonants flow.

The distinct gestural sources of intonational peaks and declination of vowelsand consonants suggest that the vocal tract is a topological substrate (c.f.Hjelmslev's expression purport) which is, in the space-time of articulation,organized in terms of complex interactions of attractors which lie on differenttrajectories. These attractors are the diverse principles of metafunctional organization to which the distinct gestural sources are attracted in the unfoldingspeech event. This would suggest that over the time-span of, say, a single breathgroup, different metafunctional parameters are held constant on sometimescale(s) of gestural activity. These parameters are the values which organize,for example, the various effects on FO into metafunctionally distinct regions ofgestural activity. For example, the local effects of the contraction and relaxationof muscles in the stretching and shortening of the vocal folds on intonationalpeak indexes feeling states of the speaker; the gradual lung deflation which isassociated with decline in both amplitude and FO indexes the rhythmic periodicity associated with breath-group boundaries; the differing timescales requiredfor the global versus local restrictions of the vocal tract may be associated withthe categorical distinction between vowels and consonants as the parameterspace is divided or 'analyzed' (Hjelmslev) into discrete regions corresponding tophoneme categories. These three examples would, on a first approximation,seem to suggest some of the ways in which listeners may distinguish the distinctgestural sources of various metafunctional values in the expression plane ofspeech. Thus, the first example relates to the expression stratum analogue ofinterpersonal meaning; the second to textual meaning; and the third to experiential meaning.

We have here a concrete illustration of the functionally diverse ways in whichmeaning interacts with and entrains the biological dynamics of the individual'srelations to its environment. Further, feelings and cognitions are not separate orautonomous functions. Rather, the metafunctionally diverse-yet-unified natureof meaning-making shows that semiosis - both 'inner' and'outer' - cross-couplesthe individual body-brain complex to its environment along a number of distinctdimensions simultaneously.


17. The Metafunctional Basis of Space and of Bodily Movement inEcosocial Space-time

All gestural activity unfolds in space. The notions of space and of spatial relationsare crucial for our understanding of the ways in which the body relates to,orients to, and is integrated to its ecosocial environment. I believe that this is soboth of gestural activity which is executed, as well as 'inner' gestural activity, ofwhich inner speech is but one specific case. All modalities of human socialmeaning-making utilize, or in some way depend on, the movement - simulatedor real - of bodily articulators. Thus, the receptors which are implicated ingestural activity coincide with those which take part in our sense of movement.These are (1) the visual receptors; (2) the vestibular receptors in the inner ear;(3) the muscular and the muscular-articulatory receptors, both of whichmeasure movement of the various parts of the body; and (4) the cutaneousreceptors of the skin which measure changes in pressure or temperature causedby contact either with the parts of the body or with phenomena in the (extracorporeal) external world (see Berthoz 1997: 33).

Space is, then, a fundamental organizing principle which may well underpinthe metafunctional basis of semiosis itself. There are typological-categoricalspatial categories (e.g. in front, behind, up, down, and so on) which enable usto perceive as invariant certain relations between objects and the parts of ourbody. There are also topological-continuous relations such as spatial distancefrom the body, which are quantitative and variable rather than invariant andcategorical. Kosslyn et al. (1995) propose that the two kinds of relations arecoded differently in the neural architecture of the brain. But there is no director causal relation between these lower-scalar neural arrangements and higherscalar ecosocial semiotic ones. We can see in these two different kinds ofrelations a possible lower-level integrative constraint on the ways in whichsemiotic modalities are themselves organized. Categories such as 'in front','behind', and so on are digitalized units of information which specify the givenobject system relative to an observer perspective. In terms of the body's spatialrelations to its environment, it defines the informational co-ordinates of thesystem appropriate to the scale of its interactions with its environment. Thus,the conceptual categorization of these spatial relations takes place on the levelof the sensori-motor samplings the body makes of its surroundings.

On the other hand, topological-continuous factors such as spatial distanceraise two orders of questions. First, spatial distance raises the question as towhich resources are best harnessed in meaning-making activity so as to overcomethe problems - perceptual and physical - that distance poses for humaninteraction. Both visual and auditory resources for interaction are subject toconstraints imposed by the material environment on their efficacy (see alsoStuddert-Kennedy and Lane 1980: 37). For example, visual gestures requireclose face-to-face contact between interactants as well as an optimal level ofambient light; acoustic signals decay with increasing distance. Secondly, spatialdistance between bodies functions as a lower-level integrative constraint onsocial and interpersonal relations (cf. closeness or intimacy versus distanceand remoteness). We may see here an analogue with the metafunctional


organization of semiotic modalities. Space and its metaphorical transformationswould seem to be a fundamental constraint on human social meaning-making.

What, then, is the relationship with the metafunctions? What relations in theecosocial environment and therefore at a higher-scalar level than the individualbody-brain favoured the further channelling and development of bodilymovements and gestural activity as modalities of social meaning-making? Itwould seem to me that the neural architecture referred to above constituteslower-level initiating conditions on the further transformations referred to above(see Salthe 1993: 216). With the emergence of higher-level ecosocial constraints,these possibilities for relating body and environment - self and non-self - incorporated more and more information of a social nature from the higher levels ofecosocial organization. This increasing specification and elaboration of the newemergent possibilities of the system for interacting with its con-specifics meantthat determinate semiotic modalities with a full-fledged metafunctional basisarose.

Researchers such as Griisser (1991) and Berthoz (1997: 108-9) havesuggested that the space in which the body moves is articulated into a number ofzones or subspaces in relation to the body. There is, according to theseresearchers, a personal space, an extra-personal space, and a distant space, to usethe terminology proposed by Griisser. Furthermore, these distinctions appear tohave a neural basis (Berthoz 1997: 108). Personal space, which is located withinthe limits of one's own body and perceived by the internal senses, is the space ofthe self. However, we can also perceive our own bodies as an external object as,for instance, when I look at my own body, or some part of it, as another externalobject in the world.

The space of grasping - both oral and manual - means that the perception ofone's body may be extended by means of a tool or implement. Berthoz (1997:108) points out that in such cases the body is in effect extended such that thebody is integrated with the physical object which it grasps or with which it comesinto contact. An example would be the use of a pen, crayon, or other writing toolfor the purposes of tracing on a treated surface. In this way, the body is extendedby the tracing implement so that the body's contact with the treated surface paper, ete. - is felt at the point of contact - the tip - between pen, say, andsurface. This shows how haptic sensations are fundamental in the integration ofthe body and its extra-somatic extensions with physical-material objects andprocesses which are external to the body. Importantly, Berthoz (1997: 109) drawsattention to the fact that the tool so used is not perceived at the point of contactwith the body but at the place it occupies in extra-corporeal space. The bodybrain complex thus constructs or projects a spatial extension of itself.

Our perception of space is organized in two main ways. The first - calledegocentric - is in relation to one's own body. That is, one's own body is thereference point in relation to which objects and their relations in the externalworld are observed. The second - called allocentric - uses the relations betweenobjects or in relation to some reference point which is external to one's ownbody. This second way of perceiving space, which appears to be confined tohumans and other primates, means that relations between objects can be constructed and perceived without recourse to one's own body as a permanent


reference point. In this way, allocentric perception is detached from themovements of my own body and allows for the internal mental simulation ofspatial displacements which occur independently of my own body. The distinction between egocentric and allocentric perceptions of space may, in my view,underpin the elementary indexical distinction between 'here' (cf. egocentric)and 'there' (cf. allocentric). This may be compared to the distinction Halliday(1992a: 20-1) makes between the 'two primary modes of experience', viz. 'outthere' and 'in here' in relation to which consciousness itself is organized. In earlyinfant semiosis, this elementary semiotic distinction appears to be a precursor ofthe textual metafunction and its further elaborations and expansions. In thiscase, the basic distinction may be something like that between 'proximate' and'distal' spatial relations. However, Halliday's distinction is also experientialized asthe first, basic distinction between 'material' processes in the external world and'conscious' processes that occur within the self. Interpersonally, this same distinction constitutes the basis for enacting interactions between an emergent'self and 'non-self.

To date, most of the effort has gone into the study of the signified or contentstratum - e.g. the lexicogrammar and semantics oflanguage - of social meaningmaking. Further, phonology and the much less practised science of graphologyhave tended to remain separate subdisciplines which have not been integratedinto a truly ecosocial theory of the body and its role in social meaning-making.Both phonology and graphology have categorical-typological resources such asphonemes and graphemes; they also have field-like prosodies which are stronglyinterpersonal-orientational, having to do with affective investment, emotionalcommitment, ete.; and they also have wave-like or periodic beginning-middleend type structures, boundaries between units, and so on. All of these interfacewith and construe bodily processes and/or extra-somatic extensions of these.They also project these into the ecosocial environment of the interaction,contributing to the construction of discursive space-time and its felt movementand rhythms.

3 Body Dynamics, Meaning-making, andScale Heterogeneity: Expression andContent as Cross-scalar SemioticProcesses Embedding the Body-brainin its Ecosocial Environment

Language, like personality, is a binder of time, of the past and future in 'thepresent'.

J. R. Firth, 'The semantics oflinguistic science' (1957a [1948]: 142)

1. The Dynamical Character of Expression and Content and the Crosscoupling of Diverse Scalar Levels of Semiotic Organization

Action and meaning are emergent properties of the time-dependent crosscoupling of a number of different systems on diverse space-time scales. Meaningemerges in and through the interaction of diverse semiotic modalities and thephysical-material world. In the real-time of the unfolding activity, varioussemiotic modalities: (l) selectively map salient features of the material world totheir own activity; (2) at the same time, they map selected features of othersemiotic modalities to their own dynamical processes; and (3) they map selectedfeatures of participants' perceptual-motor activities to their own activity. It is thedynamical cross-coupling in time of a number of these heterogeneous systemsthat produces the meaning-making event. These systems include the varioussemiotic resources that are deployed; objects, events, and so on, in the materialworld; and the perceptual-motor activities of participants. Action emerges in andthrough the cross-coupling and interaction of diverse semiotic-discursive andphysical-material systems. Genre-specific meaning-making activities occurbecause the cross-coupling dynamics of all these systems create, in time, aninternalized attractor space. It is in this space that participants' experiences priorto the here-and-now event, the previous stages of the same unfolding event, andhere-and-now responses to perceptual stimuli at any given moment all act insynergy to produce genre-specific semiotic performances and their resultingobject-texts.

Meaning-making is, then, a dynamic process which is determined by anumber of perspectives, all of which interact together to produce the occasionspecific activity. First, there is the perspective of what it is possible to mean interms of the intrinsic characteristics and dynamics of any given semiotic resource

BODY DYNAMICS, MEANING-MAKING, ... 109

system. Secondly, there is the sequential unfolding of events in real-time, thelogogenetic selecting and deployment of semiotic options and their crosscoupling with specific external events, as well as changes in cross-couplingstrengths in moments of transition from one phase of the activity to another.Thirdly, there are the sensori-motor reactions of participants to what they see,hear, feel, touch, grasp, point to, and so on, in their spatio-temporal purview atany given moment.

With these general considerations in mind, this chapter will focus on thequestion of how the organization of language as expression and contentfunctions to embed the body-brain in its ecosocial and bodily environments. Aswe shall see below, this requires us to examine the diverse timescales and theircross-couplings that are implicated in the intrinsic organization of bothexpression and content.

2. The Expression Stratum and the Principle of Alternation

Lernke's (1999) Principle of Alternation is a useful tool for theorizing, in termsof the three-level scalar hierarchy, both how semiotic functions are mapped ontodynamical scalar levels, and the reorganization of continuous variation intodiscrete variants. In the act of speaking, for instance, continuous quantitativevariation in sound is related to continuous quantitative variation in the sensorimotor activities which produce speech sounds in the process of articulation.However, the processes of both producing and comprehending speech soundsrequire that speakers and listeners are able to articulate and perceive thosefeatures which are salient or criterial for determining which words are utteredand/or heard on a given occasion. In other words, the continuous topologicalvariety of the sounds produced at level L-I of the perceptual-motor activitiesinvolved in the production and perceptual pick-up of speech sounds is reorganized and interpreted as discrete phonological categories, which are typologicalin character, at level L+I. In the case of speech sounds, this level is the phonological system of a given language.

In Table 3.1, I suggest some ways in which the expression stratum of spokenlanguage can be reconceptualized in terms of the Principle of Alternation.

In accordance with this Principle, the emergence of a ranked scale of phonological units and their structural-functional relations in a given language constitutes a new intermediate level N which functions semiotically to reorganize thecontinuous quantitative (topological) variety of units and interactions at thelevel (L-I) of sensori-motor activity in articulation as discrete, typologicalcategorial phonological units and relations for level (L+I). Level L-I interactions are the material, topological ground of the typological distinctions whichemerge at higher levels. Level L+1 comprises the system of phonologically salientdistinctions which function in a given language to symbolically construe differences in lexicogrammatical form on the content stratum.

These distinctions develop and are conserved in the long-term memory ofindividuals as a system of phonological values or differences on the basis of thematerial and semiotic interactions of individuals with their ecosocial environ-


Table 3.1: The expression stratum of spoken language in relation to the three-levelhierarchy

Scalar level Difference

L+l System of Discrete Symbolic Phonologicallyphonological typological- salientdifferences categorial distinctionsas SI distinctions; symbolically

difference of construetype lexicogrammatical

forms oncontent stratum

L Phonological Indexical Contrasts inrank scale rhythm,comprising prosody, intonationhierarchy of perform indexicalranked units: groundingtone group; functionsfoot; which givesyllable; phonologicalphoneme categories a point

of reference in thehere-now intra-andinter-bodydynamicsof the I-you andother aspects ofthe here-nowspeech event

L-l Continuous Topological- Iconic Continuousquantitative continuous quantitativevariation in variation; variation ofsound difference of body dynamicsmapped onto degreecontinuousquantitativevariation insensori-motoractivity inarticulation

BODY DYNAMICS, MEANING-MAKING, ... III

ment along their historical-biographical trajectory. In this way, the phonologicalpossibilities for meaning-making are developed in individuals at the same timeas they are transmitted to other individuals in and through ontogenetic andother processes of recruitment to and apprenticeship in the practices of thesocial group (Vihman et al. 1985; Vihman and de Boysson-Bardies 1994). Thephonologically salient distinctions made on the expression stratum are notrestricted to the organismic scale of the individual; they also link with other,larger-scale dimensions well beyond the here-now scale of the organism'smaterial interactions with its immediate environment. Semiotically salientdistinctions made on the expression stratum interface with kinetic (bodily)processes on level L-l. In this way, bodily dynamics are strongly cross-coupled todynamical processes existing on potentially many different space-time scales.

The phonological system of a given language is a system of interpretancewhereby the sounds we produce and hear are classified in terms of a relativelysmall number of discrete and contrasting equivalence class items. These classitems are, for example, the phoneme type-categories of a given phonologicalsystem. In this case, level L is that of the phonological forms, usually describedas a set of hierarchically related units in a phonological rank scale (see below).The L level has the function of filtering phonologically non-criterial characteristics of speech sounds. Non-criterial characteristics include aspects of the soundwhich relate to the subjective state of the speaker, the specific circumstances inwhich the sound was uttered and/or heard, the differences between thespeaking voices of individuals, and so on. All of these aspects are non-criterial forthe assignment of the sound to a given phonological category. Thus, the phonological contents of symbolic neural space constrain and entrain both vocal-tractgestural activity and our perception of the resulting sounds in terms of a systemof phonological distinctions (cf. values) on the higher-scalar level of the SI whichis in operation.

Thus, physical-acoustic properties of the sound such as variation in fundamental pitch, which is iconically related to the rate of vibration of the vocalchords at the time of speaking, may not be in any way salient or criterial for thecategorization of a given sound as belonging to this or that phonological equivalence class (see below). Such non-criterial acoustic-physical properties and theirrelated articulatory processes occur on a much smaller, faster timescale than dohigher-scalar phonological units. The learning, producing, and perceiving of thelatter by the central and peripheral nervous systems take place over a muchlonger timescale. Moreover, a system of phonological units and their relationsrequires a far greater number of interconnected neuronal networks in order tomaintain it in long-term memory, as well as to activate it in the production andperception of speech sounds.

Level L can be modelled as a hierarchy of vocal-tract trajectories that entrainlevel L-I dynamics so as to give rise to three different kinds of phonologicalunits. The notion of trajectory serves to remind us that we are not talking aboutstatic structural units and relations, but about time-bound trajectories ondifferent scalar levels that are identified with the phonological rank scale,comprising, in the case of English, syllable, foot, and tone group. Importantly,each of these phonological ranks and the units specific to that rank are seen as


dynamic, time-bound processes of cyclic or periodic vocal-tract behaviour withtheir own specific rates of completion. Thus, smaller-scale trajectories operate onfaster timescales than do higher-scalar ones. Furthermore, the notion of a vocaltract trajectory refers to the temporal cohesion of a given kind of phonologicalentity (e.g. syllable, foot, tone group) from the perspective of a cogent momentwhich spans the entire (temporal) existence of the trajectory. Syllable, foot, andtone group refer to such cogent moments of different cycles of vocal-tract activityon their respective scales. Each of these cogent moments specifies the beginningof the trajectory in lower-scalar neuromuscular activity on level L-I and terminateswith the completion of a cycle of activity which results in the entraining of vocaltract activity to a recognizable phonological entity in the given language.

The neuromuscular substrate on level L-l is the dissipative structure which isthe momentary material embodiment of the trajectory (Salthe 1993: 260).However, it is the vocal-tract trajectory on the phonological level L, rather thanon the material level (L-l) of its embodiment, which semiotically connects us tothe trajectories of others, intertwining and overlapping with them in theprocesses of logogenesis. The three phonological ranks of syllable, foot, andtone group exhibit this property of temporal existence. On the other hand, thephoneme is a discrete packet of paradigmatic features which serves to specify aparticular point in the unfolding phase space of a trajectory by categorizing thatpoint as the instantiation of that particular packet of paradigmatic features, i.e.as an instance of this or that phoneme type-category. The phoneme is, asCleirigh (1998: 41) argues, an expression of the probability of the speaker orlistener locating that particular packet of features at that point.

Cleirigh (1998: 38-9) has defined the three phonological ranking unitsnamed above as periodic cycles of vocal-tract activity in ways that fit the abovedefinition of trajectory. Thus, the first (lowest) scalar cycle is defined as a cycleof opening and closing the vocal tract 'to dampen phonation in the productionof syllables' (Cleirigh 1998: 38). The syllabic trajectory consists in the temporaloscillation between the assignment of heavy and light syllabic weight. This givesrise to a structure consisting of Onset and Rhyme. The units comprising theRhyme are potentially moraic, i.e. they have syllabic weight (moraicity) while theOnset consists of units that do not have syllabic weight. The latter are, therefore,non-moraic. In the discussion which follows, the clause complex I bought this VJand I sailed it will serve as the focus of our illustrative discussion. With referenceto the syllable sail, Figure 3.1 shows the assignment of syllabic weight. Thus, theconstituent comprising the Onset is non-moraic, while the constituentcomprising the Rhyme is potentially moraic and can therefore function as atiming unit on the next-highest level, viz. the foot. Figure 3.1 shows the alternation of non-moraic and moraic elements in the unfolding temporal trajectory ofthe syllable sail from the word sailed in our example.

The next-highest scalar level is the trajectory of vocal-tract activity thatproduces rhythm. The rhythmic trajectory is formed, in stress-timed languageslike English, by the alternation of salient and non-salient (weak) syllables. Salientsyllables are so defined by virtue of the fact that they are assigned stress throughvarying topological factors such as loudness and duration. The temporaltrajectory which arises on this level is known as the foot. Thus, a foot is a

Figure 3.1: Syllable structure of'sail', showing morale and non-moraic elements

trajectory of vocal-tract activity which, in English, results from the periodic(cyclic) assignment of salient and weak stress to syllables such that each cycle -each foot - consists of one stressed syllable and an unequal number ofunstressed ones.

M / bought this / Vee Jay / A and / sailed it

In the above example, the feet are marked off by slashes; the syllable immedi-ately following the slash is the salient syllable, or the Ictus. The carat sign 'A >

indicates a silent Ictus. Figure 3.2 shows the wave trajectory of the alternation ofstressed and unstressed syllables as the syntagm unfolds in time.

The highest scalar trajectory that I shall consider here is that which isconcerned with the creation of tonic prominence or tonicity through changes inmajor pitch. This trajectory is concerned with intonation. The intonationtrajectory consists of a cycle which alternates between 'switching tonicity on andoff, as Cleirigh (1998: 39) expresses it. In our example, the intonation cycle isinstantiated as follows:

//i I bought this / Veejay //i and / sailed it //

The wave trajectory of tonicity in this example is shown in Figure 3.3. In this way,we see phonology as consisting of time-bound trajectories on different scalarlevels. Each scale has its own characteristic temporal cycle or rate of completionand its own characteristic entities. Furthermore, the trajectories on their

BODY DYNAMICS, MEANING-MAKING, . . . 113


Figure 3.2: Trajectory of wave of stressed and unstressed syallables in clause complex,showing alternation of stressed and unstressed syallables as syntagm unfolds in time

Figure 3.3: Trajectory oftonicity wave in clause complex

respective scalar levels are integrated to each other in the production of speechsounds. This integration of diverse scalar levels leads to the production of anincreasingly complex speech signal in which the higher level of generality char-acteristic of moraicity is reintegrated to the increasingly more specified levels ofstress and tonicity. This process of integration across scalar levels increases theamount of information that is stored in the trajectory and hence the agentivedeterminability of the trajectory (Salthe 1993: 261). It is not difficult to see herehow the further integration of expression-stratum trajectories to content-stratumones amounts to a further increase in the semiotic complexity of the trajectoryand therefore of the amount of information-meaning that is stored along theduration of the trajectory. It is the content stratum which accesses storedmeaning on the ecosocial scale and which entrains lower-scalar expression-stratum trajectories to its own higher-scalar dynamics.

The emergence of the phonological rank scale on level L therefore providesstable parameters such that particular vocal-tract trajectories qua phonologicaltypes and their structures persist on many different space-time scales and acrossmany different users (speakers and listeners). The cyclic character of each


trajectory, which Cleirigh has identified, draws attention, in my view, to the selforganizing properties of the neuromuscular substrate. This is possible becauseeach scalar level has stored information that is specific to its level as well as acharacteristic cycle of activity which defines its cogent moment. For example, thesyllabic trajectory both stores information about syllable weight and, at the sametime, has a temporal existence definable as states in its overall phase space, viz.opening --+ closing the vocal tract. Furthermore, we shall see below how a smallnumber of principles of formal organization are fractally distributed across trajectories of both expression- and content-stratum activity.

The properties of the filtering or buffering system - i.e. the hierarchy ofranking phonological units at level L - is comprised of units which may havesymbolic and/or indexical properties and functions on the expression stratumof speech. On level L+1, a given phonological SI filters and, therefore, interpretsthe topological-continuous variation of speech sounds as phonologically salientindexical or symbolic differences. Differences of this kind are typologicalcategorial in nature. Indexical functions are signalled by salient contrasts in, forexample, intonation, speech rhythm, and other prosodic dimensions of thesound stream. These indexical features of the expression stratum are necessarilygrounded in the here-now scale of the speech event. They function to draw thelistener's attention to some salient aspect of that event either by presupposing itsexistence in the speech event or by creatively bringing it into existence throughthe use of the particular indexical sign. (The distinction between the indexicallypresupposing and indexically creative functions of indexical signs in languagecan be found in Silverstein 1976.)

In the case of indexicals, sensori-motor activity functions to index somesemiotically salient dimension of the ongoing speech event by virtue of the perceivable relation of co-occurrence between the indexical sign - intonational orrhythmic contrast - and the indexed object. For example, phonological systemsand structures may index a given lexicogrammatical unit on the content stratumby means of prosodic resources which extend over a given lexicogrammaticaldomain. Alternatively, a phonological syntagm may demarcate a lexicogrammatical unit, for example when syllables indicate morpheme and word boundaries(Cleirigh 1998: 24-5; Firth 1957b [1948]: 122-3). Furthermore, phonologicalunits such as the prosodic spread or extension of tone over a given lexicogrammatical unit may index that unit as having informational salience in theunfolding discourse. Rhythmic organization on the lower rank of the foot canalso be used to render specific lexical items prominent as well as to demarcatelexicogrammatical boundaries.

Phonologically indexed aspects of lexicogrammatical and discourse organization also include the 'communicative intentions' of speakers. For example,Gumperz, in Prevignano and Di Luzio (2003: 10-14), argues that communicativeintentions are inferable on the basis of indexical signs or contextualization cuesof this sort. The intentions so inferred do not, in my view, belong to the domainof internalized cognitive processes per se. I do not assume that the index is theexternal sign of an internal reality beneath the skin of the individual. Instead, Iwould say that this indexing of communicative intentions is just one dimensionof the ways in which meaning flows into and entrains a particular logogenetic


trajectory on the here-now scale in which the event is enacted and unfolds for itsparticipants. Thus, Peirce's understanding of indexical signs is upheld here.That is, indexical signs are sensori-motor activities that point to the existence ofsome 'object' on the here-now scale. Such objects may be grounded by the indexeither 'subjectively' or 'objectively' - as existing in the field of the speaker's body,including the inner domain of thought and feeling; or as existing in the field ofthat which is external to the speaker's body, Le. in the external objects and eventswhich are part of the context of situation. The indexing of lexicogrammaticaldomains by phonological units such as rhythm and tone already shows theprinciple of scalar heterogeneity at work. This is so in the sense that lower-scalarphonological systems and structures on the expression stratum integrate withthe higher-scalar lexicogrammatical and discourse levels of organization on thecontent stratum.

The use of prosodic resources such as rhythm and syllable stress may serve toindex semantic relationships which are not made explicit in the lexicogrammar.That is, the particular phonological feature may index a discourse-level semanticpattern which the listener is required to retrieve from his or her knowledge ofspecific intertextual patterns in order to understand the particular semanticrelation. Semantic meaning relations are made in discourse through particularpatterned combinations of lexicogrammatical selections in texts. There is noone-to-one correspondence between lexicogrammatical forms and the semanticmeanings they are used to make in discourse. Moreover, lexicogrammaticalforms do not always spell out the full semantic pattern in an explicit way.Prosodic and other phonological cues often provide subtle cues as to how toconstrue the relevant semantic pattern in discourse. This shows how expressionstratum features may index intertextually recoverable semantic patterns whichare necessary and relevant for the interpretation of the particular lexicogrammatical selections that are made. With this in mind, we can say that speakers'intentions are often shown to be analysable as an indexical relationship betweenprosodic or other cue and an implicit semantic pattern on the discourse levelrather than an internal cognitive state of the speaker per se.

3. Stratification in Relation to Expression and Content

The notion of the expression stratum that I am using here is a further development of Hjelmslev's (1961 [1943]: 47-60) theory of the stratal organization ofthe linguistic sign, seen as comprising the two strata of content and expression.Furthermore, both content and expression are internally stratified. Thus, theexpression stratum comprises the two levels of expression form (phonology) andexpression substance (phonetics). Likewise, the content stratum is internallystratified as content form (lexicogrammar) and content substance (discourse).The relationship between the strata of content and expression is one of realization. That is, content is realized by expression and expression realizes content.Realization is a two-way relationship between these two strata. It is a symbolicrelationship. That is, expression symbolizes content and content is symbolized byexpression.


In systemic-functional theory, the content stratum is internally stratifiedinto discourse and lexicogrammar (Martin 1992a: 14-21). Discourse, which islinguistically realized social action and text, is realized by co-patternings oflexicogrammatical selections. Discourse is a higher-level reconstrual oflexicogrammatical patterns so as to realize thematic patterns, interactive modes,and cohesive relations in texts. Following Halliday (1978a: 108-9), text is definedas either spoken or written instances of language-in-use. Discourse and lexicogrammar belong to the content stratum because they are based on varioustypes of semantic-pragmatic meaning and social (inter)action. I take a viewsimilar to that originally espoused by Firth (1957c [1951]: 220), viz. meaningcannot be defined in terms of any single stratum:

Language text must be attributed to participants in some context of situationin order that its modes of meaning may be stated as a series of levels, whichtaken together form a sort of linguistic spectrum. In this 'spectrum' themeaning of the whole event is dispersed and dealt with by a hierarchy oflinguistic techniques descending from social contextualization to phonology.(Firth 1957c [1951]: 220)

In some respects, the terms expression and content are unfortunate ones forthey may suggest the idea of a material means of expressing meaning, seen exclusively in terms of content. In my view, such a conception reflects a fundamentalmisunderstanding of Hjelmslev's notions (see chapter 2, section 1, pp. 59-67).The stratal organization oflanguage into expression and content reflects diverseprinciples of semiotic organization on different levels. Meaning is made throughthe contribution of all the levels involved in the logogenetic process ofproducing text or discourse. In the case oflanguage, this means that the phonological or graphological, lexicogrammatical, and discourse levels of organizationall simultaneously make their distinctive contributions to the overall meaningmaking event.

The expression stratum is based on sensori-motor activity. An example is thevocal-tract activity whereby speech sounds are articulated. Expression isinternally stratified in terms of phonetics and phonology. I shall use the termphonetics to designate language-specific classifications of articulatory repertoires independently of their linguistic function. Phonology, on the other hand,refers to the higher-level organization of such repertoires into language-specificfunctional systems and structures. As we shall see below, this means that phonological units and relations, in the logogenetic process of creating discourse ortext, are both integrated to and constrained by higher-order discourse andlexicogrammatical patterns of organization on the content stratum. Moreover,phonology is not simply a vehicle whereby such content is 'carried' or'conveyed'. Instead, it is a full-fledged level of linguistic organization whosefunctional structures and systems make their own specific contribution to themeaning and organization of the discourse event. This contribution will bediscussed in the following section in terms of the iconic, indexical and symbolicfunctions of phonological forms in the meaning-making process.

The phenomena of experience that are construed on the content stratum are


not perceived. In listening to speech sounds or visually scanning a sequence ofgraphemes on a treated surface, what is acoustically or visually perceived are notthe phenomena of experience that are construed by the content stratum. By thesame token, in listening to or in reading say a sequence of graphemes, we doexperience the phenomena that are construed by the lexicogrammatical- anddiscourse-level units and relations. However, such phenomena, unlikeexpression-stratum units and relations, are not present in the immediate environment as acoustic or light energy which structures the perceiver's point ofobservation. The acoustic or light energy which structures the perceiver's pointof observation plays a causal role in acting on the perceptual systems of theobserver. However, the flow of symbolic experience which is also apprehendedin the sense of what the discourse is about does not causally act upon theperceiver in the way acoustic and light energy does. In the process of articulation, for example, vocal-tract activity physically modifies the acoustic stream insuch a way that the sequence of perceived phonological invariants redounds witha concomitant flow of experienced phenomena such that, in addition to hearingthe sounds produced by the speaker, symbolically construed phenomena ofexperience are also apprehended. The stratification ofianguage into expressionand content means that, when one attends to the acoustic stream of speechsounds, one attends to two sets of informational-semiotic invariants: (I) a set ofphonological invariants that is specific to the stream of the speech sounds; and(2) a set of invariants that is specific to the semantic categories (persons, places,actions, events, and so on) that are construed by the lexicogrammatical anddiscourse units and relations. The first involves perceptual pick-up; the secondinvolves the resonance of what is picked up with symbolic categories ofexperience that cannot be perceived. Perceiving the acoustic stream duallyinvolves a symbolic experience of phenomena as a result of the ways in which theacoustic flux structures the perceiver's point of observation according to theprinciples outlined in section 2 so as to be recategorized as symbolic content.

4. The Integration of Iconic, Indexical, and Symbolic Modes of Meaning inPhonology

The expression stratum is itself a cross-scalar phenomenon which simultaneouslylinks multiple space-time scales. These diverse scales can be described in termsof a nested specification hierarchy which consists of iconic, indexical, andsymbolic modes of meaning-making.

4.1 The Iconic Mode ofPhonological OrganizationFollowing Salthe's (1993) notion of the specification hierarchy (see also Lemke1995b: 113, 2000a; Thibault 2000a: 301-3, 2003b), the iconic dimension ofspeech sounds is the least specified, innermost level of such a nested hierarchy.

Many phonologically non-salient features of the stream of speech sounds suchas the quality of the speaker's voice tell us that the speaker is sad, anxious, ill, andso on. In this case, we are dealing with continuous quantitative variation in thesound stream which directly - iconically - maps onto continuous quantitative


variation in the speaker's neurophysiological dynamics. That is, they are iconicsigns of body states and feelings. They can be interpreted as signs of affective andother subjective conditions of the speaker. These iconic signs are confined to theimmediate here-now scale. For this reason, they closely relate to the topologicalmaterial ground of our being in the world. However, even at this level, scalarheterogeneity is evident on account of the fact that these iconic signs extendbeyond the organismic level and are able to be interpreted as such by others.

4.2 The Indexical Mode ofPhonological OrganizationPhonological units and relations on level L also have indexical functions. Thismeans that the system of phonological possibilities - its categorial distinctions is also grounded in the here-now act of speaking and listening qua sensori-motoractivity. Indexical properties add further levels of complexity and specification,along with the new semiotic functions of these properties. The innermost iconiclevel is integrated to and modified by the newly emergent indexical level. Thisdoes not mean that the iconic level is transcended; rather, the indexical levelentails the emergence of typological differentiation from topological variationon the basis of the increasing semiotic differentiation of vocal-tract gesturalactivity. In this way, vocal-tract activity is used to intentionally 'point to'contextual values beyond the body-scale. Many 'non-linguistic' contrasts in thespeaking voice such as loudness may have such indexical properties andfunctions. Thus, contrasts in rhythm, intonation, and prosody can functionindexically to give phonological categories a point of reference in the act ofspeaking by specifying, for example, its interpersonal or textual significance.

For example, in English, the assignment of pitch prominence to a givensyllable in discourse is an independent variable which interacts with, though itdoes not define, the phonological type-category which the particular syllableinstantiates. Rather, the speaker's choice of pitch prominence indexicallygrounds the syllable in terms of factors such as its interpersonal-interactional orinformational relevance in the here-now of the speaking event. However, it is notsalient for the purposes of distinguishing, say, the phoneme categories /b/ and/p/ whereby the lexicogrammatical distinction between the words bat and pat issymbolized. Contrasts in rhythm, tone choice, pitch prominence, and so on, arepotentially indexical because they can point to or indicate contextually relevantvalues which are closely tied to the immediate here-now scale of the act ofspeaking. That is, they are interpretable as indexes of values which are recoverable from the perceptual purview shared by the participants in the event.

Phonological patterns also interact with lexicogrammatical patterns on thecontent stratum in ways that serve indexical functions of creating texture as wellas linking to aspects of organization on the higher-level lexicogrammaticalstratum. In this regard, Cleirigh (1998: 57) points out that systems of phonological cohesion may function to index 'the syntagmatic extent of lexicogrammatical units'. Following Firth (1957b [1948]), Cleirigh identifies three suchstrategies, demarcation, integration, and concatenation, as follows:

Demarcation is the phonological delimiting of a lexicogrammaticalboundary: initiality or finality. Integration is the phonological consolidation


of a lexicogrammatical domain as a single unit. Concatenation is the phonological annexation of two or more lexicogrammatical units within a larger(lexicogrammatical) structure. (Cleirigh 1998: 57)

All three strategies can be seen as having both intra- and extra-phonologicalindexical functions. For example, syllable boundary positions can create (1)endophoric ties that link either backwards or forwards to previous or followingsyllable peaks or (2) exophoric links to lexicogrammatical units that lie outsidephonology on the lexicogrammatical stratum. Thus, we can see here how thecreation of 'phonological texture' (Cleirigh 1998) functions to punctuate ordigitalize the topological flow of speech sounds into discrete events separated byboundaries. Such events have semiotic salience because they are interpretableaccording to social conventions which enable their interactional significance tobe recognized. The phonological resources referred to here, in punctuating thetemporal flow, constitute differences that make a difference to that flow. Theyare indexical resources which mark out a piece of the temporal flow of articulated sounds as a lexicogrammatical event. In a given discourse context, suchevents can be interpreted by the parties to the interaction as having a certainmeaning within the discourse. Indexical functions like these thus contribute tothe perception of the sound stream as an event, or a series of nested and interrelated events on various scales, rather than mere physical sensations (seeHande11989: 183).

4.3 The Symbolic Mode ofPhonological OrganizationThe symbolic dimension of the meaning of phonological forms is not the onlyfunction they have in the temporal unfolding of speech events. The symbolicpossibilities that such phonological distinctions have for construing or symbolizing lexicogrammatical distinctions on the content stratum means that thesounds produced in the act of speaking can be linked to the meaning-makingpossibilities of the lexicogrammatical and discourse levels of organization.Expression itself is linked to many different space-time scales that go beyond thehere-now scale in which speech sounds are articulated and heard.

The most specified, outermost level adds further layers of symboliccomplexity in the form of a full-fledged system of phonological contrasts whichcross-link to the content stratum of lexicogrammar and its semantics. Onceagain, the iconic and indexical levels are not transcended, but are integrated toand consequently modified by the symbolic level. This is evidenced by the factthat phonological units and relations have both indexical and symbolic functionsin discourse.

5. Rhythm and the Foot

The creation of a phonogenetic field in the dynamic processes of articulation isasymmetric in time (Cleirigh 1998: 53). For this reason, it is irreversible. In termsof biological initiating conditions emanating from level L-l, this is explainableby the fact that speech 'primarily exploits egressive airstreams' (Cleirigh 1998:


53). The initiating and terminating boundaries of a syllable on level L cannot bereversed because the principles which relate the boundary conditions of thesyllable to its internal organization are irreversible. The initiation of a syllablecannot be reversed to become the terminating position for each has consequences for the instantiation of paradigmatic features in the asymmetrictemporal flow of articulation (Cleirigh 1998: 53). At the same time, the directionality of the phonogenetic field of articulation is constrained by boundaryconditions operating on level L+1 of some phonological SI. Vocal-tract activity isintegrated to and entrained by this level such that particular phonological unitsand relations are recognizable.

Semiotically, the field perspective refers to the phase space, the organizedfield, which affords the emergence of the wave and particle perspectives. Wavesand particles are organizations of quantum field processes (Bickard andCampbell 2000: 331). The vocal tract is a topologically organized continuum inwhich vocal-tract gestural activity takes place. It is the material ground in whichthis activity takes place. As such, the topological-continuous space of the vocaltract is a phase space or an organized field in which certain kinds of organizedprocess perturb this field so as to give rise to various principles of organization.These diverse principles interact during the temporal unfolding of vocal-tractarticulatory activity.

Meaning-making is a probabilistic system for construing and acting on thephenomena of experience. Cleirigh (1998: 41) has shown how the Copenhageninterpretation of quantum mechanics can be applied to phonology, interpretedas a wave/particle complementarity in terms of probability. The application ofthis interpretation to phonology 'means interpreting waves of moraicity, stressand tonicity as measures of the probability of the observer - speaker or hearer finding the respective syllable, foot or tone group particle at that particularpoint' (Cleirigh 1998: 41).

For example, the syllables comprising the Ictus and Remiss phases of the footare comprised of clusters of vocoids and contoids. Abercrombie (1967: 85)points out that vocoids and contoids are phonetic segments which occur inparticular articulatory environments. These entail a good deal of phoneticvariation that is not phonologically salient and which is consequently filtered atthe phonological level L+1 where phonetic segments are construed as instantiations of particular phonemes. Such bundles of features correspond to thephoneme type-categories of a given phonological system. Phonemes specifyphonological type-classes of consonants and vowels which are specific to a givenlanguage system. Phonemes categorize discrete regions of the articulatory flux asparticles. A particle corresponds to a point in the unfolding wave to which aspecific packet of paradigmatic features is assigned such that the region of thewave movement corresponding to the particle is construed as an instance of aparticular phoneme type-category. Thus, phonemes paradigmatically specify agiven region of the flow of articulation - i.e. the muscular movements involved- as invariant irrespective of variations in pitch, loudness, and duration (Gibson1983 [1966]: 93). As Gibson points out, 'the stimulus information for detectingthem [phonemes, P]T] is invariant under transformations of frequency,intensity, and time' (1966: 93). Phonemes therefore construe articulatory


processes corresponding to paradigmatic bundles of distinctive features whichare perceived as invariant under change.

An analogous and parallel process occurs on the content stratum. Thus, experiential particles such as processes and participants in the clause, or Thing andEvent in the nominal and verbal groups, respectively, symbolically re-construephonological particles on various scalar levels as discrete experiential semanticcategories such as Actor, Material Process, Goal, and so on. These experientialparticles are invariant under topological transformations of the stimulus flux. Inturn, these particles of content construe the continuous flux of the phenomenaof experience as instantiations of paradigmatic categories of content.

In the following section, I shall focus on the metafunctional organization ofthe foot in English phonology with a view to investigating how the principlesoutlined above can be applied to the metafunctional organization of phonological form.

6. The Metafunctional Organization of Phonology as seen from thePerspective of Rhythm and the Foot

The syntagmatic structure of the foot is comprised of the two elements Ictus andRemiss. These two elements constitute two phases in the rhythmic elementcalled the foot in the phonological rank scale. Ictus and Remiss are themselvescomprised of syllables deriving from the next-lower rank in the phonologicalrank scale. The foot hierarchically dominates a given number of syllables in asequence and imposes varying degrees of stress on these syllables, 'according totheir location in the foot' (Catford 1998: 345). Syllables produced with greatestintensity and duration are described as [salient]; syllables with less intensity andduration are described as [weak].

The foot is the locus of rhythmic organization in the phonology of English.Rhythm can be viewed from four metafunctional perspectives, all of whichconstitute simultaneous and overlapping dimensions of the organization andmeaning of the foot. The four different modes of phonological realizationiconically map onto the four different ways in which the expression stratumentrains and construes vocal-tract dynamics in the act of speaking. The relevantenvironment of the expression stratum is the body. The iconic dimension ofphonology is the least specified, innermost level of a nested hierarchy of iconic,indexical, and symbolic modes of entraining and construing vocal-tractdynamics.

6.1 The Textual MetafunctiunalTextual structure is periodic; it is not based on constituent structure. Periodicityis a dynamic, time-bound process. Rhythm can be viewed as a perturbation in theform of a wave-front which is propagated through an articulatory field. Theperiodic nature of textual meaning shows rhythm as a periodic sequencing ofstrong and weaker elements, seen as alternating phases of Ictus and Remiss. Ictusand Remiss are alternating phases in the dynamic unfolding in time of arhythmic field (Cleirigh 1998: 80). English is a language of the kind identified


by Abercrombie (1967: 97) as having stress-timed rhythm in contrast to syllable-timed languages such as Italian. In the latter, periodic rhythmic movement isproduced by the syllable-producing process: 'the chest pulses, and hence thesyllables, occur at regular intervals' (Abercrombie 1967: 97). In languageshaving stress-timed rhythm, periodic rhythmic movement is produced by theassignment of stress. Figure 3.4 shows the organization of the rhythmic field ofthe two-syllable word sailed in terms of the Ictus and Remiss phases.

The foot is the parcelling-up of stronger and weaker syllabic elements into ahigher-order unit. The periodic, wave-like perspective allows rhythm to be seenas a dynamic, undulating movement of a particle from a strongly stressedelement to a more weakly stressed one (Cleirigh 1998: 80). These elements arethe boundaries of the temporal whole which is constituted by the foot. In thisway, rhythm structures the temporal dimension of events. The foot may be seenas the phonological 'atom' out of which larger-scale rhythmic structures areassembled on the expression stratum of speaking. From the perspective of thetextual metafunction, as Cleirigh observes, rhythm is 'a process of periodicallyswitching on stress' (1998: 81). In this sense, rhythm is viewed as a perturbationof a rhythmic field which gives rise to alternating Ictus and Remiss phases(Cleirigh 1998: 344).

6.2 The Logical MetafunctionRhythm, Cleirigh (1998: 345-6) further points out, can be viewed from the pointof view of the logical metafunction as a form of hypotactic interdependency inwhich the Remiss is dependent on the Ictus. The Ictus is the dominant or aphase of the foot; the Remiss is the dependent or (3 phase. Rhythm, as Cleirigh(1998: 81-2) has pointed out, can therefore be viewed as modification. That is,rhythm is 'the modification of articulation by initiator power' (Cleirigh 1998:81). Stronger or weaker stressed syllables impact on articulation in ways whichaffect factors such as aspiration, posture and degree of closure (Cleirigh 1998:81). In this perspective, we are concerned with the ways in which (articulatory)

Figure 3.4: Rhythmic field of the word sailed, showing Ictus and Remiss phases of itstemporal trajectory


phenomena relate to each other, rather than with how the different partsconstitute some larger whole. For example, we are not concerned with the wayin which Ictus and Remiss together make up a larger whole - the foot - but withthe relationship of interdependency between these two units. Dependencyrelations are relations between units such that a Remiss can be dependent on anIctus and hence can be modified by it, as shown below in relation to the wordnever in Figure 3.5.

6.3 The Interpersonal MetafunctionInterpersonally, rhythm can be seen as the creation of a charged vibratory fieldwhich becomes operational when a listener is within range of this field and ableto perceptually pick up its effects. Abercrombie (1967: 97) has referred to theexperience of speech rhythm as a 'rhythm of movement' which both speaker andlistener feel and which affords what Abercrombie has called 'phonetic empathy'.The 'phonetic empathy' referred to by Abercrombie is founded on the jointperception by speaker and listener of the periodic movement of the muscles ofthe breathing apparatus. Thus, rhythm can give rise to a jointly perceivedinterpersonal charge which has a range of potential emotional-affective valuesdepending on articulatory variables such as rate of tempo and degree ofsyncopation (Handel 1989: 458). The point I wish to emphasize here is that inter-actants may physically feel the bodily enactment of interpersonal meanings whichmay not necessarily correlate with any specific item on the content stratum (seeCowley 1998: 565-6 for a similar observation regarding pitch fluctuation).

6.4 The Experiential MetafunctionExperiential structure is particulate. In this perspective, the foot is a particleinterpreted as the peak of a stress wave. As Cleirigh (1998: 42) points out, thepeak of the stress wave represents the most probable location of the foot in theunfolding rhythmic syntagm. That is, the peak of the stress wave represents anincrease in the probability of the observer's locating the foot at that particularpoint in the overall wave. Each stress in the unfolding syntagm is thus inter-preted as a localized particle to which the listener assigns a discrete foot. Theparticle perspective results in a tendency to isolate the syllable on which the

Figure 3.5: Dependency relations between Ictus and Remiss phases in hever'


strong beat occurs - i.e. the Ictus - as determining the location of the foot. Thismakes the foot seem to be more determinate than it actually is because of thisperspective's emphasis on static segments or points, rather than on the dynamic,temporally unfolding phases of a process and the fuzzy or blurred boundariesbetween segments (Pike 1967; 1982: 12-13). Again, the focus is on the probabilistic assignment of the foot in relation to the point where stress is strongest.

Moreover, the particle perspective privileges a model of linguistic structurewhich is based on the notion of a constituent hierarchy. That is, larger-sizeparticles are comprised of smaller-size particles (see section 8, pp. 134-9). Onanalogy with part-whole principles of organization in the grammar (e.g. theexperiential structure of the clause as a figure consisting of functionally relatedroles [cf. particles] such as participant and process), the foot can be seen asconsisting of lower-level syllabic constituents which function as parts in the largerwhole defined as the foot on the level above that of the syllable. Thus, Ictus andRemiss are particles of, respectively, strong and weak stress. Experientially, thefoot is defined as the functional relationship between particles of strong andweak stress which are realized by lower-ranking syllables. These particles of stressare integrated to the higher-ranking foot in which they function as parts in thislarger whole.

Stress is produced in the vocal tract by varying and modulating factors such asloudness and duration. Loudness and duration are quantitative values, based ondifference in degree or topological-continuous variation. Degree of stress istherefore iconic with respect to degree of loudness and/or duration. From theperspective of articulation, both loudness and duration are iconic with respect to(1) 'the degree of force with which air is expelled from the lungs by thepulmonic air-stream mechanism while the vocal cords are in vibration'(Abercrombie 1967: 95) in the case ofloudness; and (2) the temporal durationof segments, depending on factors such as the length of time that a particularvowel posture is maintained (Abercrombie 1967: 82).

At all levels of organization - syllable, foot, tone group - phonology isstructured in terms of waves of vocal-tract activity. Waves of strong-weak rhythmicpulses constitute a fundamental organizing principle of the expression stratum.Furthermore, rhythmic waves at the level of the foot interact with waves at otherlevels of phonological organization such as the syllable and the tone group. Thefact that periodic, wave-like behaviour is a fundamental organizing principle ofthe expression stratum of speech suggests that it is adapted to the generating ofhighly patterned flows of behaviour which cross-couple with the flows of othersystems (other speakers) in ways that synchronize with each other in stable waysduring interaction. Thus, the feeling of 'phonetic empathy' or of being 'inphase' or 'out-of-phase' with one's interlocutor for the duration of a conversational event, or some part of it, is a product of the ways in which interacting wavevariables at many levels of organization are attuned to the many micro-levelfluctuations in the wave-like behaviour of one's interlocutor. These fluctuationsmay be amplified so that they have larger-scale effects on the organization of theflow of meaning-making activity. This further suggests that interaction is anemergent property of the interaction of the waves produced by the differentinteractants (e.g. speakers).


7. The Integration of Iconic, Indexical and Symbolic Modes of Meaning inLexicogrammar

Hjelmslev (1961 [1943]: 54-6) pointed out that the internal organization of thecontent stratum parallels many of the principles also found on the expressionstratum, relative to its own scalar level(s) of organization (see chapter 2, section1, pp. 59-67). Expression and content in language are describable in terms of afew underlying principles of formal organization which pervade both of thesestrata. As we saw in the case of phonology above, these principles correspond tothe periodic, particulate, wave-like, and chain-like modes of organization whichare characteristic of the four metafunctional regions on both strata. Furthermore, the iconic, indexical, and symbolic modes of meaning which are internalto the content stratum are describable in terms of a nested specificationhierarchy, as outlined above.

According to Deacon, Peirce's typology of icon, index, and symbol designatesa hierarchy of different modes of reference (Deacon 1998 [1997]: 73; seeThibault 2000a: 298-303 for some further proposals). In my view, Peirce's threeway distinction is not restricted to the referential hierarchy proposed by Deacon,but can be used to explain the various ways in which lexicogrammatical formrelates to the plurality of overlapping semantic functions (experiential, interpersonal, textual, logical) that are expressed by forms. Reference is a discoursepragmatic function of language-in-use. It cannot be explained in terms ofany single metafunction; instead, reference depends upon the synergy amongdifferent metafunctional areas in lexicogrammatical and discourse. My interesthere lies in the ways in which iconic, indexical, and symbolic properties can beused to explain various facets of the realization of the semantic metafunctions inlexicogrammatical form. Therefore, my focus in this section is on the intrinsicmetafunctional organization of lexicogrammar rather than on extrinsic referential uses of these intrinsic functions. My interest lies in drawing attention to someof the ways in which the intrinsic metafunctional organization oflexicogrammarcan be seen as a specification hierarchy which extends from the most generalkinds offunctional relationship between form and function (iconic) to the mosthighly specified (symbolic). The logic of the specification hierarchy furtherentails that iconic properties of lexicogrammatical form-function relations areretained when indexical and symbolic properties are progressively added. Thismeans that symbolic properties implicate indexical ones which implicate iconicones.

The relations between levels in the implication hierarchy are transitive(Salthe 1993: 64). This means that what is logically true at any given moment ofthe system is true of the system through all of its moments. For this reason, lexicogrammar has symbolic, indexical, and iconic properties, as reflected in itsintrinsic metafunctional organization. The more general iconic propertiesprobably also reflect earlier conditions of the system prior to the emergence ofthe hierarchy of symbolic, indexical, and iconic properties postulated below.This does not mean that the iconic characteristics of precursor semiotic systemsto language have necessarily been retained in their original form. Rather, theaddition of indexical and symbolic properties in the further development of the


system necessarily entails the reorganization of the lower (innermost) integrativelevels as higher (outermost) levels are, in time, added. The emergent system oflevels 'reflects the categorial reach of some observer, extending from the mostgeneral types relevant to some classification to the most highly specific (highlyspecified)' (Salthe 1993: 65).

7.1 The Iconic Mode ofLexicogrammatical OrganizationThe iconic dimension of lexicogrammar and its discourse-level patterningis revealed in the ways in which the four different modes of meaning - themetafunctions - proposed by Halliday are predisposed to different forms oflexicogrammatical realization. Thus, experiential meanings are realized byparticulate or constituent-like modes of lexicogrammatical realization; interpersonal meanings are realized by field-like prosodic and scopal ones; textualones by wave-like or periodic ones based on peaks of prominence andbeginning-end boundary markers; and logical meanings are realized by chainlike relations of interdependency between items such that one item modifiesanother (Halliday 1979; McGregor 1997: 77-80). The four different modes oflexicogrammatical realization iconically map onto the four different ways inwhich the content strata of lexicogrammar and discourse construct andconstrue the phenomena of experience - real, imagined, remembered, and soon - in their ecosocial environment. The iconic dimension oflexicogrammar isthe least specified, innermost level of a nested hierarchy of iconic, indexical,and symbolic modes.

Each of the four modes of lexicogrammatical realization maps onto thecontinuous topological variation of the phenomena of experience. Experienceis transformed as categories belonging to linguistic content along these fourdimensions, which are intrinsic to the organization of lexicogrammar anddiscourse. This does not mean that these different modes of meaning - the experiential, interpersonal, textual, and logical metafunctions - simply reflect naturalkinds of categories 'out there' in the real world. Rather, they constitute a firstprinciple of their semiotic reorganization according to the internal organizationof the content stratum.

In terms of the three-level hierarchy, the iconic dimension of metafunctionalorganization can be seen as the mapping of topological-continuous variation inthe phenomena of experience (level L-I) to topological-continuous variation inthe modes of realization of the metafunctions in lexicogrammatical form. Thedifferent modes of formal realization in lexicogrammar iconically resembledifferent semantic regions. This is so in the sense that topological variation inlexicogrammatical form maps onto topological variation in the correspondingsemantic region. The three-level hierarchy can be used to model the relationsacross levels described in this paragraph as follows:

Level L+1:Level L:Level L-1:

metafunctional semantic regionsdiverse modes of realization in lexicogrammatical formphenomena of experience (topological)


I shall now consider each of the four metafunctions individually in order to seehow the particular principles of structuration associated with each of the metafunctions iconically resemble a particular kind of meaning.

7.1.1 The Experiential MetafunctionExperiential meaning focuses on the position of particles and the part-wholeanalytical relations that exist between these in configurations of functionallyrelated process-participant relations. Thus, the constituent-like or particulatedimension of grammar is iconically related to the way in which experience isanalytically segmented into a number of component parts which together forma larger whole. For example, a clausally realized situation comprising one ormore participants, a process, and so on, all ofwhich perform different functionalroles in the whole. Consider the following example:

I bought this YJ and I sailed it

This clause complex realizes a sequence in which two clausally realized figuresare related to each other by the paratactic conjunction and. The first clauseexperientially construes a figure in which a process bought relates two participants I and this Vj. The figure which results from this relationship analyses thegiven phenomenon as a number of discrete, functionally related particles. Thefunctional relations among these experiential particles comprise a figure whoseexperiential grammar can be schematized as Actor/-Process: MaterialAction/-Coal. A particular experiential function is mapped on to each of thethree particles such that the given phenomenon is analysed by the experientialgrammar of the clause as consisting of three discrete components that interactwith each other.

This also means that the users of the language probabilistically recategorize,on the content stratum, the continuous waves of the acoustic flux as quanta ofdiscrete lexicogrammatical particles corresponding to experiential categories ofparticipant, process, and so on. They do so by analytically segmenting thecontinuous movement of the wave, or what is more likely, complex patterns ofcycles of overlapping and interacting waves deriving from different parametersof the vocal-tract articulatory activity operating on diverse temporal cycles, aspeaks to which the assignment of particle position is most probable. The particleperspective on the clause focuses on the wave peak as a static position or point,rather than as a wave in constant movement. In so doing, the emergent particulate organization of the experiential categorization of the phenomena ofexperience maximizes the degree of certainty of interpretation of somephenomenon.

7.1.2 The Textual MetafunctionPeriodicity waves ofTheme-Rheme and Given-New in the textual meaning of theclause envisage the clause as a dynamic movement in time comprised of variousphases. Periodic phases of clausal activity can be modelled as waves of content.That is, the clausal wave is interpreted probabilistically in terms of the likelihoodof the language user finding the respective information quantum at a given


position in the overall wave. In the clause, this means that waves of 'thematicity'and 'newness' are interpreted as indicators of the probability of the languageuser finding the respective quantum corresponding to Theme, Rheme, Given, orNew at a given point in the clause. It is in this way that the waves of Theme-Rheme and Given-New flowing through the clause can become quantitized aslocalized particles occurring at, say, the beginning and end points of the clausalwave. For example, the thematicity wave in the English clause consists of twopeaks, corresponding to Theme and Rheme, respectively. These two peaksrepresent the most probable points in the clause for the assignment of thefunctional elements Theme and Rheme. In the following example, the Themesare indicated by bold type; the Rhemes are underscored with a double line.

Figure 3.6: Thematicity wave in the English clause complex, showing Theme andRheme as two peaks of information quanta

7.1.3 The Interpersonal MetafunctionThe prosodic or scopal character of interpersonal meanings can be explained interms of the general tenets of field theory. In physics, a field is 'a region underthe influence of some physical agency' (Pitt 1977: 149). On the content stratum,an interpersonal field is the domain surrounding a given interpersonal selection- e.g. mood or modality in the clause - such that the given selection syntagmat-ically extends over and in some way modifies or deforms (influences) some otherfeature within the given syntagmatic domain, which it holds in its scope. Thegiven interpersonal feature influences and shapes (deforms) the given syntag-matic domain in order to achieve a particular interactive purpose. In thisperspective, the process of meaning-making can be seen as one of continuallyassigning quanta of interpersonal meaning to some syntagmatic domain so as to'energize' it for specific interactional purposes and effects. Thus, a given

maybe his fears came out so he opposed it violently and . . . but mum saved

some money up and she lent me some money to buy a boat

Figure 3.6 models the thematicity wave in this syntagm.


interpersonal quantum, such as a modality or mood choice in the clause, is aquantum of energy which acts on or extends over and therefore influences aclausal field. Interpersonal quanta such as mood or modality are energeticcomponents of an overall interpersonal field.

In the following example, the projecting clause I think and the modal adjunctprobably are modalities which act on and extend over the entire declarativeproposition in double square brackets so as to give the speaker's interactiveperspective on that proposition.

I think [[probably [the first thing that gave me an opportunity of understanding what I was going through was my mum saved up she made stringbags]]]

The projecting clause I think holds the projected clause in its scope in the sensethat the modality which is expressed in the projecting clause extends overthe modalized field of the proposition in the projected clause. In so doing, thespeaker provides his subjective (first person) modal assessment concerningthe possibility of the proposition that is expressed by the projected clause. Thesemantic extension of the modality can be modelled as the field within which themodality exerts its influence. Such a field can be seen as a modally chargedsemantic field. Figure 3.7 illustrates the full complexity of the choices inmodality and mood which operates in this syntagm on various levels, as well asshowing the field over which each choice's modalizing influence extends.

On the outermost layer, there is the mood choice of the projecting clause.The mood choice holds the proposition in this clause in its scope and indicatesthat it is being asserted as a proposition that can be argued about in thediscourse. On the second layer, the projecting clause is a modality of the kindHalliday identifies as having the feature 'subjective: explicit'. This kind ofmodality reveals the inherent self-reflexivity of human language. That is, humanlanguage has resources for acting on and reflecting upon itself such that themodality in this case involves me ta-linguistic awareness. In this case, the speaker'suse of the projecting clause shows a reflexive self-awareness of his own processesof modal evaluation. The third layer consists of the comment adjunct probably,which indicates the speaker's evaluation of the content of the proposition in theprojected clause. Finally, the fourth or innermost layer consists of the moodchoice of the projected clause. Once again, the choice of declarative moodindicates that the proposition in the projected clause is being asserted by thespeaker as something which can be argued about in the discourse.

7.1.4 TheLogical MetafunctionLogical relations in the lexicogrammar are concerned, for example, with theways in which clauses combine with other clauses. The resulting combination ofclauses is called the 'clause complex'. The logical relations between clausesiconically reflect the ways in which people use language to construe relations oftemporal and causal dependency between the experienced phenomena inclauses (see chapter 4, section 6, pp. 195-8). In the following sequence, thespeaker's use of the conjunction and iconically reflects the way in which the


Figure 3.7: Multiple modalizing fields acting on the syntagmatic domain of aproposition, showing the scope over which each field's modalized influence extends

experienced phenomena in each clause are related to each other. For example,the logical relationship of addition between there are some of us who are thin andsome of us who are thick and we all have a little problem iconically reflects the rela-tionship of extending the information in the first clause by adding further infor-mation to it in the second clause. In this example, the relationship of'extension'is signalled by the conjunction and. The logical relationship between I ah I ah wasgoing very well actually and until about January the 4th and we started the electioncampaign is indicated by when. In this case, the relationship is between a givensituation in the first clause and a temporal circumstance in the second clause.

7.2 The Indexical Mode of Lexicogrammatical OrganizationThe four iconic modes referred to above are integrated to the indexicalproperties of lexicogrammatical signs. The indexical properties of lexicogram-matical signs include the Finite element in the verbal group, which functionsindexically to ground the clause in relation to a reference point in the


temporally unfolding discourse event. The Finite element achieves this indexicalfunction through either primary tense or modality. Primary tense is indexicalinsofar as it involves temporal proximity deixis: it situates the process instantiated by the verbal group in time relative to the moment of utterance. Modalityis indexical in the sense that it involves what might be called attitudinal orevaluative proximity deixis: it situates the process instance in relation to evaluations of its probability, usuality, certainty, and so on, relative to the speaker indeclarative clauses and the addressee in interrogative clauses. Davidse (1997)further points out that the process instance is also grounded in terms of persondeixis by the grammatical Subject - it is linked to grammatical person (first,second, third). In this way, both Subject and Finite perform indexical groundingfunctions which link the clause to the I-you and here-now of the discourse event.

At group rank, the Thing instance, which is realized by the Head of thenominal group, may be grounded in terms of spatial proximity by means of thedemonstratives or in terms of person deixis by means of the possessives (Davidse1997). Whereas the iconic properties of lexicogrammatical signs that I discussedabove reflect general, probably universal design principles that relate specificmodes of meaning to their particular modes of realization in the lexicogrammarof all languages, the indexical and symbolic properties of lexicogrammaticalsigns are concerned with increasingly specified principles that characterizeindividual languages. For example, the grounding function of the Subject is anindexical property of English, though not necessarily of all other languages. Inany case, both indexical and symbolic properties of lexicogrammar areorganized in terms of a rank scale of lexicogrammatical forms on level L whichhave the function of filtering non-criterial continuous topological variation inthe phenomena of experience on level L-l for the purposes of its construal bythe semantic categories recognized by a given SI on level L+1.

7.3 The Symbolic Properties ofLexicogrammatical SignsLexicogrammar is also a system of symbolic possibilities for making meaning byvirtue of its systems of typological-categorial distinctions between, for instance,different experiential categories of process (material, mental, verbal, ete.) andtheir associated participant roles; between different interpersonal categories ofmood (declarative, interrogative, oblative, imperative) and their associatedpossibilities for enacting different interactive positions (addresser, addressee,giver of information, seeker of information, and so on) in the negotiation andexchange of meanings in discourse; and between different textual categories of,for instance, deictics. A symbolic system is not necessarily tied to its environments in any specific way. This is so because the specifically symbolic propertiesof, for instance, language consist of systems of categorial distinctions. Suchdistinctions are defined by the value which each distinction is assigned by virtueof its relations to all the other terms that constitute a given system of relations.These systems of semiotically salient differences in some community areorganized as language-specific functional systems and structures which symbolizeexperiential, interpersonal, deictic, and other categories. This is true for phonological systems and structures as well as lexicogrammatical ones. It is by virtue ofthe functionally organized paradigmatic and syntagmatic environments in which

More precisely, the linguistic form a symbolizes a category of determiner whichcan be more delicately subcategorized as [DETERMINER: NON-SPECIFIC;NUMBER: SINGULAR: COUNT]. Thus, a is used in the grammatical environ-ment of singular count nouns, as in the example above. The properties of aunder discussion here are symbolic properties that derive from (1) the positionof this lexicogrammatical form in a system of paradigmatic contrasts with otherforms such as the; and (2) the possibilities of this form for operating in particulartypes of grammatical structure such as the nominal group. The nominal groupis a particular subsystem in the language with grammatical units and structureswhich are specific to that level. These units and structures are specific to thefunctional tasks of the grammatical subsystem which is the nominal group.Specifically, the grammatical resources of the nominal group function to groundby deictic and other means the Thing element as an instance of the type-categorywhich is symbolized by the Thing. In the nominal group a very good sailor, thedeterminer a functions deictically to specify the Thing element as an arbitrary ornon-specific instance of the type-category of Thing in question. If determinerssuch as English a and the were pure indexicals, then they could not be freelycombined with other items such as, for instance, adjectives and nouns by virtueof the grammatical resources of the nominal group. That is, they would benecessarily tied to the specific contexts in which their indexed entities occur.However, it is the symbolic categorial-coding potential of these forms which


particular distinctions (cf. Saussure's values) occur that a given lexicogrammati-cal form can operate in very many different contexts rather than being tied to agiven context.

For example, the Head element in a nominal group, e.g. a very good sailor,symbolizes an experiential type-category of Thing, viz. [SAILOR] in the presentexample. It does not refer to a specific sailor in a particular real-world context.Similarly, the adjective good symbolizes a type-category of quality just as thedeterminer a symbolizes the deictic category 'non-specific'. Each of these itemsderives its symbolic potential from its position in a system of contrasting termswhereby the value of each and every term is defined. For example, a contrastswith the in the sense that both are more delicate instantiations of the schematiccategory [DETERMINER]. The more delicate (specific) instantiations constitutea system of contrasting terms in which the meanings qua categorial distinctionsbetween a and the are defined. Figure 3.8 shows this contrast.

DETERMINER

SPECIFIC the

NON-SPECIFIC a

Figure 3.8: Simplified determiner system, showing the categorial distingtion betwenthe values 'specific' and 'non-specific', as symbolized by English 'the' and 'a'


enables them to be used in very many different contexts. That is, the categoriesof deixis and number that are symbolized by these forms are not tied byindexical necessity to any specific context, but constitute a systemically definedmeaning potential which can be used and played with across very many lexicogrammatical and discourse contexts.

The observations I have made thus far refer to the semantic functions of itemssuch as determiner, adjective, and noun in the grammatical subsystem 'nominalgroup'. Nominal groups are also integrated to the higher-scalar grammaticalrank of the clause such that the nominal group realizes a particular participantrole within the grammatical structure of this higher ranking system. In thepresent case, the clause from which the example is taken is: I wasn't a very goodsailor. In this clause, the nominal group a verygoodsailorfunctions as Attribute ina relational attributive clause comprising the following functional roles: I(Carrier) wasn't (Process: relational: attributive) and a verygoodsailor (Attribute).It is only by virtue of this nominal group's integration to the higher ranking unitof clause that it becomes possible to see how the symbolic properties of thenominal group interact with its indexical meaning in the specific discoursesituation.

Thus, the speaker in the situation of utterance of the clause, as indexed(exophorically pointed to) by the first-person pronoun I, is also indexed, byvirtue of the semantic relationship of attribution that links I to a very good sailorin this clause, as a non-specific instance of the type-category of Thing inquestion. Thus, we see how, in the particular context, it is in the interactionbetween situational and lexicogrammatical (clause and nominal group)properties and relations that the combined indexical and symbolic propertiesand functions of these lexicogrammatical forms are revealed. The non-specificdeixis of the nominal group means that the nominal group cannot in itself haveany referential function of picking out a specific entity for the purposes ofreferring to it and subsequently tracking it through the discourse. Instead, it isthe clause-level relation of attribution which creates the link between indexedspeaker and nominal attribute that is assigned to speaker. Table 3.2 summarizesthe relevant properties of the iconic, indexical, and symbolic dimensions ofsemiosis, as discussed in this section.

8. Language as Particle, Wave, and Field

The tripartite perspective on language as particle, wave, and field first proposedby Pike (1967) and subsequently taken up and adapted by Halliday (1979) showshow both phonological and lexicogrammatical form are organized by a numberof distinct, though complementary, principles which ramify across all levels oflinguistic organization. These diverse though complementary principles oforganization illustrate two fundamental principles concerning the organizationof language form. First, the intrinsic organization of language cannot bereduced to any single principle of organization. Secondly, the complementarymodes of organization identified by Pike and Halliday show how the two external'realities' with which language form interfaces - viz. the body of the articulator

BODY DYNAMICS, MEANING-MAKING,

Table 3.2: Semiotic properties oJ icon, index, and symbol

135

Type of Semiosis Properties Scale

topological-continuousvariation in physical-material dynamics of body

iconic mapping of topological- non-intentional here-continuous variation to now sensori-motortopological-continuous dynamics of bodilyvariation through cross- states which crosscoupling of variation in boundary betweenbody dynamics to internal states of systemaffective and other and immediate environ-internal states of mentorganism

index spatial-temporalextension to and inten-tional projection intoimmediate environ-ment which can besampled and individu-ally remembered

symbolic space-time scalesbeyond here-nowinvolving history andcultural memory

typological-eategoricaldifference

and the phenomena of experience - are selectively transformed as functionallydiverse areas of meaning, each of which makes its distinctive contribution to theinternal organization of both expression and content. Thirdly, the functionallydiverse regions of meaning and their corresponding principles of structurationin both expression and content implicate iconic, indexical, and symbolic modesof meaning. As we have already seen, the iconic mode reflects the most generalprinciples of categorization and the symbolic the most highly specified.

The tripartite perspective of particle, wave, and field in the clause shows thatmeaning construal is not localizable at a single point in the clause. Instead, itis the time-bound interaction of all three perspectives as they are dynamicallydistributed over the global totality of the clause that yields a probabilistic interpretation of a given reality. Thus, the experiential categorization of phenomena,their grounding in a dynamically charged interpersonal field, and their


quantitization as bits of information relative to the temporally developingdiscourse context are all necessarily present in the semiotic process of interpreting and intervening in the phenomena of experience. It is the language user'sprobabilistic relationship to the wave/particle/field complementarity of anygiven quantum of meaning in the clause that affords the interpretation of reality.Moreover, this complementarity is a contextual one in which the internalorganization of language as expression and content, 'reality' in the form of thearticulatory (e.g. vocal-tract) and ecosocial dynamics modelled by these twostrata, and the observer are all internally connected to each other.

The stratal organization of language as expression and content reflects theway in which, at their respective scales, there is no ultimate or fundamenta1levelof reality which is supervenient to the rest. Language is not epiphenomenal tosome more basic level of organization. Instead, both expression and content aredescribable at all levels as quantum processes which exist as patterns of particle,wave, and field (Bickard and Campbe1l2000: 331). Complex interacting patternsof semiotic processes involving 'what is observed', 'who observes' and the 'meansof observation' lead to the emergence of both articulatory phenomena and thephenomena of experience on the body and ecosocial scales, respectively. Theintrinsic metafunctional diversity of both expression and content means thatlanguage and other semiotic modalities are not supervenient to any singleunderlying principle. Rather, the complementarity of the particle, wave, andfield dimensions of language form enables the many degrees of topologicalvariety on level L-1 of, for example, the neuromuscular substrate of vocal-tractactivity to be entrained to the phonological trajectories of a given language.Moreover, the particle/wave/field perspective means that the principlesgoverning their emergence can be precisely formulated. On the other hand, themajority of currently available models of meaning-making, representation, andinformation processing are founded on one version or other of conceptual orrepresentational atoms or primitives, each having its own representationalcontent (Bickard and Campbell 2000: 325). Such reductionist accounts arefounded on principles of upward causation which seek to explain, usually insome sort of causal-explanatory framework, how such representational atoms areassembled to form larger wholes.

The tripartite perspective of language as particle, wave, and field suggests thatlanguage, at all levels of its organization, cannot be reduced to fundamental representational atoms or conceptual-semantic or phonetic primitives. Language atall levels is organization of quantum field processes (Bickard and Campbell2000: 334-5). Particles or constituents are no more than local stabilizations on agiven level of principles of organization of underlying field processes. Forexample, the segmentation of the continuous flow of vocal-tract activity in theproduction of speech sounds into discrete phonemes amounts to specifyingpositions in the flow where systems of paradigmatic features are localized asprobably occurring at that point. How do we explain these stabilizations againsta background of endless flux? Typically, linguistic theories have modelledlanguage in bottom-up terms as comprising constituent particles and thestructural relations among these. The resulting constituent structure ispostulated as a supervenience base on which everything else depends. For


example, the linguistic entity 'sentence' is seen as being supervenient on certainclasses of lower-level constituents standing in certain kinds of relationships witheach other. Formal models of language, for example, are not concerned withrelations outside those between 'sentence' and its lower syntactic constituentsand the formal rules for their combination. Such a system of relations is closedto the external environments in which language operates.

On the other hand, if language is seen as a form of dynamic open system, thenit is, by definition, engaged in transactions with its environment over the time inwhich the system persists. In this alternative view,constituency (segmental organization) is seen to be the consequence of a particular and partial way of relatingto the wave. The wave is the organizing principle, rather than constituency assuch. Without such system-environment transactions, the system cannot anddoes not exist. It follows from this that the characteristics which language has asa dynamic open system cannot be derived from closed-system properties andrelations such as those seen as dependent on a supervenience base. Instead, suchproperties and relations can only be defined in terms of language qua dynamic,far from equilibrium system. In the terms of the present study, language, ratherthan being supervenient on a causal base of fundamental properties, persists intime and maintain its dynamical structural integrity and intrinsic properties andrelations in time because of the transactions that occur between its internalorganization and two environments, viz. the body-brain complex and theecosocial environment, with which it is cross-coupled. There is no supervenientsubstance-base on which language is ultimately founded. The reason for this liesin the way both of the interfaces referred to above are themselves persistences inthe organization of bodily (neuromuscular) and ecosocial processes. Instead ofbeing explainable in terms of fundamental constituent particles, such localstabilities or patterns of organization are explainable in terms of the patternednature of the organization of the relevant system - environment transactionsthat are constituted by the expression and content strata.

The Principle ofAlternation proposed by Lemke (1999) allows us to view thisquestion in a different perspective. Quantum field processes relative to thespace-time scales of the body and the phenomena of experience self-organizeinto higher-scalar semiotic levels, relative to the observer perspective in some SI,where they are construed as in some way semiotically salient or meaningful. Forexample, the topological-continuous space-time of vocal-tract activity on levelL-l is an organized field (a phase space) of articulatory processes which arereorganized as complex interacting patterns of phonological particles, waves,and fields for level L+I by virtue of the filtering or buffering function of theintermediate level L of phonological units and relations functions. In this way,emergent principles of phonological organization, themselves no more than afurther reorganization of quantum field processes on their own level, are mademeaningful for higher levels of organization, viz. lexicogrammar and discourse.

Language and other semiotic system-processes are not supervenient. Theirexistence is dependent on a number of environmental factors which can bethought about in terms of Aristotelian causality. First, they depend on biologicalinitiating conditions - e.g. neuromuscular processes - such as those whichenable vocal-tract gestural activity and their effects to take place (efficient


causality). Thus, waves of vocal-tract activity may be said to cause (i.e. efficiently)the compression of sound waves which are propagated from a source. Secondly,they depend on material environmental affordances such as available physicalmaterial resources - e.g. the air as a medium for the propagation ofsound waves,treated surfaces such as paper, the nature of human perceptual systems, itsamplification and diffusion by culturally evolved technologies such as writing,the Internet, and forms of social organization (material causes). Thirdly, thereare principles of organization internal to language such as its dependence on asemiotically salient system of differences and combinations of typical patterns ofphonological, grammatical, and discourse systems and structures (formalcauses). Fourthly, final causes are those whereby the system anticipates and isoriented to its own future along a temporal trajectory, which is the only way inwhich the system as such exists. The organization of ecosocial semiotic processesand the perceived stabilities or regularities in these processes can only take placein the presence of appropriate forms of organization of all four factors onvarious space-time scales.

Language is a transindividual system organized on many different space-timescales. It cannot be appropriately reduced to units or relations on anyone scale.Moreover, language is an instance of the general class of dynamic open systemswhich maintain their existence and integrity in time. In this way, the causal consequences of a given semiotic system also persist in time, again on a multiplicity ofintersecting space-time scales. For example, the processes that operate on the levelof expression - e.g. vocal-tract activity - are strongly constrained by processes onthe content stratum. This is consistent with the view expressed above that meaningflows from an intentional source and constrains lower-level sensori-motor systemsin the process of putting into effect a particular action. Meaning-making can thusbe seen as the entraining and reorganization across levels of processes ofexpression and content by higher-order proximate intentions in symbolic neuralspace. Action is the entrainment of such processes along a trajectory which hascausal, though not efficiently causal, consequences. This is consistent with the viewthat the brain recategorizes sensori-motor activity as conceptual routines, and soon (see section 13, pp. 162-8). Thus, the storing and maintaining of a stable - notstatic - language system comprising both expression and content on both theecosocial and individual (organismic) levels means that the causal consequencesof that system persist on many different levels.

For example, the entraining of meaning along an action trajectory, with itsconsequent cross-coupling of both semiotic-discursive and physical-materialrelations and flows, makes an asymmetric difference in the world if the emergentmeaning-making trajectory that results is able to flow into a sink of semioticallymediated activity that corresponds to the proximate intention of, say, the speaker(Juarrero 1999: 193). Signs may point to (index) and/or symbolically construe agiven contextual value as the purpose or goal of the action trajectory. This meansthat the intending agent must construct the indexical or symbolic link betweensign and intended goal or purpose in order for the latter to be the object of aparticular course of action. Thus, intentionally directed action or goal-seekingactivity which is directed to some purpose are made possible by processes ofsemiotic mediation that link what is inside the organism to what is outside it.


Purposes, goals, objects of desire, and so on, are not inherent in the phenomenaof experience per se. They are emergent consequences of the semiotic mediationbetween proximate intentions and relevant contextual values.

The neuro-semiotic recategorization of expression as content indicates thevalidity of the hypothesis that content is an emergent level of organizationrelative to the lower-level expression stratum. In this view, such properties anddynamical relations as emerge on the higher level are no less real or causallyeffective than are properties and dynamical relations on the lower level. It thusfollows that the properties and relations of the higher level will possess causalproperties that the lower level does not possess. Indeed, the scalar heterogeneity of language implicates that emergent indexical and symbolic properties oflanguage ramify across and have effects over potentially very many space-timescales beyond the immediate organismic one.

The hierarchical organization of both the expression and content strata illustrates how the many degrees of freedom on level L-l constitute enablingconditions in the form of very many degrees of continuous topological variationthat a dynamically emergent system can access through the buffering or filteringfunctions of level L. Constraints emanating from level L mean that dynamicalrelations and properties on level L-l are entrained to these higher-level constraints. For example, articulatory repertoires on the phonetic level of vocal-tractactivity are reorganized and entrained to the patterns of organized behaviour vocal-tract trajectories - that are afforded by the phonological systems andstructures on level L. The constraints arising from this level allow vocal-tractactivity to self-organize into more complex forms of activity that would not havebeen possible if the many degrees of topological variation existing on level L-Iwere not subject to the filtering functions of level L. In this way, level L-l providesa large repertoire of potential information which the higher level can access andthen reorganize (section 2, pp. 109-16). The entrainment of this variety to newlyemergent patterns of phonological activity on level L increases the organism'smeaning-making potential. This increase occurs by integrating the organism to acontextual level comprising indexical and symbolic semiotic modes. It is thesemodes which give the organism access to scalar levels beyond the immediatebody-scale along with new possibilities of interacting with scales beyond the herenow scale. In other words, in integrating the organism to higher scalar levels, levelL enables the organism to be functionally part of ecosocial relational dynamicsand structures which the level L-l alone could not bring about.

9. The Brain as Selective Recognition System: Language and Edelman'sTheory of Recategorization

Edelman (1992: chapters 8-9) has proposed that the brain is a selective recognition system equipped with a capacity to memorize what it recognizes. In bothinfant protolanguage (ontogenesis) and in the evolution of language in thehuman species (phylogenesis), phonetic repertoires gradually become linked toconceptual repertoires through processes of association. That is, some patternsof sensori-motor vocal-tract activity are selectively cross-coupled to some patterns


of concept recognition, in the process leading to the making of even finersensori-motor discriminations or differentiations on account of the reorganizingeffects emanating from the higher, conceptual level.

... to build syntax or the bases of grammar, the brain must have reentrantstructures that allow the semantics to emerge first (prior to syntax) by relatingphonological symbols to concepts. Because of the special memory providedby Broca's and Wernicke's areas, the phonological, semantic, and syntacticallevels can interact directly and also indirectly via reentrant circuits that areformed between these speech areas and those brain areas that subserve valuecategory memory. When a sufficiently large lexicon is collected, theconceptual areas of the brain categorize the order of speech elements, anorder that is then stabilized in memory as syntax. In other words, the brainrecursively relates semantic to phonological sequences and then generatessyntactic correspondences, not from preexisting rules, but by treating rulesdeveloping in memory as objects for conceptual manipulation. Memory, comprehension, and speech production interact in a great variety of ways byreentry. This permits the production of higher-order structures (such assentences in a grammar) and obviously helps with the elaboration of lowerorder sequences (such as phrases). Of course, once achieved, the sequencingbecomes automatic, as do many other motor acts. (Edelman 1992: 130; italicsin original)

The re categorization of conceptual repertoires as lexicogrammatical systems cf. Edelman's 'syntax' - paves the way for a massively expanded meaning-makingpotential on account of the qualitative differences that the global semioticreorganization brought about by the new level entails. The recategorization ofconceptual systems based on sensori-motor routines as lexicogrammatical onesmeans that the organism has dynamical properties it would not otherwise have.In turn, these lexicogrammatical systems which are memorized and elaboratedin the brain in the course of the organism's development and individuationalong its historical-biographical trajectory are integrated to still-higher-scalarpatterns of organization by way of the semantic registers, discourse genres, socialactivity structure types, and so on, that are the typical ways of making meaningin a given community, i.e. its SI on level L+1. This emergent hierarchy ofdynamical neuro-semiotic relations means that progressively higher levels selforganize. Each newly emergent level gives rise to dynamical relations andproperties specific to its level and, at the same time, the overall system's degreesof freedom - its possibilities for social meaning-making - is expanded. By thesame token, the higher levels impose their own constraints on the lower ones.That is one good reason why discourse and lexicogrammar are two distinct levelsof meaning-making within the content stratum in systemic-functional linguistics:each level makes its own distinctive contribution to the meaning of the whole,and at the same time is contextually integrated to the other levels.

Nevertheless, I would argue that Edelman tends to privilege a two-level modelcomprising sensori-motor routines and conceptual repertoires in order toexplain the emergence of language in the individual. Conceptual repertoires


obviously depend on the organism's time-bound perceptual samplings of theenvironment in the form of the extra-somatic outside world as well as theindividual's internal milieu. I do not think there is a simple linking of sensorimotor routines to conceptual ones. Both phylogenetically and ontogenetically,the cross-linking of the two is always mediated by the higher-scalar ecosocial environment in which the organism's perceptual sampling of the environment andconsequent elaboration of conceptual structures is embedded. As Lemke(2000a: 210-11) points out, there is not a simple 'upwards' progression fromsimpler to more complex forms of organization. Rather, there is always already ahigher form of organization, or an organized field of possibilities, on higherscalar levels which mediates the body-brain's transactions with its environment.Thus, the emergence of the content stratum of language as a recategorization oflower-level conceptual structures is the result of the emergent self-organizationof a more specified intermediate level coming between the conceptual structureselaborated in the brain and the ecosocial environment.

As I have argued elsewhere (Thibault 2000a), Edelman's focus on theindividual organism tends to downplay the ways in which the emergence oflanguage, both phylogenetically and ontogenetically, in the individual bodybrain takes place in the context of higher-scalar ecosocial systems which arethemselves continually evolving. Edelman is certainly right to emphasize that thedynamics of language in the individual does not depend on a geneticallyendowed knowledge of language which predicts from the outset the 'growth' oflanguage in the individual. Instead, the dynamics of language in the individualis entrained, in and through social interaction, to the dynamics of higher-scalarecosocial processes which themselves only have their existence along a timebound trajectory. Precisely the same argument also applies to the sensori-motorroutines mentioned above. These, too, only exist along a self-organizing timebound trajectory of vocal-tract gestural activity. Such a trajectory is based onnon-linear and irreversible cycles of neural and muscular activity, which are themicro-eomponents out of which vocal-tract activity is flexibly soft-assembledalong a time-bound trajectory (Kelso et al. 1986; Goldfield 1993: 53). Theassembling in time of such micro-components leads to the emergence of a coordinative structure: many interacting functional variables behave as a singleunit of activity. In this way, very many degrees of freedom - i.e. of continuoustopological variation - on the lower level of the body are transformed intomeaningful information for higher levels of organization. The self-organizingemergence of phonological systems and structures which recategorize vocal-tractactivity is a more specified intermediate level between these sensori-motorroutines and the higher level of the content stratum. On both levels, lower-levelnoise is amplified and transformed into information which is meaningful forhigher levels. The quantal theory of phonology developed by Stevens (1989)constitutes one of the important pioneering attempts to explain how macroscopic categories emerge from the microscopic detail of the articulatory andacoustic phenomena involved in the production and reception of phonologicalcategories.

The processes of re categorization discussed above depend on the system'scapacity for discrimination or differentiation. For example, the sensori-motor


interface between organism and environment relates properties and processesinternal to the individual subsystem to its higher-scalar environments: it providesthe means whereby such properties and relations relate to the contexts in whichsuch transactions take place. Sensori-motor operations are operations of discrimination or differentiation. The two-way nature of such operations meansthat they determine both the structure of the relevant system and the way itmodels its environment. Lorenzana (1993: 295) has pointed out that evolutionary systems [ESs], in general, 'need to build diversity from their simple set ofgauges ... this construction of diversity from an initial undifferentiated unity isan essential characteristic of ESs, imposing a condition of gradualism'. In thelight of this observation, we can understand the recategorization of sensorimotor repertoires as conceptual repertoires to be an instance of the emergent,self-organizing construction of a new scalar level of organization which allows thesystem to interact with increasing environmental diversity on increasingly largerscales.

The emergence of the content and expression strata of language represents afurther increase in the system's capacity to access its potential environment. Theemergence of these two intermediate levels of organization constitutes anincrease in the system's discriminatory power, thereby allowing the system toaccess and interact with increased environmental diversity across potentiallymany different scalar levels. Thus, the stratified nature of semiotic systems andtheir emergence as intermediate levels of organization between prior, less discriminatory systems is a specific instance of that type of hierarchy known as a'combinatorial hierarchy' (Lorenzana 1993: 295). In a combinatorial hierarchy,discriminatory operations construct increased typological-categorial differentiation and diversity from simpler, prior conditions of topological-continuous unity.Lorenzana points out that the 'chain of levels' in combinatorial hierarchies islimited due both to specific contextual pressures and to system-environmentrelations. For example, the further stratification of language in terms of registerand genre (Martin 1991) is the result of pressure from the ecosocial system. Thismeans that principles of social organization, the participation of differentcategories of social agents in diverse social networks, the division of social labour,and so on, constitute further principles whereby language is organized in termsof the diverse semantic registers, discourse genres, and so on, which arefunctional in different domains of social life. Figure 3.9 models the basicprinciples of the combinatorial hierarchy with respect to the expression-stratumdynamics of speech.

Each level in the hierarchy is formed out of combinations of elements on thelevel below. Level 1 refers to the oral cavity as a three-dimensional physical space,characterized by undifferentiated topological-continuous variation. Level 2shows the reorganization of Level-l dynamical processes as a multidimensionalphysiological space. At this level, new couplings from different dimensions notcongruent with physical space per sebecome possible. Level 3 shows how thesemultidimensional physiological couplings give rise to language-specific phoneticrepertoires independent of phonological function. Level 4 shows how thesephonetic repertoires are further reorganized as the phonological systems andstructures of a particular language.


Figure 3.9: Expression stratum as combinatorial hierarchy, showing reorganization ofhigher levels as recouplings of elements on lower levels thereby allowing for the emergenceof new systemic possibilities and, hence, new system-environment couplings

An analogous and parallel system of levels can be proposed in relation to thecontent stratum and the way it interfaces with the ecosocial environment. In thiscase, Level 1 designates the topological-continuous flux of ambient stimulusenergy in the environment. Level 2 refers to the cross-modal perceptualsampling and pick-up of this flux as information about events in the organism'senvironment. Such information necessarily correlates in the brain informationfrom a diversity of perceptual modalities. In this way, new couplings occur whichcombine inputs from diverse channels not congruent with a single physicalparameter. Level 3 shows how the resulting perceptual categorizations arefurther filtered and reorganized by the systems of semiotically salient differences- the semiotic values - specific to the particular semiotic resource systems ofsome community. Level 4 shows the further reorganization of these as the lexi-cogrammatical and discourse systems and structures of a given language. Theemergence of these levels on the content stratum of neuro-semiotic organizationis not without consequences for the expression stratum. Thus, the initial linkingof vocal-tract sensori-motor routines to pre-linguistic conceptual ones and theconsequent finer discriminations made by the former also brings about theemergence of specifically phonological systems and structures at level L on theexpression stratum.

A language system stored in the individual's brain is a self-organizing and dis-tributed global neural space comprising both expression and content. These twolevels are interdependent and, at the same time, each has its own internal

Level 4

Level 3

Level 2

Level 1


principles of organization and integrity. Expression faces 'inwards' to the bodyof the articulator and 'outwards' to the content stratum; content faces 'inwards'to the expression stratum and 'outwards' to the ecosocial environmentcomprising the phenomena of experience. As the three-level hierarchy principlehas shown us, each stratum is comprised of both initiating conditions andhigher-scalar constraints in the form ofan SI specific to that stratum. By the sametoken, the higher-scalar content stratum exerts downwardly causal constraints onthe expression stratum. The principle of the selective recategorization of sensorimotor routines by higher-order conceptual ones and the further recategorization of these by linguistic ones (phonological and lexicogrammatical) meansthat content imposes constraints on expression such that expression is selectivelyand functionally adapted and integrated to the requirements of content.

Overall, the number of semiotic differentiations made and their possible combinations is far greater on the content stratum with respect to the comparativelymuch smaller number of phonological systems and structures in a givenlanguage. Thus, the learning and interpreting of higher-scalar content units andrelations takes place over a much more extended timescale such as an entirelifespan trajectory at the same time as it requires a far greater number of interconnected neural networks in the long-term memory of symbolic neural spacethan is the case for units and relations on the expression stratum. Childrendevelop the phonological structures and systems of a given language over arelatively much shorter timescale compared to that of their development oflexicogrammatical and discourse systems and structures. The principle ofrealization, which nicely resonates with Edelman's notion of selective recategorization, means that units and relations on the expression stratum are selectivelyand mutually contextualized with respect to units and relations on the contentstratum.

Thus, the sensori-motor activity which is entrained and organized accordingto the phonological systems and structures of a given language restrict thedegrees of freedom of the body's dynamics. That is, body dynamics are entrainedto the phonological systems and structures of that language. This entrainment isfunctional for the purposes of making and negotiating meanings with others insocial contexts on diverse space-time scales. This means that human bodies arenow integrated to the higher-scalar ecosociallevel of organization where sensorimotor (e.g. vocal-tract) activity is reorganized so as to take on new functionalpossibilities on this level. Moreover, the emergence of properties and relationsspecific to the content stratum in symbolic neural space is constitutive of ourphenomenology of lived experience in ecosocial space-time. Meaning is not amere epiphenomenon of experience. For example, the very fast scale ofprocesses on the sensori-motor and eNS levels is integrated to and entrained bythe somewhat slower scale afforded by the rhythmic and other patterns oftemporal organization of a given language's phonological systems andstructures. In turn, these are integrated to the still-slower patterns of rhythmicorganization characteristic of socio-discursive patterns of organization evident intexts and social action (see Togeby 2000: 273-7).

In this way, to take just one example, our experience of time is structuredaccording to modes of organization that are endogenous to the dynamical


relations existing on the ecosocial scale rather than being referable to lowerscalar sensori-motor dynamics perse, seen as their biological initiating conditions.These higher-scalar phenomenal relations are not exogenous factors withrespect to the body. Instead, in integrating the body to them, they become a partof the body-brain's internal organization so that scalar heterogeneity is seen as afundamental property of the meaning-making body-brain. Thus, we can see how,at all levels of organization, language is not composed of isolable functionalcomponents assembled 'from below'. Rather, the metafunctional diversity that isattributed to unfolding acts of meaning-making in logogenesis reflectsorganismic and ecosocial constraints that influence the time-bound development of the system without recourse to pre-existing causal factors to explain thesystem's behaviour. Expression and content eo-evolve and eo-develop in thesense that the neuromuscular dynamics of the body are entrained in andthrough their interaction, both phylogenetically and ontogenetically, withhigher-scalar content-stratum processes and their functionings in a still-largerscalar ecosocial environment.

At its particular level of organization, vocal-tract gestural activity can bedescribed as a vector field having the properties of both direction andmagnitude (Norton 1995: 50; Cleirigh 1998: 52). Relations both within andbetween the expression and content strata are non-mechanistic informationalconstraints. Final causality operates here because top-down relations between,say, the content stratum, comprising lexicogrammar and discourse, and vocaltract trajectory function both to select and to constrain the lower level at thesame time as they also act to entrain it into its future trajectory (Salthe 1993:270). This is so because the cogent moment - the temporal cycle - of the higherlevel of meaning on the content stratum is greater such that the flow of meaningalong its trajectory continues well after the completion of the much fastertemporal cycle of the articulatory trajectory at the level of the syllable, say,on theexpression stratum (section 2, pp. 109-16). Informational constraints alsooperate on the relations between levels within both the content and expressionstrata. The relationship between phonology and phonetics that I discussed abovecan be seen in this light: phonological systems and structures are macroscopicrelations and dynamics; phonetics refers to microscopic relations and dynamicsat the fine-grained level of articulation (Browman and Goldstein 1995: 180). Theconstraints do not operate in just one direction, but instead are reciprocal,acting in both directions.

On the content stratum, lexicogrammar and discourse similarly operate asystem of reciprocal constraints between the microscopic and the macroscopiclevels of description. Moreover, the self-organizing properties of lower-levelrelations and properties such as those evident in articulation on the phoneticlevel give rise to emergent patterns of organization at higher levels such as thephonological level. By the same token, the emergent higher level downwardlyconstrains activity on the lower level. The realizatory relationship betweencontent and expression can now be seen as a system of reciprocal linkagesbetween dynamical relations and processes on a number of different levels oforganization, ranging from the microscopic level of articulation on theexpression stratum to the macroscopic level of discourse on the content stratum.


In this way, the (physical) act of speaking - vocal-tract activity - is shown to be asingle complex system of factors interacting in a non-linear fashion across all ofthe levels involved.

10. The Integration of Individuals to their Semiotic Trajectories

A social system, in the theoretical perspective of the present study, is lessspecified than an ecosocial semiotic system: the former exists at a lower integrative level, A social system is less specified than an ecosocial semiotic systembecause the latter can access symbolic, and notjust indexical, modes of semiosis,The historical emergence of the integration hierarchy of icon, index, andsymbol in human social meaning-making could only come about with the development of new forms of social life and therefore new ways in which humansrelate to each other and organize their social relations on different space-timescales. Thus, semiotic evolution emerged as part of the individuation ofdeveloping social life. See Bogdan (2000: 40-1) for some relevant observationson the socialized minds of nonhuman primates, based on indexical modes ofsocial interaction, as an evolutionary prelude to the 'unsituated' (symbolic)minds of adult humans.

Beginning with Edelman's view that conceptual repertoires are recategorizations of sensori-motor ones, we can say that these two levels were present in thebrains of our hominoid ancestors well before social semiosis as we know it todaycame into existence. Conceptual repertoires and sensori-motor repertoires maybe described as proto-content and proto-expression, respectively. Moreover,these two levels were functional at the interface between body-brain and environment. In turn, forms of social life emerged between these two less specifiedlevels. This development required the co-ordination of human activities onscales that went beyond the body-brain's pick-up and sampling of informationfrom its immediate physical and interpersonal environments. Body-brain andenvironment may thus be seen as lower- and higher-scalar levels in a semioticdifferentiation model of the eo-evolution of body-brain and ecosocial semioticrelations and processes. The fact that recategorization entails more fine-detailedspecification of the sensori-motor level means that the degree of specificationand differentiation of this level is amplified, giving rise to the emergence of thephonetic and phonological levels. Parallel processes simultaneously operate withrespect to conceptual repertoires. Consequently, lexicogrammatical anddiscourse structures and systems emerge in response to an expandingconceptual repertoire along with the need to pool and access the expandingpool of conceptual-semantic categorizations in socially co-ordinated ways.Ecosocial semiotic systems may be viewed in this perspective as dissipativestructures regulated and maintained in a constant me tastable dialectic ofstability and change (Lemke 1995b) by processes of meaning-making along thelogogenetic trajectories of the members of that system.

Individuals are connected to each other in social semiosis through their logogenetic trajectories. Logogenetic processes of meaning-making can be seen asjust such a process of trajectory connecting. An example is when two individuals


jointly construct some conversational event. In this way, individual trajectoriesintersect and, at the same time, a new, perhaps transient, higher-scalar trajectorycomes into view in the form of the jointly constructed and enacted meaningmaking event and its entextualized records and/or products. The individuationof social semiosis along an ontogenetic trajectory means that the individualtrajectory's meaning-making potential expands as it individuates. Thus, thetransition from proto-language (iconic-indexical) to language (iconic-indexicalsymbolic) means that the individual's agentive powers increase: he or she is ableto access and relate to an expanded range of meaning-making possibilities. Theintegration of, say, the microfunctions of the indexical phase to the metafunctions of the symbolic phase entails an increase in the amount of stored information along the individual's trajectory (Salthe 1993: 262; chapter 4, section 17).This has clear enough implications for human consciousness, as we shall see inchapter 5.

Symbolic meaning-making in particular and its associated practices of entextualization mean that core consciousness and extended consciousness paceDamasio (1999) are extended and integrated by symbolic consciousness. Thus,individual agents can construct trajectories across many different space-timescales beyond the here-now one of the body-brain in its immediate physical environment. This helps us to think of the emergence of symbolic meaning-makingfrom indexical meaning-making, along with the integration of the latter by theformer, along an individuating historical-biographical trajectory as an increase inthe individual's agentive powers or semiotic causality. This goes hand-in-handwith an increase in the system's stored information, as illustrated by Salthe(1993: 262). As I said earlier, it is this increase along a self-organizing individuating trajectory which enables individual organisms qua dissipative structures torelate to other dissipative structures, human and otherwise. Now, 'relate to' hasbeen by and large formulated in conceptual, cognitive or ideational terms - callit what you will - such that the accessing of systems of symbolic possibilities inlanguage, depiction, mathematics, and so on, is understood as the capacity of thesystem to store and to relate to more and more conceptual categories in theprocess of expanding its Umwelt. The transition from an inward-looking dyadicperspective of interpersonal transaction with the mother to a concern for theworld outside the mother-infant dyad, along with the concomitant emergenceof experiential systems of naming and classification, illustrate this move. Inactual fact, the engagement with phenomena outside this dyad is a qualitativejump in the notion of 'relate to', entailing new ways of constructing one's beingin the world.

At the same time, Salthe's notion of integrative levels reminds us that later(outer) levels always integrate earlier (inner) ones. That is, the symbolicintegrates the indexical, which integrates the iconic. Higher-order or symbolicconsciousness integrates primary consciousness, which integrates the topologicalmaterial ground of our being in and relating to the world. The latter notion isakin to Damasio's proto-self, which is immersed in the topological-continuousflux of our material (not semiotic) interactivity with the material world. This isessentially how we begin our being in the world. Moreover, it is the basis onwhich a person's self-organizing individuating trajectory begins to relate to the


surrounding world, including the first primordial differentiation of self fromnon-self. This initial differentiation is topological, not typological. Nevertheless,it constitutes the ground for later integrative levels whereby interpersonalrelations are constructed between increasingly differentiated selves and others,in an expanding diversity of social contexts, including those afforded by theinterpersonal resources of language and other semiotic modalities. Figure 3.10provides a schematized model of this process.

The topological ground ofour initial material interactivity with the world constitutes the vaguest, least differentiated basis for our understanding of and ourestablishing relations with others. It is a very wide and shallow attractor basin ofvague possibilities which was our originary primary being in the world - 'being'before 'categorizing' and 'relating to', ete. differentiated out. It is the primordialbasis of our capacity to enter into relations of empathy with others on the basisof feeling states of the body (Damasio 1999: 279-95). This least specifiedintegrative level of initial at-oneness with the world is, however, integrated to theexpansion of information-meaning at more specified levels. It gives rise to anincreasing 'potential scope of sympathy riding on that same growth of information' (Salthe 1993: 263). Thus, the earliest experiences that the infant has ofhis/her embodied being in the world and the socio-affective exchanges characteristic of primary intersubjectivity in early mother-infant dyads are the deeproots of our being and as such are shared by others.

Salthe's 'potential scope of sympathy' refers to the topological-eontinuous oranalogic character of our most primordial forms of communication. Such communication is concerned exclusively with patterns of relationship, e.g. betweenorganism and environment and/or between organism and others (Bateson1973c [1972]: 340-4; Wilden 1980 [1972]: 164). Analogic communication may,therefore, be considered as a kind of Ur-interpersonal function from which allsubsequent functions are logically derived. As the individual's trajectory expandsinto more specified levels of meaning-making on the indexical and symboliclevels, it also integrates the topological material ground of its primordial iconicbeing in the world in order to attain an understanding of how to engage in andmaintain more specified (and specific) interpersonal relations with anincreasing diversity of flesh-and-blood others. The increasing tendency inadvanced capitalism to commodify and sell these same properties as 'syntheticinterpersonal relations' in the market economy reveals the extent of themarket's categorial reach - right back to the very topological material ground ofour being in the world.

11. The Intrinsically Time-bound Nature of Semiosis: The Integration ofTemporal Dynamics Across Scales of Expression and Content

The tripartite perspective on language as particle, wave, and field allows us toclarify the nature of language as a dynamical system. The complementarity ofthese three perspectives illustrates the time-bound nature of language. As adynamical system, language undergoes a constantly changing series of statesin relation to its ecosocial and bodily environments. The wave/particle/field


Stage 4: microfunctions(indexical)

o

Stage 2: feeling states; socioaffective dyads

o

Stage 5: metafunctions (symbolic)

Stage 3: conceptualizingcategories and relating to

Stage 1: being

Figure 3.10: The integration of initial at-oneness with the world to the expansion ofinformation-meaning at more specified levels of semiosis along an individuatingtrajectory

perspective is a means of modelling the way in which the various dimensions ofthe system change in time. In logogenesis, the behaviour of the system as itunfolds in time is the changes it undergoes as the system changes from one stateto another. At any given moment, the system is in a number of different states,corresponding to different overlapping metafunctional regions. The overallnumber of states in which the system is in at any given moment defines its statespace. This means that the dynamical process of meaning-making can be seen as


simultaneous configuration of wave/particle/field in its state space. The diversemetafunctional regions interact with each other such that dynamical changes inone region affect and depend on the other regions.

Furthermore, the intrinsic metafunctional organization of language hasarisen in order that language may be selectively co-contextualized withphenomena in its ecosocial and bodily environments. Such phenomena are notexternal to the relevant system of relations, but are a part of that same system'sintrinsic organization. It follows that ongoing, time-bound changes in thecontextualizing dynamics that are in operation between unfolding text and its'external' environment entails changes in the internal configuring of therelations among the diverse metafunctional components which are at play at anygiven moment in some meaning-making event. This holds for dynamicalprocesses on both the expression and content strata. The evolution of themeaning-making act in time corresponds to a series of points and periodicitieswhich define the trajectory of the system in its phase space. Expression andcontent are coupled such that dynamical relations and processes on one stratummutually influence relations and processes on the other stratum. In turn, eachof these strata is coupled to the body and to the ecosocial environment. Themeaning-making body-brain does not interact with the external world by meansof symbolic inputs and outputs. Rather, expression and content are dynamicalsubsystems of a larger whole. The former is based on neuromuscular activity; thelatter on the recategorization of this in the CNS as semantic-conceptual content.

In the computationalist model of cognition, on the other hand, the brain is acentral control unit which is housed in a body, which is, in turn, located in anexternal world. The brain interacts with the outside world via the body's directcontact with the outside world. Thus, sensori-motor transducers 'translate' thedata received from the outside world and the body, as well as the symbolic stateswhich are the product of the brain's cognitive activity. This means thatperceptual organs convert physical stimulation into symbolic representations ofevents in the body and the external world. The motor system converts the brain'ssymbolic representations of events into muscular movements. Cognition is acontinuously cyclic and sequential process. First, sense data is received by somecognitive system; secondly, the given cognitive system algorithmically manipulates symbols so as to produce an output; thirdly, this output causes bodilymovement. The cognitive system is specialized to symbolic representations. Assuch, it is seen as separate from the body and the environment. Cognition isdisembodied and de-contextualized.

On the other hand, time is a critical factor in the dynamical approach.Meaning-making occurs in time. Therefore, time itself must be seen as a centralcomponent in the explanation of meaning-making. This goes hand in hand withthe view that all aspects of the relevant system are simultaneously changing intime. In logogenesis, it is the overall state of the system of relations which ischanging from one moment to the next. Meaning-making as it unfolds in timeexhibits properties of both continuity and discreteness. Moreover, diverse metafunctional regions are simultaneously in operation as well as interacting witheach other within the phase space of the system. The distinctive principles ofstructuration which characterize the different metafunetions are properties


which emerge over the time-bound trajectory of the system; they are not staticgivens. Moreover, meaning-making processes occur on many different timescalessuch that events on different timescales interact at the same time as these areembedded in a brain-body complex and in an ecosocial environment.

Real-time is a topological-continuous flow. It is possible to plot a potentiallyinfinite number of points along this flow. Each point corresponds to a state ofthe system. The neuromuscular activity that takes place in articulation is aprocess which unfolds along a temporal trajectory. Nevertheless, at each of thepoints in time from the beginning to the end of the activity there is a positionwhich the articulators occupy in space. In section 2, pp. 109-16, the phonemewas shown to be a point in phase space in this sense. The continuous topological processes of articulation can be broken down into finer and finer points suchthat there is a state of the system at every point. This fact is a consequence of thebio-physical nature of the processes that take place in the vocal tract.

The timing of events on diverse scalar levels in the unfolding meaning-makingevent is, therefore, of central importance in understanding the nature and functioning of such events. For example, the ranked scale of phonological units andrelations, viz. syllable, foot, and tone group, entail different scalar levels oforganization, each with its different rates of completion, its different periodicitycycles, and the different ways in which one scalar level temporally integrates withother scalar levels. As I pointed out above, vocal-tract activity occurs along atemporal trajectory in which the system of relations is required to spend acertain period of time in the phase space that is relevant to any given scalar level.Moreover, the timing of events on diverse scales in vocal-tract articulatory activityis cross-coupled with and integrated to dynamical processes on still largertimescales such as those implicated in the content stratum, along with still widerecosocial processes and relations. The rank scale of both phonology and lexicogrammar should have much to tell us about the temporal nature of meaningmaking on the different scalar levels which constitute such events. Changes invariables on the content stratum depend on and are cross-eoupled to changes onthe expression stratum. (See for example Martinec (1998) on the ways in whichrhythmic waves of expression correlate with rhythmic waves of content.)

Linguistic semiosis is best seen as being of mixed-mode character (Lemke2000a; Thibault 2003b). This means that language has both typological andtopological properties and relations. Change of state in the meaning-makingprocess can be either continuous or discrete. It is essential that the facts ofcontinuity and discreteness can be accounted for in a unified framework.Change from one discrete state to another and continuity of state (not time) areboth essential characteristics of linguistic processes. For example, the very fineneuromuscular activity involved in articulation takes place in complex andchanging environmental conditions with respect to factors such as localization ofinterlocutors, distance between them, movement towards or away from eachother, the blending of different sound sources, physical features of the environment on different scales, and so on. The terrestrial scale on which speech soundsare produced and received is that of metres rather than kilometres or millimetres unless amplified by technological means. This scale is that of the physicalenvironment which humans typically inhabit.


In the first instance, the wave train of airborne compression waves producedby the vocal tract is specific to the source of a mechanical disturbance caused bythe neuromuscular activity of the vocal tract. The resulting wave train 'is specificto the kind of mechanical event at the source of the field' (Gibson 1983 [1966]:16). The field of speech-sound waves specifies information about its physicalsource, i.e. the articulator. Such information constitutes a perceptual stimuluswhich is causally related to the environmental event which produced it. Information of this kind is iconic and/or indexical in character. On the other hand,the symbolic information which is specified in the sound event is not causallytied to its source in the same way. Instead, it depends on symbolic meaningmaking resources and their associated conventions existing on many differentspace-time scales that go beyond the causal linking of a sound event to itsphysical source. The neuromuscular substrate of vocal-tract activity must beflexible and able to adjust to many different environmental conditions. In otherwords, its own activity must be capable of flowing into and occupying a richlyvaried continuum of changing environmental possibilities on diverse semioticand material scales.

Table3.3: The metafunctional organization of bothphonology and lexicogrammar ontheexpression and contentstrata illustrating the operation of mixed-mode semiosis,combiningboth discrete typological-categorial distinctions and continuous topologicalvariation on both strata

Phonology Lexicogrammar

Difference of Difference of type: Difference of Differencedegree: type-X degree: of type:quantity-X quantity-X type-X

Experiential Degree of Particulate analysis of Degree of Particulateconformity to consonants and conformity to analysis oftype-category; vowels as correspon- type-category; phenomenafuzzy sets; ding to discrete fuzzy sets; into partsphonological phonological type- semantic and wholestopology; categories (e.g. topology; correspon-smearing of phonemes) as indeterminacy ding toprobability specific paradigmatic of complemen- discretedistributions states corresponding tary categories

to discrete articula- transitive and oftory categories seen ergative per- experience,as ensemble of dis- spectives e.g.tinctive features and participant-part-whole combina- processtions of these relationsorganized as aranked hierarchy ofphonologicalparticles


Phonology Lexicogrammar

Inter- Field-like Tone system in tone Field-like Mood,personal prosodic and group comprising a prosodic and modality,

scopal primary system of scopal lexis asrelations; five contrasting relations; categories ofcumulative, simple tones and two cumulative, (inter) actionamplifying compound tones amplifying and orienta-effects; effects; tiondegree of degree ofdeformation, deformatione.g. pitch shift

textual Wave-like peaks Tonality and tonicity Wave-like peaks Theme,of prominence in tone group; of prominence rheme;defining rhythm in foot as defining given-newboundaries; indexes oflexi- boundaries;begin-end cogrammar elements begin-endstructures; tran- as units of informa- structures;sitions; tion; transitions;culminative; demarcation of word culminativedynamic peri- boundaries;odicities of word emphasis;tonicity, stress, contrast betweenand moraicity +stress/-stress;seen as waves +moraicity/-of continuously moraicity; +tonicity/-varying tonicityamplitudefunctioning asextensionprosodies ofvariable lengthto integrate orconcatenatelexicogrammat-ical domains

Likewise, the experiential resources of the clause, say, must be able to modelthe constant flux of the phenomena of experience as instantiations, to varyingdegrees, of the categories which are intrinsic to a particular language system.The tripartite perspective on lexicogrammatical form as particle, wave, and fieldemphasizes the dynamical character of the system. Instead of discrete, staticcategories, the language system must be able to adapt to the continuousvariation of the flux of human experience. The mixed-mode character oflanguage shows that language embodies both discrete, categorial distinctionsand continuous topological change and variation. A better understanding oflanguage and its workings is obtained once it is understood that both continuityand discreteness are intrinsic to its dynamics. Table 3.3 sets out the properties ofthe expression and content strata of language in order to illustrate the mixedmode character of language.


12. Scalar Heterogeneity and the Phonological and LexicogrammaticalRank Scales

The existence of a rank scale on both the expression and content strata can beunderstood in relation to the diverse timescales which are simultaneouslyimplicated in the meaning-making process. These timescales range from themilliseconds of neuronal processes, the fractions of a second in perceptualprocesses, the seconds of co-ordinated movement, the minutes or hours of anunfolding textual performance (e.g. a conversation), the months and yearsinvolved in learning the phonological and grammatical systems of a language,the entire lifespan trajectory of the individual's movement through a givenecosocial system, the history of an entire ecosocial system. Diverse timescales arenot sealed off from each other. Instead, they are interrelated and integrated toeach other such that the dynamical processes on one timescale affect those onother timescales.

For example, the timescale involved in the articulation of syllables creates thepatterns of strong and weak syllables which allow for the emergence of thetimescale of rhythmic patterning in the foot. In turn, the patterning of Ictus andRemiss in the foot allows for the emergence of a still larger timescale at the rankof the tone group. Variations in loudness and stress produce rhythmicpatterning. Rhythm has the function, in English, of making content wordsprominent; it also creates the potential for tonicity and tone on the level of thehigher-ranking tone group (Halliday 1994 [1985]: 10; Cleirigh 1998: 12). Thetone group is the only phonological unit in the rank scale in English whichrealizes a unit on the content stratum, viz. the information unit (Halliday 1994[1985]: 295). In this way, phonological form is integrated with higher-scalar unitsof lexicogrammar and discourse on the content stratum.

12.1 The Lexicogrammatical Rnnk ScaleThe lexicogrammar of English is comprised ofa rank scale ranging from highestranking to lowest-ranking units, according to their constituency potential.Systemic-functional linguistic theory postulates the following ranks for English:clause, group/phrase, word, and morpheme. The number of ranks and the waysin which various systems are distributed across ranks differ from language tolanguage. However, such details need not concern us here. The clause is thehighest rank. It is not a constituent of some higher-ranking unit. According tosystemic theory, the relations between clauses in discourse are semantic innature, rather than lexicogrammatical. This is so because combinations ofclauses constitute semantic relations at the level of text. The clause constituteswhat Halliday has called the 'gateway' to text (see Thibault 1987: 611). On theother hand, clauses are made up of constituents from lower-ranking systems atgroup/phrase level. In turn, groups and phrases are made up of words; words ofmorphemes.

The lexicogrammatical rank scale serves to integrate smaller-scalar lexicogrammatical units into larger-scalar ones. For example, the clause functionallyintegrates lower-ranking nominal, verbal, and adverbial groups into thefunctional structure of the clause. Rather than assuming that lower ranks are


simply the smaller-scalar building blocks out of which higher-ranking units areassembled, it is no less relevant to see how higher ranks provide integrativecontexts for units on lower ranks. In this perspective, we see more clearly howlexicogrammatical units at all levels are integrated to and have their function inthe making of text in its context of situation. For example, nominal groups areintegrated into clauses as various categories of participant roles. By the sametoken, the deictic and instantiating functions of the nominal group qua grouprank have grounding functions whereby the Thing which is realized by the Headnoun is instantiated as an instance of that Thing by grammatical systems andfunctions operating at nominal group rank.

The lexicogrammatical rank scale serves to integrate lower-ranking units tohigher-ranking ones according to a scale of indexical-symbolic integration, asfollows.

Rank: MorphemeIn English, morphemes specify semantic information in the form of semanticdictionary entries whereby a given item can be integrated to a particular worddistribution class at the next-highest rank. For example, consider the twomorphemes in the word buyer. This item comprises the two morphemes [[buy] +[er]]. The two morphemes may be characterized as follows:

Morpheme: buy> verb tr, [process: material action: disposal: acquisition:transfer of Medium Thing to Agent on payment of money; non-iterative;beneficile: potential: +benefactive]e.g. he boughtJohn a new boat

Medium [Thing: alienable: transferable; material/immaterial]Morpheme: -er> nominalizing suffix [INSTIGATOR: AGENT;INSTIGATOR: INSTRUMENT],subcategorizes Thing in derived verbal noun as Agent responsible for theinstigation of an action [Instigator: Agent] or for the instigation of a process[Instigator: Instrument]e.g. teacher, buyer, voter, ruler, lawn mower, bread slicer

Rank: WordWords specify type-specifications of Things, Processes, Qualities, and so onaccording to word-distribution class. In this respect, the semantic informationthat is integrated by the two-morpheme combination [[buy] + [er]] tells us thatthe resulting word belongs to the word-class noun, symbolizing a semantic typespecification of Thing, viz. [THING: HUMAN: PERSON/INSTITUTIONAL;AGENT: ACQUIRER OF GOODS/SERVICES IN ECONOMIC TRANSACTIONON PAYMENT OF MONEY].

Rank: Group

A type-specification of a Thing (or Process, Quality, and so on) is not tied tospecific instances of the type. The common noun buyer is simply the name of asemantic class of Thing, as shown above. The resources of the nominal group


enable the type-specification to be instantiated as a specific instance of the giventype-eategory. Nominal groups, as Langacker (1991: chapter 2) has shown, allowspecific things to be specified in some discourse context according to fourparameters. The four parameters are as follows with respect to the nominalgroup those two enthusiastic buyers:

1. type specification, expressed by the Thing element in the nominal groupand non-determining modifiers: enthusiastic buyers;

2. instantiation, expressed by determiners: those two;3. quantification, expressed by quantifying items: two;4. grounding, expressed by articles, demonstratives, and possessives: those.

Rank: Clause: MajorAt clause rank, nominal and verbal groups form configurations of functionallyrelated items which experientially construe a given scene or situation in terms ofa number of interrelated transitivity roles. These roles consist of differentcategories of processes and participants according to the type of process in theclause, e.g. participants such as Actor, Agent, Medium, Senser, Sayer, andprocesses such as Material: Action, Mental: Perception, Verbal, and so on. Forexample, the nominal group those two enthusiastic buyers has the potential torealize a participant role in a clause, e.g. those two enthusiastic buyers went to theauction. In this clause, the given situation is experientially construed as a configuration of the following transitivity functions: [ActorI\Material: Action/-Circumstance: Location].

The clause is the rank where all four metafunctions come together to realizea semantic unit, called the message unit by Hasan (1996a: 119), which functionsin discourse. In contrast to message units realized by minor clauses such as hello,good day, damn, and so on, a message unit which is realized by a major clausebrings together in the one structural output the full array of metafunctionalmeanings.

Minor clauses, on the other hand, are primarily interpersonal in function inthe sense that they may function as both initiating and responding moves indiscourse as well as realizing attitudinal/affective meanings. However, they donot exhibit clearly definable transitivity roles whereby an experiential situation,seen as comprising particular classes of process and participants, is construed.Moreover, they do not have the feature of finiteness, which is means by which theclause is assigned a point of reference in the here-and-now of the speech eventthrough the resources of primary tense and modality. Minor clauses do not fullyparticipate in all of the systems which operate in the lexicogrammatical rankscale. For this reason, they are, in some respects, not unlike child protolanguagein the sense that their status is primarily interpersonal-indexical (see Hasan1996a: 118-19; Thibault 1992a). Hasan accordingly assigns the feature [punctuative] to minor clauses. On the other hand, Hasan assigns the feature [progressive]to major clauses because the full range of metafunctional options which theseclauses realize means that they participate fully in the development of discourseand its entextualized products and records. It is in this sense that Halliday hasreferred to the clause as the 'gate-way' to text; it is the means whereby systems and


structures which operate on different ranks are integrated to and realize asemantic message unit which can contribute to the development of discourse. Aswe shall now see, it is discourse constructed in this way that exhibits to the fullthe properties of scalar heterogeneity.

Let us reconsider our earlier example from this point of view. The morphemesuffix -er classifies the root morpheme, comprising the de-verbal process nounbuy, as the Agent/Instrument of the process named by the root; the word buyersymbolically names a type-category of Thing, viz. the agent or instrumentalitywhich instigates the process buy; the nominal group those two enthusiastic buyersspecifies an instance of the semantic category of Thing with reference to a givendiscourse context; in turn, the clause subcategorizes this instance of the Thingas instantiating the participant role Actor in the transitivity structure of theexperiential scene realized by the clause. In combination with the other metafunctions at play in the clause, this experientially construed scene has thepotential to participate in and contribute to the development of some discourseabout persons, actions, and locations which exist on some very different spacetime scale with respect to the moment-to-moment unfolding of the discourseitself. It is the clause which integrates lower-ranking units to discourse.Morphemes and words do not as such directly realize discourse meanings. Rather,they are mediated by the systems and structures operating on the higher ranks ofgroup and clause, which integrate them to the discourse level of organization.This is no less the case for a clause realized by a single morpheme, e.g. Go!

The preceding discussion shows how, at any given rank, the system changesstate according to the variables in operation at that particular rank. Statechanges at a given rank occur in relation to the more stable parametersoperating at the next-highest rank. In the case of the expression stratum, wherephonological systems and structures occur, these variables have to do with theneuromuscular dynamics of vocal-tract activity. For example, the foot, in relationto the lower-ranking syllable, establishes a set ofparameters on a slower timescalewhich influence and regulate changes in the state variables on the fastertimescale of the syllable. In this way, slower timescales on higher ranks modulatethe temporal dynamics of faster timescales on lower ranks. By the same token,the faster dynamics of the lower-ranking unit can also affect the slower dynamicsof the higher unit.

On the content stratum, the lexicogrammatical rank scale is not directlyimplicated in body dynamics in the same way, but in ecosocial ones. In this case,the integration of lower-ranking units to higher-ranking ones means that thecategorial reach of lower-ranking units is extended by their functional integration to higher units. This was shown above in the discussion of the way in whichthe integration of morpheme to word and word to group, and so on, entails anextension of the contextual reach of a given item. If, for example, language werelimited to lower-ranking morphemes and words per se, its contextual reach wouldbe much more restricted for the reasons already mentioned above in relation tominor clauses and protolanguage. The lexicogrammatical rank scale is,therefore, a resource which extends the contextual reach of language such thatdiverse ecosocial space-time scales are integrated by the interplay of indexicaland symbolic factors which the multiple metafunctional development of


discourse, as afforded by the clause, makes possible. The criterion of stabilitymentioned in the previous paragraph applies to the lexicogrammatical rankscale as well. For example, a morpheme as such symbolizes some semantic unit,however small, such as the meaning [Agent/Instrument] discussed above inrelation to the nominalizing suffix -er. However, it is the contextual integrationand mediation of this element by higher-ranking units which provide the morestable semiotic environment enabling its grounding in a specific discoursecontext. Table 3.4 presents a three-level hierarchy view of the organization of thecontent stratum of language.

Table 3.4: The content stratum of language in relation to the three-levelhierarchy

Scalar level Difference

L+l System of Discrete Symbolic Semantically salient dis-semantic differ- typological- tinctions symbolicallyences as SI categorial construe experience

distinctions;difference oftype

L Lexicogrammat- Indexical Grounding of clause byical rank scale Finite element in verbalcomprising group in terms ofhierarchy of temporal/modalranked units: proximity

(tense/modality) and in

clause; terms of person deixis

group/phrase; (Subject); grounding of

word; nominal group in terms

morpheme of spatial proximity(demonstratives) orperson deixis (posses-sives); both give theprocess and the partici-pant(s) in the clause apoint of reference in theunfolding (con) text bymeans of intra- and extra-textual indexical signs

L-l Continuous Topological- Iconic Particulate, field-like,quantitative continuous wave-like, andvariation in the variation; recursive modes ofphenomena of difference of lexicogrammaticalexperience degree realization iconic tomapped onto ways in which diversecontinuous metafunctionsquantitative semantically organizevariation in experiencesensori-motorand CNS activity

BODY DYNAMICS, MEANING-MAKING, 159

12.2 The Phonological Rank ScaleVocal-tract gestural activity has both spatial and temporal organization (seesection 2). The creation of emergent spatia-temporal organization occurs alonga particular trajectory. Vocal-tract activity has its existence only along a temporaltrajectory. Thus, the syllable is a unit of articulation which cannot be defined interms of discrete and static segments or positions in the overall flow of articulatory events. The dynamics of articulation are extended along a temporaltrajectory and can only be defined as such. This means that the syllable is, on itsparticular space-time scale, a dissipative structure which is entrained by itstrajectory (Salthe 1993: 260). Furthermore, the existence of a phonological rankscale shows that phonological trajectories exist at a number of different spatiotemporal scalar levels. As a dissipative structure, the reorganization of micra-levelphonetic processes as macro-level phonological ones means that the lower-scalaror microscopic phonetic processes of articulation are entrained into the macroscopic organization of phonological relations and processes. This requires aninput of energy along with specific dynamical laws so that the macroscopic orderemerges. An articulatory trajectory entails that external entropy increases as thesecond law is maintained. By the same token, the emergence of macroscopicphonological order means that the entropy production of the structure affordedby the trajectory is reduced. It is this fact which enables local sinks of lowdimensional or macroscopic order - i.e. phonological form in the present case to appear without violating the second law. Organization emerges when theproperties and dynamics of the oral cavity qua physical entity, whereby vocal-tractproperties are unco-ordinated, to a state in which phonetic properties are coordinated in a stable regime such that the activity of a given feature depends on theactivity of other features with which it is co-ordinated. In this way, each microscopic phonetic feature is entrained to the principles of organization of thehigher-order articulatory trajectory of the syllable.

The neuromuscular activity involved in articulation is thus entrained by trajectories existing at a number of different scales such as those suggested by thephonological rank scale. I would suggest that the notion of trajectory is the fundamental entity in meaning-making. The neuromuscular activity of articulationis a dissipative structure which functions as a material sign of the existence of ourand other's meaning-making trajectories (Salthe 1993: 260). It is what providesthe basis for their material connectivity. What brings about the emergence ofsuch trajectories? What enables their form and their dynamics to endure overextended periods of time? If meaning has properties of organization in bothspace and time such that sustained meaningful interaction with others occurs,then this suggests that the control parameter which organizes the lower-levelneuromuscular activity of the oral cavity as an ordered articulatory trajectoryshowing properties of macroscopic phonological organization is meaningemanating from higher-scalar levels. Thus, lower-level oral-cavity propertiesbehave differently when entrained to the ecosocial semiotic dynamics of ahigher-order logogenetic trajectory than they would otherwise do. Meaning, aswe have seen, exists in symbolic neural space in the context of the body and thehigher-scalar ecosocial environment. The increasing discrimination or differentiation which is afforded by the reorganization of microscopic phonetic


properties as macroscopic phonological ones means that the system organizesitself into a new hierarchy of levels such that new functional components comeinto being. The emergence of phonological order as the reorganization ofsensori-motor routines into more finely categorized gestural systems leads totheir further recategorization as conceptual and lexicogrammatical systems.

According to Edelman (1992: 100-1, 120-1), conceptual routines are basedon the interaction between perceptual activity and value systems. This shows thatconceptual categories are mediated by information about the external andinternal environments of the organism. Thus, the self-organization of the macroscopic phonological level on the expression stratum means that vocal-tractactivity is now dependent on the synergy among different component parts ofthat activity. Furthermore, this activity is itself enslaved to higher-order principlesof organization emanating from the content stratum. Therefore, the expressionstratum is now seen as being a means whereby the organism systematicallyextends itself into and is cross-coupled with its ecosocial environment. Thus, therecategorization of sensori-motor routines as conceptual ones and the furtherrecategorization of these as semiotic categories of expression and content,respectively, provides a means of explaining what theorists of dynamical systemsrefer to as the 'embeddedness' of the cognitive system in a body and an environment. The processes of semiotic reorganization across hierarchical levels thatI have outlined here show how the interaction of organism and environment isa supersystem in which the exchanges of matter, energy and informationmeaning constrain the lower-level subsystems (e.g. individual organisms). In thisway, both sensori-motor and conceptual routines are reorganized as semioticexpression and content such that the ecosocial level is imported into theorganism's body-brain dynamics. The entrainment of the organism's intrinsicdynamics to the ecosocial scalar level amounts to an expansion of the system'sphase space by increasing the meaning-making potential and, hence, thesemiotic freedom of the organism qua individual agent.

The question of the relation of a meaning-making to its environment can beunderstood in the light of the notion of trajectory introduced earlier. The flowof meaning from an intentional source in symbolic neural space along anunfolding action trajectory is a temporal process. Thus, the flow of meaning,understood here as semiotic content, produces and entrains bodily changesalong its trajectory. Moreover, such bodily changes are themselves coupled tocontextual ones. Thus, the dynamic temporal unfolding of articulatory gesturesoccurs at the same time as dynamical processes of neural activity on lower scalesand dynamical environmental processes on higher scales unfold in time.

As we saw in section 2, the phonological rank scale specifies a hierarchy ofcycles of vocal-tract activity (Cleirigh 1998: 38-9). The first cycle is the process ofopening and closing the vocal tract by articulation to dampen phonation in theproduction of syllables. This can be modelled as a binary oscillation between+ syllable weight and - syllable weight, i.e. between moraic and non-moraicsyllable constituents (Cleirigh 1998: 38). The second cycle refers to the varyingof properties such as loudness and duration so as to add stress. This process istermed rhythm and can be seen as 'switching stress on and off' (Cleirigh 1998:39). The third vocal-tract cycle is the applying of major pitch shifts to create tonic


prominence or tonicity. This process is intonation and consists in switchingtonicity on or off. Each of these cycles occurs at a given rank in the phonological rank scale. Thus, the syllable is defined in terms of its relations to lower-levelsegments as well as to the higher-ranking unit, foot. In turn, the syllable isintegrated to the level of the foot in the sense that it receives further levels ofcontextualization by being integrated to a higher-order unit. The foot itself isintegrated to the still-higher-ranking unit, the tone unit.

Each rank specifies a scalar level with its own temporal dynamics - faster onlower scales, slower on higher ones. Each rank in the scale may be seen as a levelof specification integrated by the dissipative structure - the organism - thatmaterially embodies the trajectory. That is, neuromuscular processes are continually reintegrated under the increasing specification of phonological structure.It is this integration to increasingly higher levels of specification that providesthe basis for the trajectory's coherence over time (Salthe 1993: 261). Theemergence of macroscopic phonological organization in the form of a rankingscale of articulatory, rhythmic, and tonicity cycles of vocal-tract activity frommicroscopic phonetic processes illustrates the principle whereby the lower-levelphysical embodiment of the trajectory self-organizes. What we have here is anillustration of the way in which articulatory activity qua bodily movement isorganized on a diversity of timescales. Moreover, this reveals a fundamentalaspect of embodied meaning-making: temporal body dynamics on differentscales face 'two ways', viz. inwards to the faster temporal cycles of neural activityin the CNS and outwards to the slower temporal cycles of the other systems - theindividuals - with which a given system interacts. Meaning-making unfolds intime and it is crucial that there are resources for attuning to the temporaldynamics of those with whom one interacts.

Thus, the physical organism qua dissipative structure is semiotically connectedto other organisms by virtue of their trajectories (Salthe 1993: 262). As we sawabove, the organism is semiotically integrated to a trajectory in the first instanceunder the increasing specification of the expression stratum. The integration ofneuromuscular potential to macroscopic phonological organization connectsthe organism to others' meaning-making trajectories on the basis of potentiallyshareable articulatory, rhythmic, and tonic resources such that it can functionallyrelate to other organisms who make and recognize the same phonological discriminations. Compare this to the newborn child whose vocal gestures are, in thefirst instance, iconic of immediate body states and feelings in the mother-infantdyad. The infant can only access the lower integrative levels of topologicalcontinuous variation in voice dynamics; he or she does not yet have access to thestored phonological information deriving from higher levels. To be sure, theinfant can both produce and respond to articulatory, rhythmic and pitch discriminations, but he or she can do so only on the basis of the lower integrativelevel of their iconic significance. The expansion in proto-language and laterlanguage of vocal activity into higher integrative levels where the indexical andsymbolic distinctions of phonology are made thus entails an expansion of thechild's bodily potential to relate to the semiotic trajectories of others.

A parallel observation can be made with reference to the content stratum.Before the onset of a lexicogrammar, the child's semiotic content is restricted to


the perceptual purview of the mother--ehild dyad and its immediate surroundings. In a sense, there is discourse without lexicogrammar. It is a discourse levelof semantic content which is integrated to its contextual environment on thebasis of indexical ties for the most part.

13. The Emergence of Linguistic Categories from the Child's PrimaryForceful Interactions with its Environment

Seen as a physiological space, the oral cavity is a neuromuscular potential forachieving specific goals. In order to perform a given articulatory gesture, theinfant must adjust the parameter values of the oral cavity in order to achieve aparticular articulatory gesture (Browman and Goldstein 1995: 181). Forexample, a particular gesture may involve a constriction in the distance betweenthe upper and lower lips combined with a stiffness setting and damping, whichis determined by the amount of time required for the system to approach thegoal of lip closure. This process requires the deployment of spurts of neuralenergy at precisely the right moment and to the right degree in order to activatethe muscles required to perform the gesture. Children learn to do so by experiencing the many different values of the oral cavity as produced in their spontaneous babbling and other vocalizations produced in early infancy in order toattain a specific goal (Thelen 1995: 80). The process of developing articulatorygestures involves exploring the range of parameter values in the state space ofthe oral cavity and selecting those values that selectively co-contextualize withrelevant environmental affordances and the goals and needs of the child.

Halliday's description of Nigel's use of very high-pitched squeaks at age 0;6 months (see also chapter 4, section 10, pp. 202-5) in order to draw attentionto the scattering of the pigeons is an example of how the system learns throughaction. The child's vocal gesture - the high-pitched squeak - is an iconicindexical act which points to some phenomenon (the pigeons) within the child'sperceptual purview. The vocal gesture is iconic in the sense that it can be interpreted as expressing bodily and feeling states of its utterer; it is indexical in thesense that it both indexes (1) its utterer as the source of the squeak qua acousticevent in the given environment; and (2) the pigeons as the object of interest orattention in the situation. The pigeons constitute an environmental affordancein the sense that the perceptual information which the child picks up about thisenvironmental event affords a contextual value, as determined by the indexicalact. This micro-semiotic act lasts mere seconds in the local context in which itoccurs. However, the cumulative effect, over time, of many such micro-semioticacts cascades into the developmental landscape, as Halliday documents. There isanother sense in which the real-time of the action is implicated in or interwovenwith developmental time. In the time required for the performance of a givenarticulatory gesture, values such as stiffness and damping ratio are constant overthe time-span in which the act is performed. On a higher temporal scale, thesevalues themselves change in the course of development as the neuromuscularapparatus of the oral cavity itself changes. The anatomy of the articulatoryapparatus changes as muscles become stronger, more efficient, and so on. The


point is, as Thelen (1995) shows in another context, that diverse timescales areall simultaneously interwoven in the same developmental dynamic. Therefore,scalar heterogeneity is already fully evident in the indexical phase characteristicof children's early protolanguage.

There is a further sense in which Halliday's example involves more than thebiomechanics of articulation. That is, the development of the articulatoryprocesses whereby the child establishes first-order contextual redundanciesbetween vocalization and perceived environmental event constitutes in itself afundamental fact about the development not only of sensori-motor processes e.g. the neuromuscular processes of vocal-tract gestural activity - but also ofhigher-order meaning-making.

Vocal-tract gestural activity, as in the example mentioned above, and theperceiving of the relevant environmental event, are cross-modally correlated onthe timescale on which the act occurs, i.e. seconds and fractions of seconds.Children vocalize and look at, touch, and listen to perceptual phenomena ofmany kinds across a large number of specific occasions. In the process ofengaging with perceptual phenomena in their surroundings, as well as withother people, infants may engage in a varied repertoire of vocal gestures rangingfrom babbling to crying to cooing, and so on. In so using the neuromuscularresources of the vocal tract, they therefore learn what it feels like to deliverdifferent levels of energy to the muscles of the vocal tract as well as what effectsthese have on their surroundings, including other human beings. That is, infantsmodulate the resources of the vocal tract and in so doing they produce salientchanges in their immediate environment. For example, in the instance citedfrom Halliday the child's high-pitched squeak both directs the mother'sattention to the pigeons and elicits a linguistic response from her.

The work of Halliday and others on infant protolanguage has shown thatvocal gestures such as the one referred to here have both expression andcontent. The point is that children, through the repeated use of such vocal andother gestures in potentially very many different situational contexts, arelearning about both their perceptual-motor systems and how these relate to theworld. In other words, their sensori-motor exploration and sampling of theenvironment constitutes information for the brain both about their bodies andabout the world beyond their bodies. The infant is experimenting by modulatingthe energy to the muscular systems of the vocal tract and seeing what kinds ofvocal gestures result, what kinds of phenomena these correlate with, the kinds ofresponses these obtain from others, and so on. In this way, the infant learns thata given category of vocal gesture such as the high-pitched squeak referred toabove can be correlated with a given phenomenon in the world by virtue of theprocess of recategorization of sensori-motor routines as conceptual routines(Edelman 1992: 89-90, 246-7). This process of recategorization produces acorrelation between the conceptual-semantic discriminations the organismmakes in its environment and the possibilities for acting on and intervening inthe environment. That is, the vocal gesture is a dialogically oriented and coordinated act whereby the infant causally intervenes in the world by, forexample, directing the attention of others to some object of interest. In thissense, the act is proto-interpersonal in character.


By the same token, the correlation of such sensori-motor routines withconceptual discriminations made on the basis of value-laden perceptual activityis proto-experiential. We see here how early perception-action routines involvingvocal and other gestures constitute active, movement-based intervention in theenvironment combined with the perceptual pick-up of relevant environmentaldiscriminations. These activities take place in the here-now scale of seconds andminutes. At the same time, they are a constitutive part of a longer timescale inwhich language emerges on the basis of these early re categorizations of sensorimotor routines as conceptual ones. Categories emerge on the basis of generalizations made from task-specific activities. Thus, the child studied by Hallidaylearns that a given category of vocal gesture qua sensori-motor act, involving thecontrol of the neuromuscular systems of the vocal tract, correlates with someaspect of the world outside his or her body. Moreover, vocal gestures exhibitvectorial qualities of direction and magnitude along their trajectory such thatthey afford orientation both to the source of the event and to the environmental affordance qua stimulus information which is the source of interest orattention. This is so in the sense that the vocal gesture is perceived as beingintentionally directed to a given environmental affordance which it indexes as arelevant contextual value. Furthermore, the vocal-auditory act cross-modallymutually correlates with the perception of inputs from vision, movement, andother information sources.

The process of mutual correlation between the sensori-motor and theconceptual domains arises on the basis of the brain's functioning as a somaticrecognition system (Ede1man 1992: 89-98). The brain recognizes andcategorizes events through its own activity, rather than through processes ofinstruction. Thus, sensori-motor repertoires and conceptual repertoires mutuallyrecognize each other. For example, the child's high-pitched squeak qua sensorimotor act correlates with the micro-functional semiotic category 'demand forexplanation'. It is this two-wayprocess of mutual recognition between the sensorimotor and conceptual domains that paves the way for the emergence of the twoway semiotic relation of realization between expression and content in the eventof language. Of course, the example itself shows the limitations of the notion of'conceptual' category as a covering term for the diversity of categorial distinctionsthat are correlated with sensori-motor repertoires. The term 'conceptual' isideationally biased. In actual fact, the categories correlated with sensori-motorroutines also and most crucially include interpersonal-interactional ones, as inthe example. In ways that will be explored in chapter 6, it can be seen that thebrain's activity of mutual correlation between sensori-motor and conceptualroutines is a contextualizing activity whereby me ta-redundancy relations arecreated between the two domains.

This last point has important consequences for how we understand therelations between language, context, and brains. As we have seen, the metafunctional diversity of language form - both expression and content - meansthat semantics is intrinsic to the internal organization of language form.Moreover, language form is characterized by the functional overlap of diversesemantic regions. Each of these metafunctional regions makes its specificcontribution to the meaning of the whole. The metafunctional organization of


language shows how language is internally related to the contexts with which itis integrated in use. Thus, context is not an external happenstance or epiphenomenon which has no systematic functional relationship to the organization ofthe content and expression strata of language. This has important implicationsfor our understanding of how language is learned and embodied in neuralnetworks. First, the contextual redundancies between language forms and theircontexts of use mean that the learning of language incorporates the contexts ofthe meanings and forms learned. Secondly, neural networks embody thecontexts, bodily activities, and social practices in which meanings are learned inthe form of the meta-redundancy relations that connect, say,vocal-tract gestures,speech sounds, lexicogrammatical forms, discourse structures, and so on, inlarger patterns of contextualizing relations. However, I shall now return to thefar more primordial, pre-linguistic forms of interaction that are the central focusof this section.

LakoffandJohnson (e.g. 1999) have discussed the relationship between ourforceful sensori-motor interactions with the world and the emergence oflinguistic categories of goal-directed action. The primordial experiences ofkicking, reaching, grasping, pulling, hitting, and so on, in early infancy involvethe use of children's bodies in order to bring some aspect of the world in linewith their needs and desires. In so doing, the children must harness and reducethe many degrees of freedom of both their bodies and the world so that somekind of match is created between the two in a given context of action. In this way,children learn to generalize categories of embodied-action schemata which theycan adapt and use across a very wide variety of different situations. A schematiccategory of the body's forceful interactions is abstracted from these diversesituations and correlated with a conceptual categorization. Moreover, Edelman,as we saw above, claims that the brain is a selective recognition system thatrecognizes events by its own activity, including its own activities of recognition.This is important because the recognition of a category such as 'forceful interaction with the world' as a means of bringing about causal effects in the worldentails, with the emergence of proto-language, a recognition that sensori-motoracts such as the vocal gesture discussed above can be used intentionally to causesomeone else to act in a causal way, both physically and semiotically. In the aboveexample, the child's high-pitched squeak is an intentional act which is dialogically oriented to causing another social agent - the mother - to provide anexplanation of the perceived phenomenon (the pigeons).

We can see here how the correlation of sensori-motor activity with aconceptual category gives rise to the schematic category [BODY-FORCEFULINTERACTION-ENVIRONMENT], as described in Lakoff and Johnson (e.g.1999: 270-6). This may be seen as a kind of Ur-experiential category which pavesthe way for its further re categorization as more delicate linguistic categorizationsin the way described above. The embodied schema described here may beanalysed into its constituent parts, as shown in Table 3.5.

The schema outlined in Table 3.5 shows that bodily activities such as reachingfor, grasping, and hitting objects can function to explore and obtain informationabout some object through, for example, haptic exploration. By the same token,they can be used to cause the given object to behave in a given way such as


Table 3.5: Emergence ofembodied categoryformation and its differentiation into'learning about' and 'acting on ' through forceful interaction with environment

Body part Vectorial control of force Object in environmentin formation of movementtrajectory towards object

Hand-arm Reaching movement Grasp or touch object forhaptic exploration

Hand-arm Reaching movement Cause object to move indesired way

Agent Process: action Affected

making it move in the desired way by hitting or kicking it. Of course, this secondpossibility also affords perceptual information about particular kinds of environmental events, in particular information concerning the relations betweenobjects and events which are implicated in a causal series. The understanding ofthe possibility of harnessing causal relations between one's bodily activities andthe behaviour of objects in the external world through actions such as hitting,knocking, kicking, and so on, is a forceful interaction between body and worldinvolving physical (efficient) causality.

However, infants also learn that their bodily actions can be deployed in orderto cause other persons to act in ways which satisfy their needs and desires. In thiscase, the causality involved is not physical; instead, it is semiotically mediated bymeans of some elementary sign such as a vocal gesture, facial expression, gazevector, or other bodily movement. Body-Forceful Interaction-Environmentschemata such as those just described emerge from the very first weeks of theinfant's life when infants begin reaching towards objects in their environment(Thelen 1995: 83). We can see how the Body-Forceful Interaction-Environmentschema can serve as a means of learning about the environment as well as ameans of acting on and bringing about causal events in the environment. In thevery earliest stages of reaching to touch objects, it is unlikely that the 'learningabout' and 'acting on' aspects are differentiated. In time, new forms andfunctions emerge from the dynamics of the earliest possibilities.

Without trying to suggest any kind of precise temporal or developmentalsequence, it can be seen how Halliday's (1975; 1978b; 1993: 109) pragmatic('doing') and mathetic ('understanding') categories, which are two of thediscrete microfunctions he postulates as occurring in the early stages ofchildren's protolanguage, can be seen as a further, more specified developmentand integration of the 'learning about' and 'acting on' functions deriving fromthe increasing topological differentiation of the originary Body-Forceful Interaction-Environment schema.

This most abstract schema therefore provides a model for the lateremergence of more specific schemata which are diversified over various


domains. For example, we can see how the most schematic experiential categoryof [Agent 1\ Process 1\ Affected] in the linguistic domain can be derived from thestill more schematic conceptual categorizations of sensori-motor experience. Inturn, the most schematic linguistic categories relate to the more specific lowerlevel experiential categories which operate in specific experiential semanticdomains such as the material, the mental, the relational, the verbal, and so on(see Davidse 1991: 14, for a detailed development of this principle with respectto material processes clauses in English; see also Halliday 1967b, 1967c, 1968).

Similarly, the abstract sensori-motor schema also provides a superordinatemodel for the emergence of more specific interpersonal action schemata whichare diversified over various semantic domains such as 'giving' and/or'demanding' linguistically construed information or material goods-and-services.The recognition of this category entails both a capacity to discriminate theforceful control of the body or some part of this in relation to desired outcomesin the world and, at the same time, the embodiment of an understanding of thepossibili ties for the self to act on the world of the non-self, including other agentswho can be recruited as eo-agents in the bringing about or satisfying of one'sneeds and desires. This means that the category is Ur-interpersonal as well. Thisaspect may be schematized as follows:

[[SELF-AS-AGENT] ACTS] --> [OTHER-As-AGENT2 1\ ACTION 1\

DESIRED GOAL]]

[[child utters high-pitched squeak as 'demand for information'] --> [motherresponds: gives information]]

Figure 3.11 illustrates the emergence of embodied category formation from thebody's primary forceful interactions with its immediate environment.

The correlation of the original sensori-motor schema with the conceptualschema 'forceful interaction' is topological in character. It constitutes a very wideattractor basin representing a very broad - i.e. schematic - category in theearliest stages of the infant's engagements with the world. The basis of thiscategory is iconic in the sense already defined, viz. topological-eontinuousvariation in forceful movement of the body selectively maps onto topologicalcontinuous variation in the world. There is differentiation, but it is topologicaldifferentiation, based on continuous sensori-motor exploration and adjustmentto continuous variation in the stimulus flux afforded by the events in the child'senvironment. The experiential and interpersonal functions in the linguisticsense are not differentiated. Instead, the broad attractor basin is differentiatedon the basis of more specific bodily engagements with the world such astouching, grasping, kicking, pulling, and so on.

The onset of protolanguage represents a dynamical shift to a new state space,itself brought about by the entraining of the child's trajectory to the higherscalar trajectory of the ecosocial system in which he or she lives. In this way, theprior attractor landscape is altered by emerging changes in the relations amongits component parts, along with changes in the boundary conditions of thesystem as it moves along its time-bound trajectory. In this context, the prior


phase of topological-continuous differentiation is reorganized, giving rise to newpatternings and, hence, new possibilities for acting and meaning in the world.The destabilization of the Ur-category described above as Body-ForcefulInteraction-Environment gives way to a shift into new attractor states such as themicrofunctions in the indexical phase of protolanguage. The appearance of theprotolinguistic microfunctions represents increasing semiotic differentiationwhich is now increasingly typological in character, and at the same time theemergence of indexical probability, which integrates iconic necessity, leads toextended adaptive capacity over increasingly larger space-time scales.

Figure3.11: Linguistic categorization and its emergence from pre-linguistic forcefulinteractions between body and environment

BODY - FORCEFUL INTERACTION-ENVIRONMENT

tBODY PART - VECTORIAL CONTROL OF FORCE IN FORMATION OF

MOVEMENT TRAJECTORY TOWARDS OBJECT - OBJECT IN ENVIRONMENT

'Act on'

t tM'~~";'

TRANSITIVITY + MOOD

~~

Linguisticmetafunctions:

'Learn about'Protolinguisticmicrofunctions:

[Agent/Process/Affected] Proposition/Proposal

Transitive Ergative Indicative Imperative

Actor-Process-Goal Instigator-Process-Medium

Goal- Goal- Event Actiou Declarative Interrogative Oblative Imperativedirected achieveing instigation instigation+ intention -intention + event + action

He cut The car The flood He walked Heeut Did he Shall I Cutthe cake struck the burst the the dog the cake eutthe eutthe the cake!

tree dam cake? cake?

Part III

4 The Semiotic Basis of Consciousness

1. First-person and Third-person Accounts of Consciousness

Psychologists and neuroscientists agree that conscious experience is a functionof brain processes. Furthermore, they agree that conscious experience is directlyaccessible only to the individual who is undergoing that particular experience.This general consensus has led to the formulation of two main hypothesesconcerning the nature of conscious experience. According to the firsthypothesis, only the subjective verbal reports of the person who undergoes theconscious experience can be taken as a reliable indicator of that experience.This experience may be 'reported' by linguistic means to someone who isextrinsic to the reported experience; however, the experience itself remains theunique possession of the person who experiences it. According to the secondhypothesis, consciousness may be explained just like any other observablephysical process. That is, consciousness is, in the final analysis, explainable interms of neural activities in the brain. These neural activities are accessible to anexternal observer who can observe these independently of the subjective reportsof the person who undergoes the conscious experience. In this second view, consciousness is reducible to and is, therefore, explainable in terms of the physical(neural) processes that take place in the brain. However, clinical researchers(e.g. Libet 1996: 101) point out that studies of brain activity using the technologyof positron emission-tomography (PET) and magnetic resonance imaging(MRI) can tell us where, in the brain, changes in activity take place. These technologies cannot, however, describe the specific neuronal activities that may beinvolved. Nor can they tell us anything about the relationship between theseactivities and our consciousness of our own or others' mental events.

A further set of questions arises at this point. The first concerns the unitaryand integrated nature of conscious experience. The human brain possessessome 100 billion neurons; each neuron may have thousands of connections withother neurons. Moreover, many brain functions are localized to specific areas ofthe cerebral cortex in the form of 'specialized columns of neurons' (Libet 1996:114). Yet, whatever neural activity in the brain does reach the threshold ofconscious experience is experienced as both unified and integrated (Libet 1996:114). The second question has to do with whether mind or conscious mentalactivity exerts any causal influence on physical brain processes. It is worthquoting Libet at length on the two approaches which have predominated inattempts to answer these two questions:


On issue A, Eccles (Popper and Eccles, 1977: 362) has proposed, based on adualist-interactionist view, that 'the experienced unity comes not from a neurophysiological synthesis but from the proposed integrating character of theself-conscious mind'. This view had, in principle, been expressed bySherrington (1940), and also by Sperry (1952, 1980) and Doty (1984). ForSperry and Doty, however, the mind doing the integrating was viewed inmonistic terms, as an emergent property of brain processes. On issue B, bothEccles and Sperry have proposed that the 'mental sphere' could influenceneuronal function but in dualistic vs. monistic terms, respectively. The viewheld probably by most neuroscientists (and perhaps modern philosophers) isa monist-deterministic one, in which conscious mental experience is simplyan 'inner aspect' of brain function (identity theory); it is fully determined byknowable physicochemical processes in the brain, and its apparent ability toinfluence brain function is a subjective illusion with no actual causal powers.Although each has explanatory power and each can be shown to becompatible with (not falsified by) the available evidence, none has beensubjected to appropriate or adequate experimental testing in a format thatcould potentially falsify it. (Libet 1996: 114)

Libet then discusses the notion of 'binding' as a possible solution to the first ofthe two questions discussed above. The notion of binding is based on thediscovery of global patterns of synchronization of neural activity, which someresearchers see as 'the neural coding for a unified mental image in an otherwisechaotic background' (Libet 1996: 114). However, Libet cautions as to the needto distinguish between "binding" at a cognitive level (that mayor may not involveconscious experience) and the binding that refers to unity experienced inconscious awareness' (1996: 114). Libet hypothesizes the notion of a unified'conscious mental field' (1996: 115) which would solve the question of both theunity and the integration of conscious experience. Furthermore, the consciousmental field, produced by diverse neuronal activities, can causally affect andchange neuronal function (see Thelen and Smith 1994: 189 for an alternativeaccount based on direct cross-modal communication; chapter 5, section 3, forfurther discussion).

Many brain processes are not conscious even though they relate to functionally differentiated brain functions which have to do with memory, perception,learning, and so on. However, the recognition that there are both conscious andunconscious brain processes has not led to any scientific consensus concerningthe nature of consciousness. The reductionist view that consciousness can beexplained in terms of physical brain processes remains unable to relate these topersonal experience or to explain the link between such experiences and thephysical brain processes to which these are causally related. Bateson (1980[1979]: 39-40) points out in this regard that we are aware of the products ofperception; however, we are not aware of the neurophysiological processes whichmake perception possible. Bateson's observations can be extended to conceptualand semiotic, including linguistic, processes. Thus, we are aware of the thoughtsthat we attend to in inner awareness whereas we are not aware of the neuronalactivity which realizes these thoughts. Moreover, many aspects of linguistic

THE SEMIOTIC BASIS OF CONSCIOUSNESS 173

processing in the brain are not conscious. This fact has long been recognized bylinguistics. Typically, the participants in linguistic interaction are consciouslyaware of the speech sounds which they attend to, as well as the sense of ameaning which they understand and attend to as the interaction proceeds.However, many aspects of linguistic processing are not normally available toconscious awareness and would greatly impede the process of consciousmeaning-making if they were.

2. The Representation of Subjective Experience in Consciousness inRelation to the Higher-scalar Environment of the Individual

Theories of consciousness are concerned with the problem of how we representour own subjective experience as well as that of others. Deacon formulates thisproblem as follows:

The problem with other minds is that the glimpses we get of them are allindirect. We have a subjective experience of our own thought processes, butat best only an imagined representation of what goes on in others' subjectiveexperience. When we speculate about others' 'inner' states, the only data wehave to go on are what they tell us and what we observe of their physical states.Like the subject in the Turing text, we are forced to make assessments onrather limited and indirect data. We can, it seems, have direct knowledge onlyof ourselves. In philosophy, this argument is aptly termed solipsism (from theLatin solus, alone, and ipse, self). (Deacon 1997: 424)

... the problem of representing the subjective experience of another and theproblem of representing one's own subjective experience both entirelydepend on the nature of the representational processes involved. Theproblem is not whether some knowledge is representation and some is directand unrepresented. The problem is, rather, what sort of representation isinvolved, and what knowledge this provides of our minds and the minds ofothers. (Deacon 1997: 425)

Deacon argues that both first-person and third-person perspectives on one's ownand other's experiences are both dependent on 'representational processes'.The notion of representation, in contrast to the notion of 'direct and unrepresented' knowledge (and experience), raises the further question as to howconsciousness can be theorized as a highly specified meaning system on allintegrative levels (iconic, indexical, symbolic) in the inner perspective of the selfwho contextualizes and interprets the given experience as meaningful in the perspective of the self. Consciousness is based on internalized self-self transactionsthat have emerged from and are a further reorganization of self-other transactions that the self has engaged in with others (see Thibault In press. Thissuggests that, rather than representation, dialogue and/or its less specifiedanalogues on the iconic and indexical levels are fundamental here. As we haveseen, Libet, Bateson, and others point out that our consciousness of our


thoughts, perceptions, and so on, does not amount to consciousness orperceptual awareness of the neurophysiological processes that underpin andsustain these. Instead, we are conscious of the 'products' or the phenomenological manifestation of the experience itself. This fact suggests that the centralissue here is, once again, that of the interpretation of meaning in the perspective of the self.

The observations of Libet that I discussed in the previous section suggest thatconsciousness entails the global reorganization of neuronal activity in the brainacross many different scalar levels of organization. Emergent, higher-scalar levelsof neural organization reorganize the information on lower levels as meaningfor the self. The higher-scalar levels constrain the lower-levels of the body-braincomplex to its own dynamics. Moreover, the highest scalar level, whichconstrains and interprets the lower levels, is the self. The self is realized by themost complex, most global (higher-scalar) level of neuronal organization in theindividual's brain. The self can be seen as a system of interpretance thatinterprets and enacts meaningful experience at the same time as it constrainsand modulates sensori-motor activity along its trajectory within the perspectiveof the self. Consciousness, in my view, just is a highly specified system ofmeanings in this sense in the brain of the individual. However, this does notmean that the self is reduced to (higher-scalar) neuronal organization perse. Thepoint is, rather, that very complex global levels of neural organization providepersons with the neural resources to contextualize and, therefore, to givemeaning to experience. They give meaning to experience in ways that may haveconsequences for the signs which the individual outputs into the environment asspecific action trajectories. The self is the embodied person who acts and meansand is recognized by others as acting and meaning in ways that can be related tothe self as the source of these actions and meanings. Selves are integrated totheir contexts in ways that are interpretable as meaningful from their own selfperspective as well as from that of others.

Deacon (1997: 426) alludes to the notion of 'theory of mind' which has beendeveloped in cognitive psychology to explain how one learns, though enculturation, to view things from another's perspective. The 'theory of mind' firstemerged in discussion concerning ape language and cognition (Premack andWoodruff 1978). It then became a dominant theme in developmental studies ofhuman infants (e.g. Astington, Harris and Olson 1988). It has also been influential in autism research (e.g. Baron-Cohen 1989a, 1989b). For a criticalreflection on this entire debate, see Battacchi, Battistelli and Celani (1998).According to Deacon, it is the development of 'symbolic referential abilities'which enable humans to interpret 'information from another's behavior to beinterpreted as representing another mind' (1997: 426). Species that lack suchsymbolic abilities are limited to the representation of indexical associationsbetween stimuli from both self and others. Deacon explains this symbolic referential ability as follows:

As novelists and poets amply demonstrate, the range of personas and experiences that can be conveyed through symbolic media is unbounded. In a veryreal sense, this gives us the ability to share a virtual common mind. Because


symbolic representation maintains reference irrespective of indexicalattachment to any particular experiences, when an idea or a narrative orsomeone's experience is reconstructed by another, it can be regrounded, soto speak, by interpreting it in terms of the iconic and indexical representations that constitute the listener's memory. Symbolic reference is interpreterindependent, because each interpreter independently supplies the nonsymbolic ground for it. (Deacon 1997: 427)

Deacon adds to the standard account of 'theory of mind' the notion that this ismade possible by the development of symbolic referential abilities. The varioussteps in Deacon's line of argumentation may be summarized as follows:

1. Each person's mind is separate from and independent of every otherperson's mind;

2. Access to another person's subjective experience - i.e. to that person's mind- is limited and indirect;

3. Minds represent subjective experience;4. Minds are possessed by self and other;5. Representation is the means whereby we know about our own mind and the

minds of others;6. In order to have access to others' minds, we must be in possession ofa theory

of mind which is based on symbolic referential abilities.

Deacon assumes that representation is the key to unlocking the subjective experiences that are housed in someone else's mind. Moreover, the level of first focusin this account remains that of the individual mind-brain. However, individualorganisms are not separate from other organisms; they are constitutive parts ofhigher-scalar systems of interpretance. A system of interpretance is a means ofdeciding whether information deriving from either self or nonself is potentiallymeaningful to our own observational perspectives, as well as to others whopotentially share these same observational perspectives. Individual body-brainsdo not exist as independent entities. They are always integrated with higherscalar systems and, at the same time, constitute the interpretative framework forsmaller-scale ones. This is so both biologically and socially. In such a view, therelevant focus of study is not the individual body-brain per se, but the way itis embedded in and is a constitutive part of its higher-scalar ecosocialenvironments.

The individual body-brain complex is a system of meaning-making potentialrelative to its higher-scalar environments. Moreover, other body-brain complexeswith which the individual interacts are integrated to and embedded within thesame overall supersystem. In such a system, the individual subsystems contributeto the meaning which is stored in a given subsystem through interaction with it.Individuals learn about other individuals who share the same system of relationswhen, on the basis of dialogically co-ordinated interaction between two individuals, the individual system's Innenwelt is changed. The individual subsystemsdo not directly interact with each other; the interaction between them isalways mediated by and entrained to the dynamics of higher-scalar systems of


interpretance and their associated practices. Change in the individual's Innenweltis a result of the meta-redundancy relations - the patterns of contextualizingrelations - that are constructed between individual (self) and other individuals(nonself) on the basis of such transactions. Thus, meaning is stored, not at thelevel of the individual perse, but at the level of the contextual configurations e.g. the dyads - which integrate individuals to their ecosocial environments and,therefore, to the systems of interpretance that are embedded in these.

The notion of self-organization (chapter 1, section 2, pp. 8-11) provides ameans of overcoming the tendency of the representational theory of mind tooveremphasize the individual's experience of his or her environment. Selforganization entails an increase in a system's overall semiotic complexity andspecification as a result of the collection of the products produced by its owncascading (Salthe 1993: 176). Following Halliday's (1993) notion of the 'interpersonal gateway' to meaning-making in the process of the young child's'learning how to mean', we can see how the emergent self is dialogically linkedto its Umwelten in such a way that perceived events in the child's purview impingeon the child as indexes, in the first instance, of environmental events. Theseindexical signs of environmental events are, in turn, built up and elaborated inthe child's Innenwelt in the form of linguistic and other semiotic models as tohow one can act on and construe the world. In this way, there emerges adeveloping system of possibilities for action which afford the child a repertoireof possibilities for acting on, intervening in, and constructing his or her world.As noted before, these processes always occur in and through the mediatinginfluence of the dyads which organize the child's interactions with the nonself.The child responds to events in the world, including the activities of other conspecifics, with an expanding set of possibilities for action. Dialogically organizedinteractions with others lead to the development of models of the world whichmediate the child's subsequent interactions with the world. Such models includethe linguistic system which is elaborated and stored in the individual's Innenweltin the course of development and individuation.

3. Locating the Seat of Consciousness

Where is the seat of consciousness? According to both dualists and reductionists,it is in the brain. But this conflicts with our consideration of self-organization, asexplained above. Higher-order consciousness entails the emergence of a sense ofself (Edelman 1992: 165-72; Edelman and Tononi 2000: chapter 15). This isdependent upon the emergence from initial proto-serniotic vagueness and indeterminateness of ever more specified and determinate symbolic possibilities foracting and meaning. That is, the individual is defined by the entire life-spantrajectory along which the self unfolds and individuates. (For related notions seeHarre's (1979: 312-34) theory of the individual's social trajectory or moralcareer; Sartre (1969 [1943]) on the project; Salthe (1993: 180-5) on the individuating ontogenetic trajectory; Riegel (1979) on developmental lifespanpsychology.) The projects to which the individual is committed are more highlyspecified according to the higher-scalar meaning systems, or systems of interpre-


tance, which are available to the individual in his or her ecosocial environment.In the human case, the self of sense emerges in and through the individual's participation in dialogically organized exchanges between self and nonself. In thisway, the self is constructed, elaborated and maintained in the brain in the formof higher-order, self-referential meta-representations such that the systemconstructs meanings about its own internal states and its relations to others(Flohr 1991; in this volume, chapter 6, section 2, pp. 283-9).

The central argument of this chapter is that consciousness is a structuredsystem of meanings whereby we organize our experience around a notion of theself along a trajectory (Harre and Gillett 1994). It is the most global manifestation of neuronal self-organization. Because of this, it has the capacity both toconstrain and to interpret the activities and experiences of its own lower-scalarcomponents in its own perspective. The self is, then, a self-organizing system ofmeanings in and through which the individual interprets the effects of its ownself-organisation (Salthe 1993: 158). Harre (1983: 146) talks about the self asconstituted 'in a discourse made up of speech acts having the form of self predications'. I would rather emphasize the building up of meanings along atrajectory through dialogic exchanges with other selves with whom one'strajectory intertwines. The point is that consciousness, in the view I am puttingforward here, is necessarily founded on dialogic exchange processes which makepossible the emergence of self-referential perspectives. It is these self-referentialperspectives which enable experience to be sourced at or centred on a determinate 'self'. Consciousness, in particular higher-order consciousness, entails theself-organization of perspectives which, when viewed globally, are seen as havingrelative coherence and order. However, the local manifestations of consciousnessare not necessarily orderly. Consciousness is itself an emergent consequence ofdifferent levels of semiotic organization - both scalar and integrative.

The further question arises at this point as to the eo-evolution oflanguage andother semiotic modalities and the human brain. Given my premise that consciousness is not reducible to its neuroanatomical enabling conditions, I arguethat the various levels of human consciousness qua system of meanings arestructured according to metafunctional principles of organization or theirvaguer, less specified analogues on lower integrative levels. As we shall see in moredetail in chapter 5 in relation to Damasio's theory of consciousness, the metafunctional organization of consciousness qua internalized semiosis in the perspective of the self is an appropriate means for the structuring of the experienceswhich are sourced at a given self. The metafunctional structuring of consciousness is therefore able to account for the necessarily dialogically constituted natureof the relations between self and nonself, as well as the embodied grounding ofthe meanings attributed to a self relative to its surrounding milieu. The notion ofan intentional 'consciousness of' relation which holds between a person and anintentional object (see Harre 1983: 148) is a meaningful relation. We shallexplore in chapter 5 how this relation takes the form of the metafunctionalsemiotic organization postulated here. For such a relation to occur, it follows thatthe person concerned has the requisite modes of semiotic organization stored inhis or her central nervous system. Moreover, the objects of conscious awarenessand attention on all levels in the postulated integrative hierarchy of iconic,


indexical, and symbolic modes of semiosis (see chapter 1, section 8; chapter 3,section 7) are experienced as either topological-eontinuous variation or typological-eategorial distinctions which make a difference to the person who has theconscious experience, relative to a given system of interpretance. Thus, one doesnot have conscious awareness of the brain states and processes - the neuralactivity - but of the meanings (iconic, indexical, and symbolic) which are variablyrealized by these and which are made available to conscious awareness in theinterpretive perspective of the self (see above). This point applies indifferently tophenomena of experience of all kinds, whether they exist in perceptual awarenessor are activated by non-perceptual awareness in the processes of imagining,hypothesizing, pretending, and remembering.

The relational structure of consciousness is therefore shown to have what Ibelieve to be revealing parallels with the metafunctional organization ofsemiosis. This does not mean that all consciousness is able to be 'reported'by linguistic means. Moreover, the distinction between 'reportable' and 'nonreportable' experience (Harre 1983: 155) does not mean that that which is nonreportable by linguistic means is outside the realms of conscious experience andawareness. This would be to underestimate the importance of perceptual-motorforms of categorization at the iconic and indexical levels of the integrationhierarchy of semiosis. Furthermore, the assumption of non-reportability woulddisallow the importance of topological-continuous variation as a means ofmakings differences that make a difference to some individual. Not all consciousness is necessarily 'reportable' in terms of the typological-eategorialsemantic distinctions made by the lexicogrammar of natural language. So called'non-propositional' awareness and knowing (Harre 1983: 155) is all about thiskind of non-symbolic construal of conscious experience. In any case, the notionof 'reportability' suggests that there is something empirically prior whichlanguage simply packages up and reports as the more essential experience.Instead, the metafunctional form of consciousness at all levels in the integrativehierarchy shows that the act of consciousness constitutes the object of consciousness relative to the categories of the system of interpretance which is usedto contextualize the experience.

The dialogic exchanges between self and nonself are possible because, in thefirst instance, the biological organism belongs to that class of physical systemswhich are dynamic, open systems (Prigogine and Stengers 1985 [1984]: 143;Lemke 1984a, 1984d, 1995b; Wilden 1981: 2). Such systems exchange matter,energy, and information-meaning with their environment and, at the same time,maintain their specific identity in relation to their environment by virtue of thesesame system-environment transactions. However, these exchanges of matter,energy, and information-meaning do not mean that the organism is simply 'in'an environment with which it exchanges matter, energy, and information. It isthrough these exchanges that the organism participates in the active constitution of its environment at the same time as it is constituted by these same selfenvironment transactions. It is in this way that organisms expand their Umwelten(Harre 1990: 301). In expanding their Umwelten, they define their own identityand structural integrity in and through these exchange transactions. In theseexchange processes, human individuals exchange both structure and meaning


with their environments. The intrinsic organization of these self-environmenttransactions and the ways in which they integrate both with the body-braincomplex and with the higher-scalar ecosocial environment can be explained interms of the metafunctional shape of consciousness that I first postulated above.The environment is, in the first instance, perhaps better formulated as thenonself in relation to the self. The self selectively orients to, engages with,interacts with, and interprets the nonself. The nonself includes other personsand objects, events, and so on. In exchange, the structure so exchanged isimplicit; verbal meaning is, in large part, explicit in fully contextualizedoccasions of discourse, though non-verbal forms of semiosis such as facialexpression, body movement, and so on, may be much less so. Importantly,meaning is distinguished from information by that fact that meaning is alwaysconstrued in relationship to the observational viewpoints and perspectives of theparticipants on the ecosocial scale in question. Information is not dependent onsuch viewpoints, but is a statistical measure of the variety in relation to which agiven event occurs (Wilden 1981: 17-18). Again, the importance of meaning andthe interpretation of experience in the self-perspective as meaningful forconscious experience are suggested by the relevance of viewpoints and,therefore, interpreters to meaning as distinct from information.

On the human scale which is of concern here, meaning requires humanperspectives and viewpoints which can be shared with others on the basis of thedialogically co-ordinated exchanges between self and others. Dynamic opensystems, of which human individuals are a specific case, maintain their structuralintegrity through the processes of obtaining required matter, energy and information with a high order of structure from their environments and returninglow-order matter and energy to the environment in the form of disorder. In thehuman case, higher-scalar ecosocial structures and processes integrate andextend the structures and regulatory principles of the individual. In the earlymother-infant dyads characteristic of primary intersubjectivity (Trevarthen1987, 1992), dialogic exchange processes integrate the child into the meaningmaking practices and the systems of symbolic possibilities of the group. Thus, theindividual's emergent symbolic or higher-order consciousness entails thebuilding-up and elaboration of internal structures of meaning and interaction.This has the dual function of (l) integrating individual consciousness into thesocial semiotic system and (2) providing individual consciousness with symbolicresources which expand and extend the individual's own possibilities beyond thehere-now scale of its material interactivity with its environment. Integration intothe symbolic possibilities of the social group is therefore necessary for theemergence of higher-order consciousness in the individual (Edelman 1992:131-6).

By the same token, higher-order consciousness in the individual is alsoessential for the social group. This is so for three reasons. First, the symbolicmeaning-making resources of the social group afford the continuing integrationof apprentices (the new-born, new members, etc.) into the group. Secondly, it isalso the means whereby consciousness is made shareable with other conspecifics,and in ways which ensure the maintenance of a higher-scalar collective consciousness in historical time. Thirdly, these resources make possible the social


semiotic construction of individual historical-biographical coherence withreference to the individual's unfolding lifespan trajectory in some community.These three points illustrate the need to view higher-order consciousness in theindividual as an integral component of a still-higher-scalar system of ecosocialrelations and processes which exist on much greater scales than the individualbiological organism.

The integrated nature of conscious experience, as discussed above in relationto Libet, can be seen as an instance of the non-linear operations of a complexsystem whose closed in ternalloops of neural activities and functions on very manycomplexly interrelated scalar levels of neuronal organization give rise tometastable states or fields of consciousness. Consciousness is a metastable state ofvery high complexity which enables strong cross-couplings between the individualbody-brains and selected aspects of his or her inner and outer environments.Such cross-couplings can only occur through semiotically mediated transactionsbetween individual and environment. Moreover, rather than seeing symbolic consciousness as the mere bottom-up aggregate of many individual body-braincomplexes, it is possible to conceive of the full-fledged emergence of symbolicconsciousness in the human species as the evolutionary consequence of the historically emergent differentiation of symbolic modes of social meaning-making(language, gesture, depiction, etc.) from the iconic vagueness of prior systems.The higher degree of specification of symbolic modes of meaning-making can beseen as the result of the increasing differentiation of the topological-continuousvariation that is characteristic of the prior system as it differentiates into a numberof more specified subsystems. In this way, the prior system constitutes theimmediate environment of the newly emergent system and therefore provides theinitial systemic conditions whereby the increased specification of the symbolicmode can emerge (see chapter 3, sections 4 and 7). This suggests that theevolution of symbolic consciousness in the human species arose from theincreasing differentiation of vaguer prototypes of exchange between organismand environment. The symbolic brain, in this view, is not the unit whichgenerated language, but is the evolutionary product of the further differentiationof previously existing semiotic modalities in their contexts of use.

Symbolic modes of meaning-making entail a higher degree of explicitnessand specification than do the iconic and indexical modes from which theyemerged and which they contextually integrate to their own (symbolic) level(chapter 3, sections 4 and 7). The emergence of the symbolic level in no wayimplies that the iconic and the indexical levels have been transcended. Linguistically mediated discourse is fully contextualized language in operation inrelation to other contextual features, including other semiotic modalities whichare relevant to its co-contextualization. It is a context-specific restriction of themeaning potential which is stored and elaborated in the brains of individuals atthe same time as it is, to varying degrees, shared by them by virtue of their beingembedded within the same system of interpretance in a given ecosocial semioticsystem. This meaning potential is stored in the individual's brain in largelyunconscious ways until is selectively activated according to the requirements ofsome context. Thus, explicitly meaningful and consciously enacted discoursecan be seen as a critically important resource for selectively cross-coupling the


individual to the ecosocial environment. Discourse is a powerful agency both forrecruiting individuals to relevant social practices and for integrating consciousness with its contexts in relevant ways. This is equally true of both 'inner' and'outer' forms of meaning-making. Linguistically mediated social meaningmaking is always to some extent explicitly meaningful for the participants in theexchange, whereas the iconic-indexical modes of posture, gesture, facialexpression, and so on are much more implicit and, hence, 'unconscious' to agreater extent.

Higher-order consciousness is characterized by the use of symbolic modes ofmeaning-making such as language. From the perspective of the individualsubsystem, it entails the capacity for the linguistic and other semiotic mediationand construal of experience. I have argued that this process crucially depends onthe integration of the individual organism into higher-scalar ecosocial processesand trajectories. At the same time, it is clear that some of the semioticallymediated exchanges that take place within the individual organism in the formof 'inner speech', linguistically mediated 'thought', internalized visual semiosis,and so on, are specialized to the 'inner' realm of the individual and set againsttransactions that occur in the 'outer' realm such that the former are seen asprivate or individual. However, it is important to recognize that even those formsof meaning-making which have been specialized to the inner realm are definableby and integrated into the semiotic resource systems of the higher-scalarecosocial system. If conscious experience is mediated by and integrated into ahigher-scalar system of interpretance, then it seems legitimate to ask how suchsemiotic resource systems, which are stored at the higher-scalar ecosocial level,give shape to consciousness itself.

The representational view of consciousness focuses on the individualorganism which represents the world as conscious experience. This has resultedin an overwhelming emphasis on the bottom-up brain-mind mechanisms andprocesses which 'cause' conscious experience in individuals. Similarly, thoseapproaches which focus on the ways in which our sensori-motor activity andperceptual systems constitute the body images and schemata in and throughwhich we construct the world continue to emphasize the individual organism asthe focus of theoretical inquiry (e.g. Lakoff and Johnson 1999). Alternatively, Ihave argued that consciousness cannot be characterized in terms of individualsper se. Instead, it requires a shift in focus to a three-level scalar hierarchy view ofthe systems relations that are involved (Salthe 1993: 36-46; Lemke 1999;Thibault 2000a). That is, it requires us to ask how the intrinsic dynamics ofhigher-scalar semiotic resource systems such as language and the social practicesin which these are embedded and deployed play their part in the shaping ofconsciousness.

In terms of the principles of scalar hierarchy theory and scalar heterogeneity(chapter 1, sections 3, 9), it is possible to re-think the locus or seat of consciousness in relation to the body-brain complex of the individual organism. Biologistssuch as Salthe (1993: 181) have pointed out that the molecules and cells whichmake up our bodies are constantly replaced. From the perspective of thebiological organism, we are not the same organism that we were just a few yearsearlier or even a year ago. Nevertheless, we construct a sense of continuity over


time so that our consciousness of our selfhood is maintained and enhanced inspite of these constant changes to the organism. In the process, we encounternew historical contingencies, which provide us with new memories and experiences. Salthe poses the question as to the locus of selfhood and agency in thisperspective. This locus is the entire ontogenetic trajectory of the organism,starting with birth and continuing until death. Any moment-by-moment actionof the organism is no more than a local manifestation of this trajectory (1993:181). To quote Salthe:

It would be convenient for present purposes to describe our own ontogenetictrajectories from a scale delivering a cogent moment of about one of ourlifetimes. In that case the entire trajectory is present simultaneously as a singleobject. This object can be taken to be the seat of self-organization, agency, andselfhood because these inhere in its ontogenetic trajectory rather than in theorganism, which is continually changing (besides being at any momentmerely a mechanism, which can't have those properties). (Salthe 1993: 182)

The entire trajectory is a 'cogent moment', as viewed from a still-higher-scalarperspective, with respect to the very different scale of the moment-by-momentexistence of the individual body-brain complex. The latter is a lower-scalar entitywhose lower-scalar dynamics operate on a much faster timescale than those ofthe higher-scalar dynamics of the trajectory as a whole. Each of the two scales hasproperties which are ontologically specific to its own scale and the respectiveentities and processes on that scale. It is symbolic or higher-order consciousnesswhich makes the emergence of such a trajectory possible. Moreover, theorganism which sustains the trajectory is a dissipative structure which is characterized by the historical emergence and accumulation of meanings along itsentire trajectory, Increased dissipation to the environment goes hand in handwith increasing structural complexity such that fluctuation to higher-order statesis increased, thereby allowing the evolution of new metastable states (Lemke1984c: 29). The trajectory is a self-organizing system. Self-organizing systems aredefined by the fact that they exchange matter, energy, and meaning with theirexternal environments in wayswhich bring about the spontaneous emergence oforder and pattern in the system (see above). Self-organizing systems have historyand individuality (Prigogine and Stengers 1985 [1984]: 176; Salthe 1993: 142;Lemke 1995b [1993]: 112-13).

Consciousness is, then, a trajectory-in-time. It is the trajectory which is the seatof selfhood and agency (Salthe 1993: 184). The trajectory of consciousness isboth meaning-based and regulated by memory. The body-brain complex of theindividual is the moment-by-moment realization of this trajectory. The two - Le.consciousness and the body-brain - exist on different scalar levels. The linkbetween them is one of realization, rather than causality. The relationship ofrealization specifies (1) that the trajectory of consciousness is realized by thebody-brain complex; (2) that the body-brain complex realizes, or embodies, consciousness, therefore the body-brain complex is a constitutive part of the contextof consciousness on account of the fact that consciousness is always embodied inthe body-brain dynamics of individuals. In other words, the body-brain is part of


and is embedded in a still-wider system in and through which meaning is made.The body-brain constitutes and embodies the lower-scalar enabling conditions ofconsciousness as well as being the lower-scalar environment of the trajectory ofconsciousness. In a sense, the trajectory of consciousness can be thought of asthe meaning - the content stratum - which is realized by the expression stratumof the body-brain complex.

The self-organizing character of this trajectory depends crucially, as I pointedout above, on the ongoing exchange of matter, energy, and meaning betweenthe system and its environment. In the case of symbolic or higher-order consciousness, this is founded on the dialogically organized meaning-makingexchanges that take place between self and others who share the same ecosocialperspectives and viewpoints. The trajectory of higher-order consciousness is,therefore, constituted against a background of the dialogically organizedmeaning exchanges in which the individual has participated along the course ofhis or her trajectory. The temporal organization of the trajectory is thus madepossible by its dialogic synchronization with the trajectories of others with whichit engages (Riegel 1979: 167). The evolution of new metastable states ofconsciousness occurs on the basis of conflict or contradiction which must beresolved for progression to a new state to occur. Dialogue is driven by thenegotiation of such difference or contradiction, which the participants in thedialogue seek to resolve in some way. Resolution, or what Bakhtin (1986: 76-7)called 'finalization', can never be achieved in any ultimate way. The dialogicnature of consciousness means that new questions, new doubts, new differenceswill always arise such that the trajectory of consciousness remains open-ended,therefore constantly engaging in {dialogic} interaction with others.

Physical brain processes and states are inadequate to explain consciousnessprecisely because they cannot deal with the principle of final causality, which isnecessarily entailed by the dialogic basis of consciousness. Physical brainactivities are cause-effect relations within just one level of relations in the scalarhierarchy. Final causality, on the other hand, is a higher-scalar constraint whichoperates 'downwards' on lower levels. Such constraints, which emanate from thehigher-scalar levels, are, increasingly, semiotic, rather than physical, in character.Final causality does not derive from the lower level of physical brain processes,which would be efficient causes in Aristotle's schema. The latter operate on amuch faster temporal cycle than does the trajectory of individual consciousness,which spans a lifetime. This means that the much faster rate of the lower-levelbrain activities in a given cycle of such activities comes to completion while thehigher-scalar-level trajectory of consciousness unfolds according to its muchslower dynamics, and at the same time continues into the future. Thus, thehigher-scalar level of the trajectory, which is the seat of consciousness, operateson and selects the lower-level brain activities at the same time as it pulls theseinto the future (Salthe 1993: 270).

The individual agent is dialogically oriented to the nonself along the entireduration of its trajectory. Therefore, the forever open and incomplete characterof this relationship is the means whereby the trajectory of the individual self ispulled into its future by the need to resolve ('finalize') local contradiction anddifference through dialogic engagement with the nonself. It is in this way that


the trajectory self-organizes at the same time as it individuates. In terms ofthe three-level logic of the scalar hierarchy (see chapter 1, section 3, pp. 14-16),this means that level L - the level of first focus, or the ecosocial scale onwhich individuals interact with each other and their material and semiotic environments - is constituted by the dyadic character of self/rionself transactions.These transactions can be modelled as a hierarchical triplicate of levels, asfollows:

L+1: Historical-biographical trajectory of individual consciousness (historicalbiographicallifespan)

L: Dialogic basis of ecosocial self!other transactions along logogenetictrajectory of unfolding text-time (moment-by-moment)

L-l: Physical brain activity in individual organism (milliseconds)

4. The Meaning-making Capacity of the Body-brain Complex through theDiscrimination of Difference

The fact that the world is not something 'out there', an object in itself, having itsown reality independent of our ways of making meaning about it, has importantimplications for the body-brain. The recognition of this fact requires the abandonment of the representational theory of mind and cognition and its replacement with a view in which the world and its meanings are actively produced byus. Further, the systems of perceptually, conceptually, and semiotically salient distinctions constituted by our perceptual systems, the conceptual distinctionselaborated in primary consciousness, and language and other semiotic resourcesystems in higher-order consciousness are inseparable from the architecture ofour body-brains and, hence, the embodied nature of our ways of makingmeaning (chapters 2 and 3). Edelman's (1989) notion of an 'unlabelled' worldand Saussure's (1971 [1915]: 156) notion of the 'amorphous flux' (see alsoThibault 1997a: 166-71; 1998a: 25-6) are apt metaphors for expressing the indeterminate nature of pre-semiotic vagueness before it is made into somethingmore determinate in and through our deployment, in context, of the systems ofcategorization made available to us by some meaning system. In other words, wetransform indeterminate vagueness into something more determinate andmeaningful for us by actively exploring and intervening in the environment thatwe inhabit.

Organisms produce the world, so to speak, through their capacity forembodied exploration and interpretation of it. In so doing, they make theproperties - perceptual, affective, and motivational - of the given aspect of theworld which is the focus of their attention and interest emerge. In selectivelyfocusing on and showing interest in some things rather than in others, theyimplicitly assign value and, therefore, motivational salience to some thingsrather than to others. They do so by means of the various sensori-motor andsemiotic systems which enable them to construe, potentially, very many degreesof difference in the topological vagueness of the ambient flux which surroundsthem. Thus, the human hand, in comparison with the hands of other animals,


in conjunction with the system of visual perception, has a greatly enhancedability to discriminate very many degrees of difference, and therefore fine detail,as compared to other animals. Human and non-human primates have a moredeveloped and therefore more discriminating grip than other animals. Thisincreased capacity for discrimination affords the kind of detailed explorationand differentiation referred to above. Human and non-human primates alsohave a more highly developed pre-frontal cortex. The combination of these twocapacities of the hand and the pre-frontal cortex means that humans and othernon-human primates have enhanced capacities both for construing more andmore salient and, hence, potentially relevant and meaningful differences - bothtopological and typological - in the world.

The pre-frontal cortex is the most recent product of primate, includinghuman, cortical evolution. It is a further cortical development of the motorsystems responsible for direct motor control. This means that it affords abstractsymbolic exploration and investigation of the environment in the absence ofactual movement. It is this capacity for abstract exploration which, in turn, allowsthe organism to place itself in the position of others and, therefore, to viewthings as others do, as well as to abstractly look ahead and reflect on possiblecourses of action and their consequences before these are put into action asactual movement and behaviour. The interpretation of another's gaze vector is acase in point (see section 9, pp. 201-2). Our actions in the world and our experiential construals and perceptual-motor categorizations of the phenomena ofexperience through sensori-motor activity and exploration are, then, increasingly abstracted from sensori-motor activity so as to produce symbolic 'thought'.We see, then, that direct sensori-motor exploration, along with explorationwhich is abstracted from this in the form of 'thought', are capable of acting onand construing and constructing the phenomena of experience in our 'inner'and 'outer' worlds. The greatly enhanced capacity for such exploration inhumans on account of the resources of symbolic memory and social meaningmaking means that our very being - at once social and biological - is linked tothis capacity.

The question of the embodiment of the brain raises the question of thestructure of the brain and its relationship both to the body and to the ecosocialenvironment which lies 'outside' our bodies. The body is represented in the neocortex both in terms of the exploratory motor activity of perceptual systems andthe performatory activity of executive systems (Gibson 1983 [1966]: 46; Edelman1992: 105). With reference to the latter, it is significant that the face and thehand are massively and disproportionately represented in the neo-cortex withrespect to other areas of the body (Thelen and Smith 1994: 136-7). The reasonfor this lies in the fact that a quite large proportion of the motor cortex isdedicated to the control ofthese two parts of the body, both ofwhich are capableof very sophisticated and finely differentiated exploratory (and performatory)activities in comparison with other parts of the human body. They are thuscapable of discriminating very many more degrees of difference than are otherhuman body parts. Take, for instance, the crucial case of the human hand withits opposable thumb. This feature of the human hand affords a vast number ofpossibilities for the exploration and manipulation of objects, for tool use, for the


symbolic construal (imitation, etc.) of actions and objects which are not present,for pointing and imitation, and for the articulation of gestures and the realization oflinguistic meanings in sign Uohnston 1992; McNeillI992). Similarly, thehuman vocal tract and its cross-modal connections to facial movements alsoafford greatly expanded possibilities for articulating very finely differentiatedexploratory and performatory activity.

In the case of writing, hand-arm-joint-eye kinaesthesis (Gibson 1986 [1979]:275) enables the writer to interact with others across very diverse spatial andtemporal scales. The ability of the hand to articulate complex motor acts orgestures and the resulting graphic traces on a treated surface means thataddressers and addressees can integrate meanings across diverse spatio-temporalscales and assimilate them to their own Innenwelt. Moreover, the complex motoractivity which is involved in human speech can be seen as a further developmentand extension of the human ability to analyse phenomena into their componentparts and to interpret them. It is the interplay between face and hand, on theone hand, and the brain, on the other, which has enabled human beings tomassively expand their meaning-making potential and, hence, to intervene in,shape, and expand their Umwelt across diverse space-time scales (Latour 1996a,1996b; Lemke 1999; in this volume chapter 3, section 1, pp. 108-9).

Symbolic consciousness and the body parts which afford active exploration ofthe environment enable humans to orient to the indeterminate and theunknown (Prodi 1987: 20). The hands and face are able to discriminate verymany degrees of (topological and typological) difference. The delicate musculature of both the face and vocal tract and the hand-thumb system are able tomake and discriminate complex systems of differences and their combinationsand re-combinations according to the requirements of specific ecosocialcontexts. This is so of both the performatory activity of these bodily systems quaexecutive systems, as well as the exploratory activity of them qua perceptualsystems. In this way, social meaning-making becomes possible. From an evolutionary perspective, this discriminatory capacity was further enhanced by thedevelopment of a vertical posture: this development allowed for an enlargedvisual field, and at the same time the upper body was no longer tied to therequirements of moving around in the environment. Instead, the upper bodycould dedicate itself more and more to the selective orientation to and exploration of the environment in collaboration with the various perceptual systemssuch as the haptic, olfactory, auditory, and visual systems (Gibson 1983 [1966]:72-4). The increasingly delicate and discriminating patterns of difference thatcould be made in and through these bodily resources enabled emergentmeanings to be expressed. These meanings could, in turn, be addressed to andshared with others on the basis of higher-scalar boundary conditions whichemerged as a consequence of the interactions between the organism's internaldynamics and its ecosocial environment. In this way, the same patterns ofdifference are constrained by the same attractor such that they could be takenup and adaptively modified or re-worked by others. The taking up of the samepatterns by others meant that the meaning-making potential of these patterns ofdifference was able to link individuals across diverse space-time scales. Thislinking of individuals across diverse scales is, in the first instance, possible


because there always were precursor higher-scalar attractors and boundaryconditions which functioned to entrain human body movement in sociallysignificant ways. In this way, systems of differences become socially shared andconstrained by virtue of their being entrained to the same higher-scalar attractoron the ecosociallevel (Salthe 1993: 270, 277-81; Lemke 1995b: 116).

The synergy of hand and eye in hand-arm-joint-eye kinaesthesis meant thathumans could explore, act on, interact with, intervene in, and experiment withthe emergent possibilities of an unlabelled world. Moreover, the cross-modalsynergy of hand and language in both speaking through the co-contextualizationof vocal-tract and manual-brachial gestural activity and writing through the cocontextualization of hand and eye in hand-arm-joint-eye kinaesthesis and thevisual-graphological traces on a treated surface vastly extended this capacity. Theemergent and increasingly discriminatory capacities of this kinetic interface inthe physiological processes of articulation in speaking, writing, and gesture are,as I mentioned above, related to the massive representation of the activities ofhand, face, and vocal tract in the neo-cortex. The connection of such bodilyprocesses to activity in the neo-cortex, as evidenced by the massive neural activityassociated with the use of these body parts, suggests an intimate connection toconsciousness, and especially to symbolic or higher-order consciousness. Thus,the ability to intentionally control and modulate the activity of the hand, faceand vocal tract goes hand in hand with higher-scalar cortical activity. This, inturn, suggests an intimate tie-up between consciousness and those parts of thebody most closely associated with symbolic meaning-making.

The very fine discriminations made by vocal-tract and other (e.g. manualbrachial) articulatory activity can be co-ordinated with the articulatory activity ofothers in both space and time. Furthermore, the integration of these motorfunctions and the discriminatory potential that they afford to higher-scalarsemantics in symbolic neural space means that the exploratory activity ofvocal-tract activity, as well as that of the hand, has implications for higher-orderconsciousness. The very fine discriminations made by vocal-tract activity, theface, and the hand are all forms of expression that can be recategorized ascontent (linguistic or otherwise) by virtue of the brain's capacity to act as asomatic recognition system. This is no less true of inner speech, which involvesthe categorization of the brain's own activity as imagined motor routines corresponding to the expression stratum of speech, rather than the categorization ofinputs from external sources of auditory and other perceptual informationinvolved in listening to another's speech (chapter 5, section 9, pp. 271-2). In thecase of inner speech, this means that the brain's own internal activity categorizespast samplings of speech sounds as specific types of phonological routines thatcan be activated by the brain's own internal activity and matched with the appropriate structures on the content stratum by virtue of the brain's capacity tofunction as a somatic recognition system. This capacity therefore applies both tothe categorization of the perceptual inputs from external sources of speechsounds and to the phonological routines that occur in different global mappingsin the brain as a result of the brain's internal activity. In all these cases, the finedifferentiations made by the vocal tract, the face, and the hand in the process ofexploring and/or acting in the environment have contextual redundancies with


categories of content to which they are integrated. In this way, the exploratorypotential of these parts of the body has direct implications for the categories ofconsciousness itself.

Thinking, as I pointed out above, is symbolic exploration abstracted fromsensori-motor activity. However, the fact that, in thinking, direct exploration ofthe environment through sensori-motor activity does not take place, does notmean that thinking is not also a form of activity. In my view, thinking is bothaction and reflection, to use Malinowski's (1923, 1935) terms, which has beenspecialized to the 'inner' domain. It is a kind of action and reflection which cantake place without directly subjecting the body to the hazards and risks entailedby a given course of material action. Verbal thinking - that is, language activityspecialized to the inner realm of 'mind' - means that reflection on action andits consequences, the discourse genres of, say, abstract reasoning and argumentation, particular ways of interpreting and understanding, and so on, are not thesole property of a single individual. Luria describes syllogistic reasoning as aform of higher-order rational thinking in precisely these terms:

One of the objective devices that arises in the process of the development ofcognitive activity is the syllogism - a set of individual judgments of varyingdegrees of generality in certain objectively necessary relationship to oneanother. Two sentences, of which the first ('precious metals do not rust') is inthe nature of a general judgment ... while the second ('gold is a preciousmetal') is a particular proposition, are not perceived by the developedconsciousness as two isolated phrases in juxtaposition. A human being whosetheoretical thought processes are well developed will perceive these as acompleted logical relation implying the conclusion, 'hence gold does notrust.' This conclusion does not require any personal experience; it is arrivedat through a syllogism created objectively by historical experience. A considerable proportion of our intellectual operations involve such verbal andlogical systems; they comprise the basic network of codes along which theconnections in discursive human thought are channelled. (Luria 1976: 101)

The 'networks of codes' are the discourse genres which give determinate shapeto 'discursive human thought'. The 'completed logical relation' between theimplied conclusion and the first two parts of the discourse structure is a logicosemantic one which can be glossed as CONDITIONACONSEQUENCE. Thus,the rational thinking which is involved in logical reasoning and inference ismediated by genres structures which become internalized as procedures forhigher-order thought. Furthermore, what may appear, viewed in this way, as themonological property of individual consciousness is, in actual fact, a dialogicalresponse to some prior situation or text. The tendency to naturalize these internalized genres as the monological properties of a single consciousness isreinforced by the tendency to consider written texts as the unique property of asingle individual. The fact that the first two sentences are put in a particular kindof logico-semantic relation with the third one also shows the internally dialogicnature of the text as a whole. Thus, the first two constitute a particular generalizing stance on a given situation-type while the third sentence is a particular


dialogical response to this situation from another point of view, i.e. from thepoint of view of the specific characteristics of gold as an instantiation of theschematic category 'precious metal'.

Symbolic thinking, in the form of, for example, 'inner speech', links theindividual to the accumulated ways of thinking built up over many generationsin the course of the constant and adaptive modification of thought on the partof all of those individuals who have contributed to it. Symbolic thinking, whichis thinking mediated by a semiotic system of linguistic or other potentiallymeaningful differences in a given community, affords differentiated thought andthus extends and enhances our ability for active exploration and expansion ofour Umwelt. In this view, thinking is a specific use of neuronal capacities forgiving pattern and meaning to experience and for organizing one's responses toevents in both our inner and outer environments as specific courses of action.

As Saussure demonstrated in his discussion of the mediating influence of thevalue-producing categories of a given language system on 'thought' and 'sound'(chapter 2, section 1, pp. 59-67), such mediation enables new possibilities ofmeaning to emerge from the vague and 'amorphous' flux of pre-semiotic'thought' and 'sound'. That is, we produce our world in and through oursemiotically mediated interventions in and engagements with this vague andundifferentiated mass of possibilities, which we can never, in any case, know orexperience directly, but always only mediately. The projection into the world ofthe highly delicate differentiations afforded by the sensori-motor activities of thehand and face-vocal-tract systems also means that we endow the phenomena ofour experience with value, affect and motivational salience. That is, we assignthem with value in terms of what they entail for us as action and interpersonalinteraction (Thibault 1992b).

5. Language Functions and the Cortical Organization of the Brain:Implications for Higher-order Consciousness

The human brain is organized in terms of two cortical hemispheres. The righthemisphere regulates our responses to the unknown; the left is more concernedwith actions which are carried out in response to the known and familiar.Researchers such as Davidson (1984,1992) have shown that the two hemispheresare differentially specialized as regards value and affect in connection with theway the individual orients to the world. Positive affect (e.g. attraction) correlateswith increased neural activation in the left frontal cortex. Negative affect (e.g.repulsion) correlates with increased neural activation in the right frontal cortex.This suggests that the two hemispheres differentially regulate our affective andvalue-laden responses to the world. The right hemisphere regulates ourresponses to that which has not yet been assigned a specific value or motivationalsalience. The left hemisphere, on the other hand, is concerned with our orientation to that which has already been assigned a specific value and motivationalsalience. In the first case, the negative evaluation of the uncertainty andindeterminateness which the individual encounters constitutes the basis onwhich active exploration may occur. In the second case, we are on familiar, safe


ground and already know what to expect and therefore how to act. In both cases,value-laden responses are linked to sensori-motor activity and, therefore, tospecific courses of action and their likely consequences.

The lateralization of brain functions has frequently been cited as evidence forthe specialization of language functions to the left hemisphere of the brain. Inthis view, the right hemisphere is specifically adapted for the production ofvisualand auditory images, for global pattern recognition, holistic thinking, and so on.Table 4.1 presents the lateralization of brain functions in the left and right hemispheres according to Deacon.

The right hemisphere appears to be associated with topological-continuousvariation, with the vagueness of the not-yet-fully specified or determinable, thussuggesting what Salthe calls 'signifying possibility rather than necessity' (1993:143) on account of its indeterminate character. The left hemisphere, on theother hand, is specialized to the typological-categorial analysis of phenomenainto discrete elements - that is, with the specification of that which is determinate. Typically, the left hemisphere has been seen as specialized to languagewhereas the right hemisphere is considered to be concerned with holisticpatterns and images. (See for example Corina et al. 1992 for arguments in favourof the view that left hemisphere specialization of language respects a distinctionbetween linguistic systems such as American Sign Language (ASL) and nonlinguistic gestures.) However, Deacon (1998 [1997]: 311) argues that thehypothesis of hemispheric specialization for language is erroneous and that whatis really at stake is a question of the way different linguistic and other brainfunctions are differentially distributed between the two hemispheres (see alsojakobson and Santilli 1980; Cowley 2002: 79-80). Deacon sees this as 'anotherreflection of the role of competitive processes in determining the ultimaterepresentation of functions in the cerebral hemispheres during development'(1998 [1997]: 311).

Table 4.1 summarizes Deacon's observations concerning the typical ways inwhich various language functions tend to be distributed across the left and righthemispheres in the development of perhaps the majority of human beings. Aclose examination of the various items in Table 4.1 reveals that all of thefunctions that are presented in this table are intrinsic to language form andfunction. Moreover, the functions that are prototypically associated withlinguistic processing and assigned to the left hemisphere, in actual fact,represent a very partial and distorted view of language. It is a view which is basedalmost exclusively on formalist criteria, as well as on a narrowly mono-functionalview of language as being about information transmission or the cognitive,ideational, or representational function of language. My point is that all of thefunctions presented in Table 4.1 are relevant to language. It follows from thisthat language is globally distributed over both hemispheres of the brain.Therefore, the whole brain is involved in the functioning of language in theindividual (see also Peng 1994: 110).

Table 4.1 shows that a number of different, though complementary,functional semantic domains ofianguage are differentially distributed across thetwo hemispheres. Given the limitations of the model of language which inspiresDeacon's account, it seems necessary to explore even more deeply the implica-

THE SEMIOTIC BASIS OF CONSCIOUSNESS

Table 4.1: Lateralization of language functions in the left and right hemispheresaccording to Deacon

191

Left Hemisphere Right Hemisphere

Word and phase-level processing and Large-scale semantic processing ofanalysis language of larger symbolic constructions

that words and sentences contribute to:complex ideas, narratives; and arguments(Deacon 1997: 312); compositionalcomplexity

Contextual logic (re jokes, humour)(Deacon: 1997: 312)

Specific details Implicit organizational logic (Deacon 1997:313)

Phonological processing Processing of prosodic features of speech(rhythm, pitch, emotional tone) (Deacon1997: 313)

Phonemic processing Correlation of changing pitch, volume, andrate of production with level of arousal;changes in quality of vocalization asindication of type of interaction (hostile,submissive, ete.); overall phrasing withrespect to breath control (Deacon 1997:314)

Ultra-rapid analyses of sound changesand control of rapid, precise, skilledmovement sequences

tions of his argument that the human brain eo-evolved with language rather thandetermining the nature of human language. From this perspective, it makessense to ask how the intrinsic design features of language form and function arereflected in the human brain. In systemic-functional linguistic terms, Table 4.1suggests that the different metafunctions are differentially specialized to the twohemispheres. Following Deacon, this does not mean that the relationshipbetween linguistic metafunction and brain hemisphere is fixed. Instead, itdepends, as Deacon suggests, on developmental factors rather than on in-builtgenetic ones. Moreover, the evidence he cites concerning bi-lingual hemisphericspecialization in trained interpreters also suggests the role of individuation inthe way language functions are distributed across the two hemispheres in thebrain.

Following Deacon's suggestion, we can hypothesize that diverse metafunctional domains are differentially distributed across the left and right hemispheres in the interests of overall processing efficiency in the brain. The righthemisphere is specialized for the interpersonal and textual metafunctions; the


left for the ideational (experiential and logical) metafunction. The differentialdistribution of the metafunctions across different cortical hemispheres showsthat meaning is a global level of organization of the brain's self-organizingdynamics. Furthermore, the global cortical distribution of the metafunctionsconstitutes a means for integrating language forms both with the body-brain andwith the world in ways which reflect the intrinsic organization of language itself.If language is intrinsically organized in ways which integrate it to the dynamicsof the body-brain and to the world, then this suggests that the metafunctionalbasis of language and other forms of semiosis has important implications for theself-self transactions on which consciousness is founded. This also raises thequestion as to whether different individuals have different personal dispositionsto different metafunctional dimensions of linguistic meaning, along withdifferent dispositions to integrate language with other non-linguistic semioticmodalities, or even different dispositions towards a given semiotic (e.g. visual,linguistic, kinesic) modality with respect to others.

In fact, language is never independent of other semiotic modalities.Language and other semiotic modalities are always eo-deployed in an integratedway. Spoken language is combined with manual-brachial gestures, posture, gaze,body movement, facial expressions, and so on. Each of these semiotic modalitiesco-contextualizes the others in the making of some discursive event. Cocontextualization means that one modality is not simply added to some other.The assumption that they are simply added to each other rests on theassumption that the different modalities are separate and independent to startwith. Rather, different semiotic modalities have both eo-evolved with each otheron the phylogenetic timescale and eo-developed and co-individuated on theontogenetic scale. They also co-contextualize each other on the timescale ofparticular discursive activities and the textual productions that derive fromthese. The co-contextualization dynamics of the diverse semiotic modalities isshown in the ways in which they form typical eo-occurrence patterns which arerecognizable as being typical or regular combinations in some community.Moreover, the combining of an item from one modality with an item from someother can alter the meaning that each item has on its own. Drawing on someobservations of Bateson (l973e: 319-20) concerning the combined effects ofmultiple changes on somatic flexibility, Lemke (1998) has drawn attention to themultiplicative character of meaning-making: in multimodal forms of discourseand/or text different semiotic modalities co-contextualize each other, ratherthan being merely added to each other. The combined effect of the variousmodalities is multiplicative rather than merely additive.

The multimodal character of meaning-making activities is commensurate withthe global organization of meaning in the brain. In this perspective, the globalcortical distribution of the metafunctions provides support for the hypothesis ofa more general metafunctional basis to all modalities of semiosis and theirorganization in the brain. In systemic-functional theory, there is a growing bodyof evidence that the metafunctions constitute the intrinsic design features ofother modalities such as depiction (O'Toole 1994; Kress and Van Leeuwen 1996;Lemke 1998; Thibault 2000b, 2001), movement (Martinec 1998), gesture, andsign language (Johnston 1992). This does not imply that the typological charac-


tensncs of language can be used to explain the very different topologicalprinciples of organization of, say, depiction. The fact that every discourse eventand every text is multimodal derives from the ecosocial reality that meaningmaking events and texts are always simultaneously material and semiotic(chapter 1, section 2, pp. 8-11). Material and semiotic processes are always crosscoupled in meaning-making activity. The distinctions we make in our theoriesbetween language, depiction, sound, gesture, and so on, qua distinct systems ofdifferences (typological or topological) are abstractions from the materiality ofall meaning-making activity. Abstract systems of differences define and crossclassify types of relationships on the basis of multiple hierarchies of metaredundancy relations (Lemke 1984c: 35-9). Yet, meaning-making is alwaysgrounded in material couplings, the many degrees of topological variation ofwhich can never be exhausted by any system of typological-categorial distinctionssuch as the theoretical abstraction of a language system that is uncoupled fromthe material. It is this inexhaustibility of the topological-continuous ground ofsemiosis which allows for new meanings to be made. This is because semiosis isnever coupled to the material in entirely predictable ways. Therefore, new eocontextualizing relations across modalities can emerge at the same time as new,previously latent material possibilities can be semioticized. That is, they canbecome differences that make a semiotically salient difference to the membersof a given community. The global integration of meaning across the right andleft cortical hemispheres shows that the brain is actively involved in the integration and contextualization of the material-semiotic couplings referred to here.The global, 'imagistic' or topological-continuous character of right-hemisphereprocesses suggests that this hemisphere is concerned with partly nonsemioticized material couplings whereas the left hemisphere is concerned withthe semiotic categories that are involved in their semioticization. The metafunctions, now generalized as an organizational principle that is common to allmodalities of semiosis, can be seen as a principle of integration whereby diversesemiotic modalities are integrated and instantiated at the level of situationalcontext or social activity. There is not, therefore, a one-to-one correlationbetween a given metafunctional selection in, say, language and a givencontextual feature. Instead, contexts are semiotic constructs which are definedby global combinations of resources from different semiotic modalities and theirmaterial couplings. In this way, social activities and contexts, which are globalconstructs, are themselves organized according to metafunctional principles.The brain is thus involved in the process of contextualizing the ways in which themetafunctional components of different semiotic modalities are eo-deployed inorder to produce a particular action on the basis of specific semiotic andmaterial couplings. Therefore, the ways in which gesture, pointing, gaze,posture, and other bodily activities are integrated with language and with eachother in speech can be seen to be a consequence of the ways in which verygeneral semiotic principles of metafunctional organization in symbolic neuralspace constrain and entrain to specific action trajectories the differentcomponents of all the meaning-making and material resources that are eodeployed in the enactment of that activity.

Table 4.2 presents four accounts - two recent, two not so recent - of language


Table 4.2: A comparison offour accounts of language form and function in context,showing an emerging consensus concerning the multifunctional nature of languageand its contextual motivation

Malinowski Biihler Halliday Silverstein

language as action symbolic field: in terpersonal interactional(l) expression/ metafunction coherenceappeal

language as (2) representation experiential/logical de notationalreflection metafunctions coherence

context of situation deictic field textual metafunction entextualization/contextualization

form and function. The four accounts show a striking convergence, in spite ofmany differences in detail, in their overall conception of the relationshipbetween language form and the meanings realized in its contexts of use. Inemphasizing the convergence on the three general types of meaning in thesefour accounts, I am endeavouring to suggest the possibility of a broaderconsensus regarding the significance of the four meta functions - experiential,interpersonal, logical, and textual - for the shape of consciousness.

Language is a multidimensional semantic space in which diverse functionalregions are simultaneously realized in the same lexicogrammatical or textualform. The fact that there is an emerging consensus on the multifunctionalcharacter of language, if not on all the specific details of how the variousfunctions are realized in the formal organization of language, has importantimplications for a theory of consciousness. If symbolic consciousness is semiotically mediated, then it becomes necessary to theorize consciousness on the basisof a semantically enriched view of both language and consciousness. Generallyspeaking, the models of language which inform the discussion of consciousness,when language is considered relevant at all, are narrowly focused on 'referential', 'cognitive' and 'conceptual' theories of meaning. These theories arefounded on the formalism of the sentence and the forms of propositionallogicthat are associated with formal logic. But semiotic mediation is a discursiveprocess; it is linguistically mediated social activity and/or text which areintegrated to their contexts. Moreover, social activities and the texts that derivefrom these are shaped by genre forms and conventions according to the requirements of specific social situation-types. This fact requires that the full metafunctional diversity oflanguage form and how this relates to the contexts of both thebody (expression) and of our experience of the world (content) needs to be partof the explanation of consciousness itself.

In the different, though related ways, and to varying degrees of generality andabstraction, the functional theories of Malinowski, Buhler, Halliday, andSilverstein draw attention to the fact that context can never be defined in terms


of a direct relationship between any given functional component in languageand the activities that language and other semiotic modalities enact. Rather,social activities and the situational contexts in which action unfolds eo-deployresources from diverse semiotic modalities such that a particular experientialselection in language can be eo-deployed with, say, an interpersonal selection ingaze that has implications, not for the given linguistic or gaze selection as such,but for the global meaning of the activity in which these features are combinedand embedded. This is possible because particular experiential and interpersonal values from different semiotic modalities cross-classify each other at thelevel of the activity structures in which they occur. The ways in which selectionsfrom different semiotic modalities co-contextualize each other in this waydepend on the contextualizing relations that define the activity structures of agiven community. The fact that action is enacted by diverse semiotic resourcesand that these semiotic modalities are globally integrated by general metafunctional principles that are found in different semiotic modalities, albeit accordingto different principles of formal realization, provides a powerful argumentin favour of the view that the metafunctions are homologous with the globalorganization of meaning in the brain.

According to the metafunctional theory of language developed by Hallidayand others, language is a resource for construing human experience, forenacting interpersonal relations, for giving coherence and texture to ourmeaning-making acts and the ways these connect to their contexts, and forconstruing logical and temporal relations between events. Table 4.2 comparesfour functionally motivated accounts oflanguage form in order to suggest a basisfor such a consensus.

The proposal that consciousness is shaped along metafunctional lines drawsattention to the fact that a number of interrelated factors are all at play, simultaneously. First, one is aware of some object of consciousness as an object which isconstruable as belonging to such and such a category of experience. The idea of acategory of experience refers to a difference which make a difference - topologicaland!or typological - at any given level of the implication hierarchy of icon, index,and symbol. There is no suggestion that such categories necessarily correspond toa linguistic term, though they may. The experiential resources of language are aresource for construing phenomena of experience as semantic 'objects' of consciousness. Secondly, consciousness entails an active and intentional orientation tothe objects so experienced. Thirdly, consciousness is experienced and felt as atextured whole which relates self to its surroundings in a situated and unified way.These three aspects of consciousness will be discussed in sections 6, 7, and 8 belowin relation to the metafunctional organization of language.

6. Experiential Meaning and the Assimilation of the Phenomena ofExperience to Knowable Categories

A given object of consciousness is not simply a representation of somephenomenon. Perceptual, conceptual, and semiotic categories construe specificinstances as being instantiations of this or that category. In the case of language,


the experiential resources of the clause construe some phenomenon - realor imaginary - as an instantiation of a given category of process-participantconfiguration in the clause. The term 'construal', as distinct from the notion ofrepresentation, therefore refers to the ways in which the phenomena of ourexperience, including perceived events as well as imagined, remembered, orhypothesized ones, are construed as instances, to varying degrees, of thecategories of language or some other meaning system. A process-participantconfiguration is a semantic figure, as defined by Halliday and Matthiessen, whichis realized grammatically by the clause. For example, the clause Two hundredyears ago, rococo was all the rage in Europe consists of the Process-ParticipantCircumstance configuration Circumstance: Time-Participant: Carrier-Process:Attribution-Attribute: Circumstantial. Experiential categories such as the ones inthis example symbolically construe the phenomena of our experience as instantiations of the categories that belong to a particular language system. The notionof representation may suggest that language represents something which is priorto it. For this reason, the term 'construe' is to be preferred.

The experiential metafunction in language is concerned with the symbolicconstrual of experience as categories and relationships of experiential meaning.This metafunction interprets the phenomena of both inner and outerexperience in and through the experiential structures and categories of theclause and clause complex. Halliday and Matthiessen have formulated thisrelation as follows:

A phenomenon is the most general experiential category - anything that canbe construed as part of human experience. The phenomena of experienceare of three orders of complexity: elementary (a single element), configurational (configurations of elements, i.e. a figure) and complex (a complex offigures, i.e. a sequence) ... (Halliday and Matthiessen 1999: 48)

Sequences, figures, and elements are semantic constructs. As Halliday andMatthiessen point out, their typical realizations in the lexicogrammar are asfollows:

sequence ~ clause complexfigure ~ clause

element ~ element in structure of clause (group/phrase)

Consider the following example:

Sequence:

When travelling to Fassifern by bus, buy a bus ticket to Fassifern.

Figures:

When travelling to Fassifern by bus

buy a bus ticket to Fassifern

Elements:

when, travelling, to Fassifern, by bus, buy, a bus ticket, to Fassifern


In this example, the two figures are related to each other in a sequence by thelogico-semantic relation of CONDITIONAACTION. The individual figurescomprise the following experiential configurations:

Actorol'rocess: Material: Action-vCircumstance: Location/Circumstance:Manner (clause 1)

Actor/Process: Material ActionAGoalACircumstance: Location (clause 2)

Both of these figures construe the 'you' (the addressee) as the Actor whoperforms a sequence comprising two actions in determinate circumstances oflocation and manner. The central construct in the semantics of the experientialcomponent of the lexicogrammar is the figure. A figure is a symbolic construalof some phenomenon of experience in the inner or outer domains - real orimagined - as a configuration consisting of a process, one or more participants,and associated circumstances. This is not a question of some direct relationshipof representation or correspondence between phenomena (entities,happenings, events, ete.) in the world and selected aspects of language form.Rather, the phenomena of experience are symbolically construed by the experiential categories according to a relatively small set of process-participantcircumstance configurations in a given language system. For example, in thegrammar of English, a very small set of process types - being/having, doing,sensing, and saying - configure with their respective participants to form acluster of related semantic models of human experience. The figure is based onthe following principle: the figure as a whole is comprised of a configuration ofparts which function in the whole to which they belong. The semantic figure isthus analysed into its constituent parts and the functional relations whichoperate both between the parts and the whole and between the various parts.That is, the configuration is comprised of part-whole and part-part relations.

The notions of sequence and figure, as defined by Halliday and Matthiessen,show how the constituent elements of, say, a given figure cannot symbolicallyconstrue the phenomena of experience unless they are placed in what Biihler(1990 [1934]: 206) called the 'symbolic field' of language. The symbolic field ofexperiential meaning is constituted by the configurations of figures andsequences of these such that their constituent elements receive their symbolicfield value on the basis of the larger-scale configurations and sequences they participate in. The individual elements alone, independent of the symbolic fields inwhich they are placed, do not symbolically construe the phenomena of humanexperience. This important fact highlights the misleading nature of the claims asto the symbolic value of words per se. Thus, the noun 'cat' per se does notsymbolize a particular phenomenon of experience - i.e. a given instance of thetype-category [CAT]; instead, it symbolizes the schematic category [CAT]whereby all instances are already fully specified to some relevant degree ofgenerality. Thus, the noun 'cat' does not refer to or instantiate a particularindividual cat in some referent situation. The symbolic construal of thephenomena of experience depends on further specification of the given sign.Such specification is, in the first instance, provided by the symbolic field of the


figure (and sequence) in which the given element is located and in terms ofwhich it derives its field value. In the case of experiential meaning, the constituent elements of the figure, say, are grounded in their symbolic field valuesthrough the transitivity relations of the type of process-participant configurationin which they participate, as we saw above. The symbolic construal ofphenomena through the experiential categories of a given language systemimplicates the processes of naming, classifying, and taxonomizing. In this way,the phenomena of experience are selectively attended to and categorized so thatthey may be assimilated to the domain of what is known and, hence, in somesense already familiar.

7. Interpersonal Meaning as Exploratory and Orienting Activity in Relationto the Ground

However, there can be no symbolic construal of the phenomena of ourexperience by means of the experiential resources of language without the givenconstrual also being given a determinate point of reference in the here-&-now ofthe speech event. The here-&-now of the speech event is the ground (Langacker1991: chapter 6; see also Halliday 1994 [1985]: 75) in relation to which the experiential construal in the clause is given a point of reference. The clausallyrealized experiential figure is grounded by the resources of the Finite elementin the verbal group. In English, the Finite grounds the particular processparticipant configuration in terms of either primary tense or modality. Primarytense is what Davidse (1997) calls 'temporal proximity deixis': it locates theprocess-participant configuration as an instance in time relative to the ground the here-&-now - of the utterance. Modality, following Davidse, is a form ofmodal proximity deixis. It grounds the given process-participant configurationwith reference to the addresser's evaluation, in the case of declarative propositions, and the addressee's evaluation in the case of interrogatives, as to howlikely, certain, possible, usual, obligatory, necessary, and so on, the processparticipant configuration is. Again, these modalized evaluations are made withreference to the ground, as defined above. Moreover, both temporal proximitydeixis (primary tense) and modality proximity deixis (modality) ground theproposition (declarative or interrogative) in the way described here in theperspective of the selves who are dialogically co-ordinated by the exchange ofmeaning which takes place in and through the uttering of the given proposition.

In the clause, 'Two hundred years ago, rococo was all the ragein Europe, the declarative proposition is located as being temporally remote from the ground (pasttense). Davidse, following Langacker (1991: 93), points out that the Finite codessubjectively. That is, the Finite locates the proposition in terms of the addresserand/or the addressee's subjective orientation to it along the parameters oftemporal and/or modal proximity. That is, the addresser and/or the addresseeis positioned as being subjectively near to or far from the processparticipant instance. In this sense, the process-participant configuration can besaid to be grounded in the perspective of the SELF. The notion of SELF is beingused as a cover term here to refer to either the addresser or the addressee in the


exchange. In the above example, it is the writer, in the first instance, who locatesthe proposition as being temporally remote from him or her, Le. some twocenturies ago in the past. However, the addressee also orients to the propositionand re-grounds it in his or her perspective so that you and I qua readers, forexample, also orient to the proposition as being about events which aretemporally removed from the here-now ground of the act of reading. In this way,the addressee is able to dialogically co-ordinate his or her self-perspective withthat of the addresser. This means that the resources of temporal and modalproximity deixis in the Finite enable addresser and addressee to orient to theproposition in terms of the same temporal or modal frame of reference withrespect to the ground. Thus, even though I may read the above clause, whichoccurred in an article in Cathay Pacific's in-flight magazine, in a context whichis different from the original one of its writing, I can re-ground the clause in myperspective because I can subjectively orient to the temporal proximity deixissuch that the referent situation specified by the process-participant instance isunderstood as being temporally remote from me as well. This means that I canmake that referent situation a symbolically construed object of consciousness inthe perspective of my own self at the same time as I dialogically orient to it.

The process-participant configuration is also grounded, Davidse argues, byvirtue of its being tied to the Subject of the clause. In other words, the clause isgrounded with reference to the person deixis of the Subject. It is tied to itsgrammatical person, which can be first, second, or third. Person deixis links theprocess-participant configuration to the intersubjectively and dialogically coordinated I-you axis in terms of which the exchange of meanings take place insome determinate context. The I-you axis is, therefore, a further form of groundin terms of which the proposition is referenced. In the above example, theSubject rococo is a third-person participant that is located as being outside theintersubjective I-you relation in terms of which the exchange of meanings isnegotiated between I and you. In other words, it is specified as being remotefrom the ground in the sense that it is a third-person other. Once again, we cansee how the grounding function of the Subject takes place within the perspectiveof the selves who are implicated in the exchange process. In this case, rococo isnot a first- or second-person participant in the exchange of meanings. Instead, itis marked as being removed from this, i.e. as belonging to the third-persondomain of the other to which I and you orient. The Subject objectively groundsand actualizes the proposition by instantiating the entity - rococo in the aboveexample - in terms of which the proposition is grounded. That entity can be theaddressee (I) or the addressee (you), or it can be objectively removed from theI-you ground and grounded as a third-person other which addresser andaddressee jointly attend to as the element in which the speaker invests the modalresponsibility of the clause. The Subject is the entity on the basis of which thespeaker expresses a modalized perspective in the proposition and in termsof which the addresser seeks to influence the addressee's take on the sameproposition.

The interpersonal metafunction is what Halliday (1993: 103) defines as the'active' principle of language. It is language functioning to enact interpersonal,dialogically organized interactions between some 'I' and some 'you'. The


interpersonal metafunction is therefore centrally concerned with language asaction. Closely linked to this is the fact that the resources of the interpersonalmetafunction are fundamentally concerned with systems of social values wherebyinteractants orient both to each other in and through their interaction, to theutterance itself, and to still wider systems of social values that the utterance mayinvoke. In other words, interpersonal meaning is concerned with the linkbetween value and action and the implications which the former has for thelatter when interactants orient to and engage with the world of the nonself(Thibault 2002). This further implies that the interpersonal metafunction isconcerned with exploratory activity whereby the individual negotiates andengages with the indeterminate, the not-yet-specified, and the unknown. Thus,linguistic propositions (declarative, interrogative) and proposals (oblative,imperative) are a means of enacting this kind of exploratory activity. In theprocess, the individual generates value-laden hypotheses about the unknownwhile assimilating it to some more determinate experiential framework (see alsoHalliday 1993: 102).

Thus, interpersonal exploration of and engagement with the unknown, theunfamiliar, or the indeterminate also leads to the increasing specification of thepreviously vague phenomenon through its construal by the experientialcategories of language. If, hypothetically speaking, language was interpersonalinteraction and action per se, there would be no resources for convertingtopological-continuous variation into more determinate and manageabletypological-categorial distinctions of the experiential kind. The existence ofthese typological-categorial distinctions means that diverse experiences can allbe categorized as being similar on account of the system of values of the givenlanguage system (Thibault 1997a: chapter 7).

8. The Textual Metafunction as Semiotic Means for Giving Unity andWholeness to Meaning-making

The textual metafunction is concerned with the way in which language isorganized as discourse which is operational in some context (Halliday andHasan 1976; Hasan 1980). It is dually concerned with (1) the ways in which thesign-tokens (words, clauses, ete.) which comprise a given text achieve a relationalcohesion with one another through such resources as topic maintenance andcontinuity, continuity of referents, and so on; and (2) the subjective orientationof interactants within what Biihler (1990 [1934]: 117) referred to as the 'deicticfield' of language. The deictic field consists, in the first instance, of the herenow-I system of subjective orientation to a co-ordinate source (e.g. the speakerof the pronoun 1) in some context of situation in which the given discourse isoperational. More generally, the textual metafunction is concerned with bothhow the various parts of the discourse are related to each other as a texturedwhole (a text) such that the resulting whole is perceived as having achieved adegree of relative coherence as a whole and (2) the ways in which the participants in the temporally unfolding interaction manage to achieve systematic andregular links between some emergent context and the unfolding linguistic or


other text. Moreover, it is important that the participants perceive the patternedcontextualizing relations between the two as (situationally, generically) appropriate or effective realizations of text-in-context (Silverstein 1997: 270).

9. Experiential and Interpersonal Meaning in Gaze

When we attend to someone else's gaze vector, we can see how similar metafunctional principles operate at the lower, less specified integrative level ofperceptual awareness. In attending to and attempting to interpret the target ofsomeone's gaze vector, perhaps in order to determine his or her interests orintentions with respect to the target, I implicitly analyse the structure of theother person's gaze in terms of the person who is the source of the gaze (Gazer),the directional vector of the person's gaze, and the Target or goal of the gazevector, understood as the entity which the Gazer is attending to. In this perspective, the gaze is analysed as a process-participant configuration on analogy withexperiential meaning in the clause. Table 4.3 provides an analysis of thisproto-experiential dimension of the meaning of gaze.

Table 4.3: Proto-experiential meaning in gaze vector

Participant

Gazer

Process

Directional vector of gaze

Participant

Target

The gaze is also related to the here-now ground relative to the observer (the self)who interprets the other person's gaze. In this case, the Gazer is the objectivegrounder or the actualizer of the gaze. The gaze vector that extends from theGazer's eyes to the target, on the other hand, subjectively grounds the gaze inthe here-now of perceptual awareness and perceptual modality (visual); theCazer's current interests, as interpreted by the observer; distance from theGazer; and the Gazer's possible subjective orientations to the Target. Forexample, is the Gazer looking at a member of the opposite sex because of sexualinterest in or attraction towards the Target? The Target, on the other hand, doesnot participate in these grounding functions, which are based on the Gazer-GazeVector nexus. Instead, the Target further specifies the overall process-participantconfiguration by specifying the specific entity that is the focus of the gaze. TheGaze-Gaze Vector nexus is, therefore, similar in function, though on a lowerintegrative level, to the Subject-Finite structure in the Mood element of theclause. However, gaze is much more immediately tied to its here-now ground inways that language is not on account of the way in which the interpretation ofsomeone's gaze depends on the interpreter's here-now perception of the other'sgaze. Table 4.4 illustrates the grounding function of gaze, as discussed in thisparagraph.


Table 4.4: Proto-interpersonal meaning in gaze vector, showing ground functions

Objective Ground

Gazer as objectivegrounder and instantiatorof gaze vector

Subjective Ground

Gaze Vector grounded inhere-now of perceptualact; indicating subjectivestance on Target (e.g.sexual attraction)

Target as specification ofentity that is focus ofattention

10. Proto-interpersonal Meaning and the Child's Exploration of ItsEnvironment

Interpersonal meaning, as we saw in section 7, is implicated in the initial stagesof the exploration of the vague, the unknown, and the indeterminate. It isconcerned with hypothesis formation. Consider the following revealingdiscussion by Halliday:

A human infant engages in symbolic acts, which I have referred to as acts ofmeaning. Children are predisposed, from birth, (a) to address others, and beaddressed by them (i.e., to interact communicatively); and (b) to construetheir experience (i.e., to interpret experience by organizing it intomeanings). Signs are created at the intersection of these two modes of activity.Signs evolve (a) in mediating - or, better, in enacting - interaction withothers, and (b) in construing experience into meaning; specifically, inexploring the contradiction between inner and outer experience (betweenwhat is perceived as going on 'out there' and what is perceived as going on 'inhere', within the child's own consciousness (cf. Trevarthen, 19S0).

Thus, typically, at 0; 3 to 0; 5 months (years; months) babies are 'reachingand grasping', trying to get hold of objects in the exterior domain and toreconcile this with their awareness of the interior domain (they can see theobjects). Such an effort provokes the use of a sign, which is then interpretedby the adult caregiver, or an older child, as a demand for explanation; theother responds in turn with an act of meaning. There has been 'conversation'before, but this is a different kind of conversation, in which both parties areacting symbolically. A typical example from my own data would be thefollowing, with the child at just under 0; 6 months (Halliday, 19S4a, p. 2):

There is a sudden loud noise from pigeons scattering. Child [lifts head,looks around, gives high-pitched squeak] Mother: Yes, those are birds.Pigeons. Aren't they noisy!

(Halliday 1993: 94-5)

In Halliday's example, it is the sudden loud noise made by the pigeons as theyscatter which impacts on the child's perceptual systems. This noise is an index information in Gibson's sense - of an environmental event. As such, it stands in


an iconic-indexical relation of necessity to that event. However, it is not simply aquestion of the brain's detecting and processing of specific (auditory) sensationswhich are, in turn, interpreted as representations of the world, and which theninform the mind as to what is going on in the world. The point is that once theinformation which is carried by the nerve fibres from the peripheral receptorsenters the central nervous system, it is no longer locked only onto the stimulus(Wall 1996: 176). In other words, it is no longer a question of a purely indexicalrelationship between environmental stimulus and the organism's response. AsWall writes, the information which is received by the brain 'depends on the contingencies of other events in the periphery and on the brain state which ordersand permits and shapes the arriving messages' (1996: 176-7). Thus, the sensorymotor response which is made is not passively locked into the stimulus, butdepends on an active process of selection in the brain (Wall 1996: 177). Thisprocess of selection means that the transduction of the stimulus informationwhich is picked up by the receptors results in its translation into various possibilities of sensori-motor response. This can only occur on the basis of a storedsystem of signs in the Innenwelt whereby the child uses stimulus informationabout environmental events as signs for interpreting the world. The child'sresponse to an environmental event - his squeak - has potential symbolic valuebecause there is no necessary relationship between this sensory-motor responseand the way it may be interpreted by his interlocutors. That is, the squeak isalready interpreted as belonging to a system of symbolic possibilities, howeverelementary. This means that the squeak derives its value from its position in asystem of elementary possibilities for meaning-making. Furthermore, the child'svocalization can be seen as a dialogically co-ordinated response to a particularnon-verbal aspect of the context of situation.

The system of symbolic possibilities which are stored in the child's Innenweltthus regulates his sensori-motor activity as value-laden possibilities for interacting with and dialogically engaging others. The value-laden character of thesepossibilities means that they are endowed with actional and affective significance. The squeak qua squeak is in itself an environmental event whichgenerates stimulus information about its source. Again, such information is bothiconic and indexical. However, the fact that the child's interlocutor construes thesqueak as a 'demand for information', as indicated by her own response, meansthat the squeak is also construed as having symbolic value. For this reason, thesqueak is not necessarily (indexically) locked into specific environmental events;instead, it affords possibilities for symbolic action which can be deployed inother contexts, in response to other environmental events. This is so because themeaning 'demand for information' of the squeak is not locked into a particularhere-now context, but derives its symbolic value from its position in a developingsystem of other contrasting values and their modes of realization.

The child's high-pitched squeak, which Halliday glosses as a 'demand forexplanation', is a situation-specific instance of adaptive motor behaviour. It is ameans whereby the child explores his immediate environment. It is anelementary form of hypothesis, which elicits its further modification on the partof the other (the mother) in the exchange. This modification, of course, comesin the form of the mother's far more highly specified response. In this way, the


child learns about the world in these jointly constructed acts of interpersonalengagement and experiential construal. Moreover, the mother's response islikely to have affective consequences as well. The assimilation of the unknownphenomenon which had perturbed the child's perceptual purview to knownexperiential categories is potentially laden with affective valence and motivational significance. Thus, the satisfaction of the child's 'demand for explanation'and the assimilation of the given phenomenon to the jointly shared andsustained experiential categories of the dyad boost positive affect and, hence,motivate the desire for further exploration.

The loud noise caused by the scattering of the pigeons is an environmentalevent which is picked up by the child's perceptual systems. The fact that the childlifts his head and turns around to attend to this event indicates this quite clearly.The head movement constitutes a response to the event, which is seen as beingin some way interesting, surprising, and so on, and is, for this reason, tracked bythe child's head movement. In this way, the event - the scattering of the pigeons- gets constituted as something which is worth orienting to and responding to.The child's high-pitched squeak, therefore, becomes the child's comment onwhat he has selected as the focus of his attention. The interesting and revealingthing about this exchange is the way in which the child's high-pitched squeak isrecognized and construed by the mother as being a comment about the pigeons.That is, the mother is attuned to the interests and experiences of the child inthree specific ways. First, she recognizes that the child's squeak is a particularorientational stance. Secondly, she recognizes that this stance is aboutsomething. Thirdly, she provides, in making her response, a link between thesymbolic possibilities of the higher-scalar meaning systems of the culture and themother-child dyad. All three aspects are fully revealed by her response, whichshows both that she is tracking and is tuned into the child's stances towardsparticular phenomena and that she can co-ordinate the child's stance with herown (linguistic) construal of the event. The overall structure of this dyadicexchange may be glossed as follows:

PERCEPTUAL PICK-UP OF STIMULUS INFORMATION ABOUTENVIRONMENTAL EVENT

CHILD'S ATTENDING TO EVENT AS A FOCUS OF INTEREST

CHILD'S RESPONDING TO EVENT WITH PROTOLINGUISTICSIGN/ENGAGING INTEREST AND ATTENTION OF MOTHER

MOTHER'S RESPONSE INTEGRATES EVENT AND CHILD'S RESPONSEINTO HIGHER-SCALAR ECOSOCIAL CATEGORIES


The child's response to the event in the form of his looking around followed bythe high-pitched squeak is a meaningful act which is directly dependent uponthe experienced phenomenon. It lacks the symbolic organization of a lexicogrammar, which enables acts of meaning to be freed from direct dependenceon the phenomenal (Halliday I992a: 23, 25). Nevertheless, his compositemeaning-making act may be seen both as construing experience and, at the sametime, his dialogically engaging the mother in his experience, in the processenacting an intersubjective framework of interpretation, which the motherfurther elaborates. The two-part nature of the child's response - viz. lookingaround followed by squeak - can be interpreted as a protolinguistic analogue ofthe interpersonal function of mood, which specifies the speaker's interactivestance on some proposition or proposal in the adult language. Thus, thescattering of the pigeons is constituted as a kind of proto-Subject about whichsome comment is made in the form of the squeak. The squeak may therefore beseen as a proto-proposition which is made about the designated Subject. BothHalliday's analysis and the mother's response draw attention to the way in whichthe latter is attuned to both the interactional status of the child's sign as well asto its experiential significance.

11. Bogdan's Theory of Mental Sharing and Topical Predication

Bogdan (2000: 78) argues that the infant's selection of a target of attention is atopicand that the representation of the target is a comment. This results in whatBogdan calls 'topic-comment parsing or segmentation and the mental routinethat effects the parsing is topical predication, ... ' (2000: 78). Bogdan considerstopical predication to be pre-linguistic and based on mental sharing:

Work in linguistics, psycholinguistics, and philosophy has shown that topicalpredication cannot be reduced to logical or grammatical predication (seeClark and Clark 1977 and Ninio and Snow 1996 for psycholinguistic surveys;see also Bogdan 1987 and Dretske 1972 about implications for propositionalattitudes). This irreducibility can be grounded in the hypothesis that topicalpredication is prelinguistic and rests on mental sharing by sentimentalbonding and physiological regulation. Also supportive is the fact that autisticpeople, whose minds are not socialized intersubjectively, can do logical andgrammatical but not topical predication.

To account for these developmental truths and the reasons for them, it helpsto start with the liberal notion that topical predication indicates a mental take(comment) on something of interest (topic). The topic can be whatever is atarget of attention or interest - a thing, event, property, relation, scene,sentence, thought, image, picture, sign, or complex structures of such. Thetopic can be out there in the world or up here in the mind, whether one's ownor somebody else's. The comment is some mental take on or mental relationto a topic, expressing an experience, emotion, stance or attitude. Dependingon whether the topic expressing an experience is a worldly or mental item,the comment can be a representation of something out there in the world or


a representation of a representation in the mind, respectively. There are manyways in which topics and comments can be represented - in perception,memory, thought, mental imagery, bodily posture, action, with or withoutlanguage. (Bogdan 2000: 78)

The idea of a pre-linguistic topic-comment structure, with its correlative of thesubject's mental takes on the world, suggests that there already exists an at leastweak classification of self as distinct from the world, including others. Furthermore, it suggests that there exists a prior, pre-linguistic structure on whichlinguistic forms of predication are subsequently built. In the dynamics systemsview, on the other hand, the differentiation of self and world depends on morethan the prior existence of a topic-comment structure. Halliday's notion of theinterpersonal gateway to meaning draws attention to the initial primacy of prelinguistic proto-interpersonal or dyadic engagement. The infant's earliest experiences of the world are topological and require a gradual and progressivedifferentiation from a less differentiated whole. The self is experienced as not yetsharply differentiated from the world, though in some kind ofvague relationshipwith it. That is why I prefer the proto-interpersonal interpretation given above.If, on the other hand, there exist 'prior topic-predication formats' (Bogdan2000: 78), then this would amount to saying that the child comes alreadyequipped with a primitive means for segmenting the self/nonselfrelationship interms of some target of selective focus in the world (internal or external) andone's propositional take or attitude on this in the form of a comment. Self andworld would already be construed as different in this view.

In my view, the construal of self and world (nonself) as different is not alreadygiven; instead, it is a form of developmental emergence (Salthe 1993: 244).Thus, the shift from the child's primordial experience of self-in-relation-tothough-not-yet-differentiated-from-the-world to the experience of se1f-asdifferent-from-world involves the forming or coalescing of a higher-scalar entity- e.g. a dialogically organized proto-proposition - around a number of variablesthat now appear as the lower-order components - neural values, environmentalstimulus information, visual tracking by head-turning, protolinguistic gesturesand vocalizations, the presence of others, meanings from other conspecifics out of which the new property has emerged. By the same token, there is also acorresponding shift in the observer/agent's attention to the typologicalcategorial distinctions belonging to the higher integrative leveL This means thatthe newly emergent proto-propositional resources and attitudes mentionedabove, with their typologizing of self/nonself relations, are now seen as closer tothe observer's scale which is shared by other members of the culture, rather thanto the vaguer state of initial topological non-differentiation. The latter, however,remains implicit at the lower scalar levels, which are now constrained by thehigher ones. In any case, it is the dialogically organized and therefore sociallyshared act of proto-meaning which exists and which is most visible in the supersystem transactions between self and other. Nevertheless, the mental andindividual bias illustrated in Bogdan's description of topic predication indicatesthe hankering after inferred objects - mental entities - once the observer'ssupersystem categories have been described with reference to their leveL


The child's protolinguistic engagements with the world make possible furtherdifferentiations of the reentrant loop that maps these engagements onto thechild's inner neural activity. The child thus moves from an initial, primordialtopological state of non-differentiation (vagueness) between experiencing andmeaning, to an experiencing of experiencing, to the progressively more differentiated construal of meanings that are attributed to experience (Lemke p.c.).Bogdan's focus on pre-linguistic topic predication could suggest a built-in teloswhere none may, in fact, be necessary in order to explain what is going on. Hisfocus on mental sharing, physiological regulation, and sentimental bonding saysrather less about the physical-material processes - e.g. stimulus informationabout environmental events - which inform and constrain the self-organizingproperties of the child's neuro-motor system. Such environmental events and thestimulus information they afford the organism are contingencies with which theself-organizing properties of the child interact in ways the probabilities of whichare pre-set by biologically in-built criteria of value (section 12 below).

In my view, the prior existence of a topic predication structure would be toorigid an imposition and would not allow the child to respond to constantlychanging environmental contingencies. As the above discussion of the examplefrom Halliday begins to show, the key to understanding what is going ondepends on understanding the ways in which individuals adapt to openness andnovelty in complex dynamic systems comprising very many fluctuating variables.In our example, these include the interactions between an environmental event,the child's perception of this, his attending to it in looking around, his squeak,and the mother's response. It is the mother's more specified linguistic responsethat provides the link to higher-scalar ecosocial semiotic categories and relations.In so doing, she provides the child with access to resources - symbolic possibilities - and entrains him into their contextually appropriate use. The furtherexploration and entraining of these possibilities brings about the increasingdifferentiation of topological vagueness and wholeness and the development ofan expanded repertoire of relations between self and other.

The above analysis serves to show that it is not a question of prior, prelinguistic topic-comment structures which permit 'mental sharing'. The analysisshows that the ability to focus on and form a proto-proposition about some eventis an emergent property of the overall activity structure that takes place. As theanalysis shows, this may be broken down into at least the four componentsspecified above. The individual is thus regulated by his or her exchange transactions with the ecosocial environment, including other conspecifics. Both thephysical environment and the social environment function as structured sourcesof potentially meaningful information. In the example, the physical environment provides stimulus information about an event - the scattering of thepigeons - to a potential observer; the child picks up and attends to this information, as demonstrated by his tracking the event through head-turning. Hissqueak dialogically engages the mother, who responds with a linguisticutterance. In my view, the mother's contribution to the exchange provides astructure which functions to complete or further elaborate the child'sdeveloping structures. (See also Cole 1996: 187-90 on the notion of routineactivities as proximal environments for development.) Thus, the much more


implicit structure and meaning of the child's contribution to the dialogue elicitsthe more complete response of the mother. The latter has the function ofextending and modifying the implicit structures of the child by elaborating andfurther specifying them.

In this particular example, the mother's utterance, Yes, those are birds. Pigeons,is a declarative clause whose demonstrative pronoun Subject those indexes thesame phenomenon which the child had initially responded to. That is, theperceived phenomenon is construed as a linguistic entity about which a proposition is made in the rest of the clause. In this case, the attributive clausefunctions to specify the type-category to which the pigeons seen by the childbelong and, therefore, to classify or name them. Thus, the relatively low orderstructure created by the child - viz. 'looking around followed by high-pitchedsqueak' - is itself a kind of proto-proposition in relation to which the perceivedevent (the pigeons) is the proto-Subject about which some proto-proposition ismade. My point is that this proto-proposition is not an internal mental structurealready possessed by the child. Rather, its emergent properties are distributedacross various components of a muItimodaI social activity. The mother's furtherelaboration of this structure in the form of an explicit linguistic propositiontherefore entails its further elaboration and completion. Table 4.5 provides amultimodal analysis of the transaction considered here.

Table 4.5: Multimodal eo-deployment ofperceptual-semiotic resources in the exchangebetween the child and his mother in Halliday's example

Perceptual-semiotic Movement Visual Auditorymodality

The dialogically eo-ordinatedcomponents of theunfolding event

The scattering of Spontaneous Optical informa- Auditory informa-the pigeons as envi- movement of the tion about an envi- tion about an envi-ronment event pigeons as they ronmental event ronmental event

scatter available to a point available to a pointof observation of observation(e.g. the child and (e.g. the child andhis mother) his mother)

The child's dialogi- Turns head-body to Directs gaze vector Utters high-cally co-ordinated attend to and to event pitched proto-response visually track the linguistic squeak

event

The mother's Co-ordinates head- Co-ordinates gaze Utters linguisticresponse body movement with child's to response

with the child's to jointly attend tojointly attend to eventevent


There is no need to insert mental functions between biology and culture in theway that Bogdan does (2000: 31). Here is how Bogdan distinguishes mentalfunctions from neural functions:

The contrast need not be between biology and culture, with nothing inbetween. Culture cannot be grafted directly on unsensitized neural functions.Left out are mental functions, however simple. The problem is not that suchsimple functions are neural; of course they are, in an implementational sense.The problem is that mental functions cannot be reduced to biological orneural functions, insofar as the former (but not the latter) have psychologicalor mind-world value. That psychological value concerns the relation betweenorganism and world, both physical and social, and the impact of that relationon behavior. This is how psychology (but not biology or neuroscience) looksat organisms. (Bogdan 2000: 31)

The problem in Bogdan's account lies in his assuming that neural functions areneutral or value-free. They are not. Moreover, neural functions are concernedwith the relationship between organism and world. There is no need to postulatemental functions, however primitive, which mediate between biology andculture from the outset. Biology is neither value-free nor non-social right fromthe word go. Value-laden neural functions predispose the child to participate inself-regulating exchange transactions with their ecosocial environment from theoutset. The child's high-pitched squeak, in our example, is a dialogic act whichis oriented both to an environmental event and to the mother. It shows how thechild's active and dialogic orientation to his environment works to seek information and meaning which will affect his further development. The child'sproto-linguistic squeak constitutes an active intervention in, rather than a passiveresponse to, his environment. The squeak is fully dialogic, if not linguistic,because it is intended to obtain an appropriate response from the mother orother primary caregiver. In this way, the child's own agency constitutes anattractor space in which the context-bound responses of caregivers are organizedand channelled. The attractor space therefore constitutes a domain in which thesocial practices of attending to and interacting with infants interface withbiological values and functions. It is the interfacing of the two which enables thefurther development of the child's own self-regulating structures. In this way,mental functions such as topic predication are built up on the basis of thereentrant mapping of the child's exchange transactions with its environmentonto internal neural activity. They are not part of the mechanism whereby suchstructures are acquired. The means of acquisition of such structures is betterhandled by Edelman's notion of value, as we shall now see.

12. Interpersonal Meaning, Value, and Action

What is it that impels the infant to ever more differentiated symbolic possibilities? Is it necessary to postulate pre-linguistic topical predication as the key to'mental sharing'? Pre-linguistic topic predication would point to the existence of


an already existing in-built initial differentiation between self and other, howeverprimitive. Recent neurobiological evidence does not favour this view. Forexample, Damasio advances the notion of the 'proto-self, which he defines as a'preconscious biological precedent' (1999: 153) of the self. I shall return to thisnotion in chapter 5, section 5. Damasio's neurobiological arguments do notsupport the notion that there exists an already prior, initial polarization of selfand world as being categorially distinguished. This seems to suggest some kindof categorization of this fundamental distinction prior to experience. Edelmanhas proposed the notion of some very general biological values which areintrinsic to the organism from the start. Edelman's values are not based on suchcategorial distinctions. Values simply function to provide the initial impetus tothe organism's ongoing, time-bound engagements with the world. They do notin themselves provide the organism with ready-made categories for distinguishing self from other through experience or for experiencing one's own experiencing, let alone for giving meaning to experience. The progressivedifferentiation of self and other and the elaboration of the self's experience ofthe world is based on epigenetic principles. On the basis of experience, the differentiation of self and other occurs through the reentrant mapping and gradualstrengthening of neuronal groups whereby the category of self-other relationsand transactions is elaborated. Values function to nudge the developingorganism along certain developmental pathways. For this reason, they may beseen as kick-starting the interaction between the organism's self-organizingdynamics and the environmental contingencies it encounters. For example,Thelen and Smith (1994: 315) cite research undertaken byJohnson and Morton(1991) which proposed that 'infants' strong attraction to faces begins with aninitial orienting bias in newborns' (1994: 315). Thelen and Smith continue asfollows:

This bias, the behavioral evidence suggests, favors the arrangement of severalhigh-contrast 'blobs' corresponding to the relative location of eyes andmouth, likely mediated through subcortical pathways. Given that infants seesuch configurations in real faces close up during nursing, changing, andsocializing, even a small initial valence toward that configuration would leadto rapidly strengthened neuronal groups in the presence of faces. As higherbrain centers become more involved in the second and third months, thesystem will already be attuned toward facelike stimuli, insuring that infantswill be positioned to acquire more information about faces. Indeed,preference for particular faces emerges by 4 months or so.

What is minimally needed to start such a process moving is a sensory systemwith some relatively unspecific tuned parameters and a value wash that keepsthe organism preferring certain stimulus configurations over other values.Such values are not specific tutors. Having a bias in the system that says lightis better than no light or human voices are better than auto horns does notendow the system with knowledge modules or conceptual primitives or theunderstanding of objective properties. Rather, these are valences or trophismssimilar to those exhibited by simple organisms and even by animals withoutnervous systems. That such valences might be contained within the architec-


Perceptual-motorgrounding of body-brain in Umwelt

Physical-material

Symbol

Entextualizationand contextual-ization of deicticfield

Vague, unboundedtopological-continuous

Specified, boundedtypological-categorial

ture of the system does not put a ghost in the machine, or if it does, it is theghost of life itself, for similar biases are exhibited even in single cells. (Thelenand Smith 1994: 315-16)

The initial orienting bias to the 'blobs' mentioned in the above citation indicatesthe purely iconic and topological-continuous nature of the child's initialorientation to the other. Thus, the values which favour the active attending toenvironmental events through the perceptual systems, the participation indyadic exchanges with mothers, and so on, bias the self-organizing dynamics ofthe system. The infant participates in activities which pull the infant away fromthe initial attractor space of relative topological (iconic) non-differentiationbetween self and other towards the increasingly differentiated indexical andsymbolic differentiations afforded by the mother's access to the higher-scalarcategorial (indexical and symbolic) differentiations of the ecosocial system. Indoing this, the child's initial, less differentiated relations between self and otherbecome progressively more differentiated. The child progresses from proto-imperative followed by proto-indicative modes which are tied to the phenomenaof experience to the full range of interactional possibilities afforded by theinterpersonal systems of the adult lexicogrammar. Figure 4.1 illustrates thisdevelopmental emergence as a presupposition-cum-implication hierarchy oficon, index, and symbol with reference to the developmental emergence ofinterpersonal semiosis.

This suggests that the interpersonal metafunction is concerned, in the firstinstance, with the vague, the unspecified. It deploys prosodic and scopal modesof realization so as to engage with and resolve, through dialogic negotiation (cf.Bakhtin's concept of finalization), that which is not yet assimilated to somedeterminate classification. In the first instance, the interpersonal metafunction

Figure 4.1: Integration-cum-presupposition hierarchy of iconic, indexical, and symbolicmodes of grounding

Contiguity with selec-tively attended tocontextual values inhere-now spatio-temporal purview ofaddresser and addressee

IndexIcon

Semiotic-discursive


is concerned with the assigning of affective valence and motivational relevanceto the not-yet-defined phenomenon. It is concerned, above all, with questionssuch as 'what value do I assign to this thing?', 'what should I do with it?', 'howshould I orient to it?', and so on. In other words, interpersonal meaning orientsinteractants in terms of the given phenomenon's value-laden salience for action.It is only through acting in and on the world - i.e. the nonself - that we are alsoable to construe it experientially in ways which make it understandable and,therefore, able to be assimilated to the stock of human social meaning-making.It is the ability, initially, to formulate hypotheses about the unknown and thevague or indeterminate which constitutes the first stage in our adaptation to it.In this way, our modalized orientations to phenomena constitute a structuredresponse which functions to inhibit and channel our responses to the indeterminate at the same time as we elaborate our experiential understanding of it. Itis the experiential metafunction which completes and complements this processby abstracting from our subjective investments in particular courses of action soas to create a more explicit experiential framework of knowledge.

13. Procedural Knowledge, Declarative Knowledge and the Semiotic Spiraltowards Symbolic Consciousness

The distinction that psychologists make between procedural and declarativeknowledge is relevant here. Procedural knowledge is concerned with actionoriented to intervening in the world and bringing about desired change. It isbased on the active sensori-motor exploration of the world in order to developappropriate and lasting patterns of action for dealing with the world and its contingencies. It is knowledge which is implicit because acted on only in direct interactions with other objects, other people, and so on. The right hemisphere isresponsive to and productive of the global or holistic imagery which provide theinitial topological-continuous variation - the proto-meanings - on which a text'smeaning is built. The left hemisphere re-interprets these values as explicitstructure and increasingly abstract taxonomic hierarchies of experientialcategories. It is significant, then, that the procedural basis of interpersonalengagement and action emerges before declarative knowledge, both phylagenetically and ontogenetically. Declarative knowledge, on the other hand, isconcerned with what is explicit and known; it includes knowledge which iscreated in and through the ideational resources of language. As Halliday,Trevarthen and others have shown, the child first learns, though protointerpersonal interaction, the procedural knowledge associated with thenegotiation of the T I'you' relation in, say, the mother-child dyads characteristic of primary subjectivity. Only later is the infant's attention turned towards theexperiential construal of the world which lies outside 'me' and 'you' in the realmof the third person. Thus, early interpersonal interaction and the dyads whichgive shape to this entail, in the first instance, the entraining and adaptation ofsensori-motor patterns of activity prior to the more abstracted forms of symbolically mediated activity that are characteristic of later stages of development.Table 4.6 schematizes some of the salient stages in this process.

THE SEMIOTIC BASIS OF CONSCIOUSNESS

Table 4.6: Types ofknowledge and stages of semiotic development in early infantsemiosis

213

Time in Months Type of Knowledge Semiotic Mode

0.0-0.8; 0.8-0.9 Sensori-motor knowledge; Iconicpractical knowledge;adaptation of reachingresponse to object;implicit recognition ofself-object distinction

0.9-18.0 Representational Iconic-indexicalknowledge; means-endand cause-effect relations;inten tional activity;memory

18.0-24.0 Emergence of abstract Iconic-indexical-symbolicmeta-representations;construing remembered,unshared information;emergence of I-me dis-tinction, i.e. child knowshimself not only as apoint of action (I) actingon others, but as otherssee him (me)

Procedural knowledge, which is based on sensori-motor activity and the protointerpersonal dimension of meaning, is not essentially representational incharacter. Rather, it is concerned with the enacting - initially in interpersonalcontexts - of social activity-structures (cf. genres). However, it is, in its own way,a form of knowledge which is transferred between individuals in and throughthe processes of exchange. Mimicking and imitation, for example, (Piaget 1946;Battacchi 1999; Perinat and Sadurnf 1999) can occur with reference to otherpersons as well as in relation to inanimate objects. In play, children may interactwith the nonself, which may be another person or some object. In the process ofdialogically engaging with the given object, both structure and meaning areexchanged. Initially, the child projects some qualities onto the object of interestin his or her purview. The child then imitates his or her own interactions withthese objects; in the process the child ritualizes his or her own activity as he orshe adaptively modifies his or her own actions in and through their exchangeswith the object. In this way, ritualistic activities emerge as the precursor ofsymbolic activities (Thibault 1998b; Lorenz 1969 [1963]). That is, genre anddiscourse structures, perhaps specific to the precursor world created andinhabited by children (Harre 1983: 251-2), emerge and function as extensionsof the self.


Interpersonal moral orders are grounded in procedural activity and only latermade more abstract through experiential reflection on action and interaction.Thus, declarative memory produces experiential construals of the implicit logicof interpersonal exchange in the procedural system. The elaboration andstorage of such meaning is essential to the formation of interpersonal moralorders in which one's own and others' actions can be explained and accountedfor within a network of responsibilities and agentive positions and theirassignment to individuals. In this way, interpersonal-procedural knowledge isabstracted and codified as forms of moral reasoning (see Hasan 1992a; Thibault1992b). The declarative-ideational system has thus elaborated a meta-discourseconcerning procedural-interpersonal activity through a process of abstractionfrom the here-now of interpersonal engagement and exploration. As the work ofHalliday, Trevarthen, and other researchers shows, the earliest dialogic acts ofinfants in the form of sensori-motor acts (gestures, vocalizations) and imitationoccur prior to the emergence of any explicit meta-semiotic awareness of theprinciples which regulate interpersonal exchange. The declarative-ideationalsystem allows for reflection on and further elaboration of the meaning-makingprocesses that occur in early sensori-motor exploration, imitation, and play.

The procedural-interpersonal system constitutes a form of memory for actionand interaction. This memory is built up and elaborated on the basis of verymany individual exchanges which have been integrated into the patterns ofinterpersonal engagement which are appropriate or otherwise acceptable in agiven interpersonal moral order. The integrative nature of interpersonalexchange means that the often competing motivational demands of agents areresolved with reference to the interpersonal moral orders in which agentsinteract with each other. It also means that individual needs and desires areregulated by social norms and values which the individual internalizes. Thismeans that the system of sometimes conflicting, sometimes aligned values,judgements, viewpoints, and so on, in the system of social heteroglossia, areorganized and internalized in terms of hierarchies of values such that themorality of specific courses of action can be evaluated and re-evaluated. As Ipointed out above, the enactment of such value hierarchies, which are usuallyimplicit in specific acts of exchange, is developmentally prior to the experientialconstrual of the actions in which these values are implicit. In this way, explicitexperiential reflection on action and value becomes possible, and in ways whichcan act back on and alter action itself. Explicit experiential reflection meansincreased symbolic abstraction. With increased powers of symbolic abstractiongo the capacities of self-monitoring, self-control, and self-awareness that aredefining characteristics of higher-order consciousness.

This shows very clearly how both the interpersonal and experientialdimensions of meaning-making are implicated in and are indispensable to theworkings of higher-order consciousness. Thus, the emergence of symbolic orhigher-order consciousness (Edelman 1989, 1992) comprises a spiral that beginswith sensori-motor exploration and procedural-interpersonal engagement withthe other before symbolic consciousness emerges. These earlier forms are thenexperientially construed as symbolic acts which can be shared with others. Theseobservations accord with Halliday's (1993: 103) notion of interpersonal meaning


as the 'gateway' to later forms of meaning-making: new meanings are firstenacted interpersonally before being construed ideationally. Symbolic sharing,in turn, constitutes the basis of the further abstraction of interpersonal interaction from the here-now of sensuous and embodied activity. This means thatprocedure can be dialogically reflected from the social viewpoint of others - realand virtual- then enacted in subsequent interpersonal exchanges. The result isthat interpersonal interaction is itself modified, thereby leading to furtherchanges in its experiential construal. This process can take place in the courseof interpersonal interaction itself, or internally by virtue of its abstraction fromsensori-motor activity and its consequent specialization to the 'inner' realm ofsemiotically mediated thought, which is itself a specific form of discursive activity(section 4, pp. 184-9).

14. Interpersonal Meaning, Goal-seeking Activity, and the Goal Hierarchy

The interpersonal grammar of mood shows that the goal and the means toattaining this are not constitutively separable elements. Each of the basic moodchoices, along with the speech functions that the various mood choices enact,entails a present state, a desired future state, and the resources - selective crosscouplings of the physical-material and the semiotic-discursive - that could beused to attain the goal-state, viz. to transform the former into the latter. Thegrammar and semantics of the interpersonal metafunction shows that what areoften called 'desires', 'wishes', 'needs,' 'wants', and so on, and seen as mentalpredicates in the theatre of the individual mind, can be seen as dialogic interacts(Martin 1992a: chapter 2, 1992b; Bakhtin 1981, 1986; Thibault and VanLeeuwen 1996). The interpersonal grammatical resources discussed above arealso deployed in the service of larger-scale goals on different space-time scales.Thus, specific semantic interacts and the clauses and larger units which realizethese interacts or moves are seen as nested in this larger-scale structure in termsof relationships of the kind Silverstein designates as 'interactional coherence'(1997). At any given moment in the unfolding interaction, the number ofpossible variables that can function as means and end are not totally openended, but are limited in number. A particular choice of speech-function movein the interaction entails criteria of inclusion and exclusion by virtue of thedialectical duality of paradigmatic and syntagmatic principles of organizationwhich operate in discourse.

However, the open and dynamic character of discursive negotiation (Thibault1999a) means that the range of possible relevant variables is not fixed, butflexible. The principle of goal-seeking means that we have, to varying degrees, arelative semiotic freedom (Wilden 1981: 25) to make decisions and to choosebetween possible alternatives. The human pre-frontal cortex, which is fundamental to goal-seeking activity, makes this relative freedom possible by virtue ofthe way it organizes the temporal arrangement and sequencing of our actions asmotor plans. The cortical appendage, the basal ganglia, plays an important rolein this sense (Edelman 1992: 105-6). It does so by taking account of relevantcontextual variables and matching that information with sensori-motor activity.


Thus, selections in the interpersonal grammar are, in the ongoing temporal flowof discourse, cross-coupled to the physical-material processes whose flows theyentrain and co-ordinate. The basic principle underlying this operation is the wayin which the sought-after goal - the desired future - is modally evaluated inrelation to the undesirable and negatively evaluated present.

The linguistic resources of modalization and modulation play a key role here.In discursive activity, dialogic interacts are organized in relations of interdependency (one thing follows after another, e.g. Question-Answer, and at thesame time there exists a relation of semantic interdependency between the twoparts of the dialogic interact), as well as hierarchically (some goal-seekingpathways dominate others). Within this complex network of ecosocial relations(Lemke 1995b), consciousness may be seen as an integrity (Salthe 1993: 51)which interrupts the hierarchy of relations in the way I proposed in chapter 2,section 1, pp. 59-67. The logic of scalar hierarchies tells us that one level- theecosocial one - is the one we engage with most naturally. It is in relation to thislevel in the overall hierarchy that actions, objects, and so on, are interconnectedwith other events, objects, and so on. Further, these interconnections constitutean entire system of relations potentially extending over many different spacetime scales. They can function both independently of our body-brains as well asdirectly in relation to them. Higher-order thinking requires the distinctionbetween self and nonself.

Lukacs (1980 [1978]: 26) has described how the separation of subject andobject, which emerges in the labour process, entails the 'channelling' of theresulting reflection of reality into consciousness. Thus, 'being' and its 'reflectionin consciousness' constitute a break with the prior, unitary form of being. Theresulting antithesis leads to the emergence of a qualitatively new level of being,viz. social being. Similarly, Cole (1996: 118-21) discusses the notion of artefactmediated action as first developed by Russian psychologists such as Luria (1976)and Vygotsky (1986 [1934], 1987 [1934]) in the cultural-historical tradition.Cole draws a distinction between 'natural' or unmediated functions and'cultural' or mediated ones as two ways in which the subject and object (environment) are related to each other. In my view, there is no such thing as anunmediated relationship between subject and object. Rather, the question isconcerned, above all, with the different kinds of mediation afforded by theintegration hierarchy of iconic, indexical, and symbolic modes of semiosis. Thisshows us that the purely iconic mapping of topological-continuous variationonto topological-continuous variation of the kind found in perceptual-motorcategory formation prior to the emergence of indexicality remains linked to theimmediate here-now scale. On the other hand, indexical and symbolic forms ofmediation progressively free the agent's consciousness from the here-now scale.The so-called 'direct', 'natural' relation is best seen as the most primordial formof iconic relationship between self and nonself in which no meta-level experiencing of experience is apparent.

The linguistic resources of modalization and modulation function to grounddialogic interacts in relation to the here-now of the speech event (Finite) and,therefore, in relation to the ecosocial space-time scale of human social meaningmaking. Modal systems such as 'probability' and 'usuality' in English ground the


number of variables that can be contextually managed in order that the numberof variables does not run away from us. The point is that modality deals with thecontingent and the unknown by allowing participants in discourse to orient toand evaluate it as if it were predictable, known, or possible, and so on. Theunknown, the uncertain, the doubtful, and so on, constitute sources of friction.Interpersonal exploration and negotiation is brought into play so as todetermine the most appropriate level of analysis for the evaluation andresolution of both semiotic and material friction (Thibault 1995). Moreover, thenesting of discourse moves within dialogic interacts, and the nesting of these inturn with still-larger-scale interactional episodes, means that specific contingencies are not simultaneously evaluated at all levels of the whole hierarchy ofrelations. The interpersonal grammar oflanguage is organized in such a way thatinteractants can orient to a goal on a specific level and act as if only that levelwere contingently relevant. In this way, the values of things in the world can beselectively attended and discursively fixed so that action is possible.

15. Consciousness and Semiotic Stratification

I would now like to consider consciousness in relation to the stratified nature ofsemiotic systems (chapter 1, section 9). Stratification is the means whereby theagent is cross-coupled with and integrated with the ecosocial system (chapter 1,section 9, pp. 39-46; chapter 2, section 1, pp. 59-67). This is so because theexpression and content strata interface with and integrate the agent's Innenweltand Umwelt, respectively. This means that the agent has internal, semioticallymediated, and progressively elaborated and individuated models in his or hercentral nervous system for acting in the world. Moreover, individuals canintervene in and expand their Umwelten through the externalization of thesemodels in the form of social meaning-making activity. In this way, the socialsemiotic system is integrated with the small-scale dynamical processes of theagent's body at the interface of the body with the expression stratum. The agentis also integrated with the dynamics of spatial and temporal scales that aregreater than those of the agent's own bodily processes. The integration withthese much larger scales is possible because of the interface between the agentand the content stratum. It is this second interface, as we have seen, thatconnects the agent to the phenomena of experience in his or her Umwelt.

Higher-order consciousness is semiotic through and through. The twointerfaces which are constituted by the expression and content strata are themeans whereby the individual organism and its dynamics is cross-coupled withthe very different dynamics of space-time scales which can extend well beyondthe organism along a semiotic-material trajectory into its ecosocial environment.In this sense, language and other symbolic modes of meaning-making constitutea resource for the integration of the here-now activities of the organism withlarger space-time scales. Peng accordingly distinguishes between motorfunctions and motor activities. The latter may appear, superficially, to 'make useof the cortico-bulbar pathways for efferent neuronal connections to innervatethe musculature of the vocal apparatus (including the facial musculature), if


movements of the muscles are the main objective' (1994: 114-15). Thedifference between 'motor functions' and 'motor activities' lies in the fact thatthe former are part of language in the individual, whereas the latter are notrelated to language activity in the individual organism. Both motor functionsand motor activities, as defined here, are performatory activities of theorganism's executive systems (Gibson 1983 [1966]: 46). In the case of speechproduction, the motor functions which are involved in the articulation of speechsounds during phonation are semantically integrated with other facets oflanguage in the brain of the speaker in order to produce a linguistic event.Motor activities are not semantically integrated in this way. Thus, the motorfunctions of phonation are integrated with informational invariants concerning(1) the sound shape of the word to be uttered, (2) their correct sequencing onthe expression stratum; and (3) the lexicogrammatical form and semantics ofthe word on the content stratum (chapter 3, section 11). Peng points out thatthe motor functions consist of the mapping of these various components ontoeach other. To quote Peng:

... when the speaker wants to produce the sequence of sounds, Ip/, lce/, andI d/, the speaker must have the meaning of the word padin the cerebral cortexfirst, for which the sequence of sounds is intended, before it is produced; then,the sequence of sounds must have the brainstem function, that is, goingthrough the cortico-bulbar pathways, coupled with the extrapyramidalfunction and the cerebellar function, when it is being produced.

In contrast, however, when the lips are to be closed for the purpose of mastication, in order to prevent food from coming out of the mouth, the closingof the lips is simply a part of the motor activities of the nervous system, distinctin nature from what I have called motor functions which are a part of thespecific brain functions pertaining to language. (Peng 1994: 114)

The use of the cortico-bulbar pathways in the production of human speechsounds entails a functional loop which connects the cerebrum or neo-cortex tothe vocal apparatus through the cortico-bulbar pathways (Peng 1994: 117). It isthis loop of functional connections leading from the neo-cortex through thecortico-bulbar pathways to the vocal apparatus which enables the speaker tomake the very fine segmentations and distinctions necessary for the articulationof speech sounds. In this way, the speaker produces the invariants of stimulusinformation which listeners can discriminate from the topological-continuousvariety of the acoustic signal. Therefore, the listener is able to tune into thesesame invariants even when the speaker, the time and place, and so on, ofspeaking and listening, are different from one occasion to the next (see Gibson1983 [1966]: 278; this volume, chapter 3, section 2, pp. 148-53).

The motor functions, as theorized by Peng, thus integrate the expression andcontent strata of language within the central and peripheral nervous systems ofthe individual. By the same token, the individual organism is semioticallyintegrated with space-time scales which extend beyond the material body.Stratified semiotic systems such as language and depiction face two ways, i.e.both 'inwards' to the body of the articulator of the sign and 'outwards' to the


phenomena of experience in the ecosocial environment. It is from the latter thatthe articulator picks up stimulus information by means of his or her perceptualsystems (chapter 1, section 3, pp. 11-18). In fact, both the body and the worldare environments with which the expression and content strata, respectively,interface in the processes of transducing input and output between the body andthe ecosocial environment with which the body-brain interacts. These inputs andoutputs can be classified according to the schema presented in Figure 4.2.

Figure 4.2: Classification of input and output impulses involved in the symbolic trans-duction of stimulus information and sensori-body activity in semiosis (adapted andmodified from Gibson 1983 [1966]: 46)

Inputs: content(afferent impulses)

Outputs: expression(efferent impulses)

Exteroceptive (stimulus information activelyobtained from environment by perceptualsystems as proto-meaning)

Proprioceptive (stimulus informationobtained from own body, e.g. inner sensations,feelings, inner visual and other imagery)

Attending to (listening, looking, etc.)meaning-making acts of others by means ofactive perceptual exploration

Performatory activity of articulators (vocal-tract, manual-brachial gestures, arm-hand-joint-eye-tool kinaesthesis in tracing, graphicact) in the externalized execution of signs

The inputs and outputs which cross the boundary between organism andexternal environment are the means whereby matter, energy, and informationare exchanged between the two. These inputs and outputs contain differentorders of information. These orders can be classified according to my re-interpretation of Deacon's hierarchy of reference in terms of the presupposi-tion/implication hierarchy discussed in chapter 3, section 7, pp. 126-34. This isso in relation to both the expression and content strata. Moreover, the informa-tion-meaning which is afforded by these two strata is both exteroceptive andegoreceptive. In providing information to the other (exteroception), thespeaker, say, simultaneously provides information to him- or herself about his orher own meaning-making activity (egoreception). This means, for example, thatspeakers hear their own voices while speaking and at the same time, feel themuscular movements of their face and vocal tract. Gibson (1986 [1979]: 115)points out that all perceptual systems are both exteroceptive and egoreceptive:


the body of the perceiver is also a source of information about the perceiver'sbody during the perception of some environmental event. Both environmentand self are simultaneously specified by information from the two sources in theact of perception (Gibson 1986 [1979]: 116). The complementarity of these twoperspectives is based on the fact that the information for specifying the two is thesame. Similarly, the body of the speaker in the process of articulation is both asource of information about the speaker to observers (e.g. the addressee) in theexternal environment and simultaneously also a source of information to thespeaker about his or her own body.

As we shall see in chapter 5, section 4, pp. 246--8, the reentrant mapping ofthe information from these two sources - self and nonself - in the act ofperception creates higher-order maps in the brain of these lower-orderself/nonself mappings. The resulting higher-order or global mappingsconstitute a basis for the internalization and consequent reorganization of thisperceptual information in ways which are relevant to the structure of (self)consciousness. The local outputs of multiple maps that are connected to theexteroceptive and proprioceptive sources of multimodal stimulus informationare reentrantly connected to form a global mapping. This global mapping formsa dynamic loop that correlates the information about the organism (proprioception) to the information about external events (exteroception). In this way,the selection of neuronal groups in particular local mappings results in theemergent distinction between an 'in here' perspective on what is 'out there'.Edelman points out that 'sensorimotor activity over the whole mapping selectsneuronal groups that give the appropriate output or behaviour, resulting in categorization' (1992: 90). Appropriateness of output or behaviour is determinedby internal criteria of value that constrain the ways in which the organisminteracts with its environment. In this way, categorization emerges on the basis ofexploratory activity that matches, in biologically appropriate ways, the physiological life-supporting systems of the organism (Edelman 1992: 90-1). Thedistinction between 'self and 'nonself' is not, as we shall see in more detail inchapter 5, a given; instead, it emerges as a consequence of the ways in which theindividual's sensori-motor engagements with the nonself, including mostcrucially other conspecifics, select neuronal groups that eventually result in thecategorical distinction between self and nonself. The fundamental importanceof the orientation to the other in the infant's very early proto-semiosis suggests,once again, that this proto-interpersonal orientation is founded on internalcriteria of value which define these engagements with the other - most criticallythe mother or other primary caregiver - as an appropriate domain of activity.That is, as an activity that is essential for the infant's further development. Thus,the infant's internal value criteria constitute the biological initiating conditionson level L-l of a semiotic trajectory on focal level L which has its beginnings inthe infant's performing of behavioural routines directed at fulfilling its mostbasic needs as a newborn - nourishment, warmth, affection, contact with otherconspecifics as sources of information that is needed for the further development of its ontogenetic trajectory.

The perceiver cannot avoid perceiving, and, therefore, being aware of him- orherself in the act of perceiving something. This means that the perception of


some environmental event (exteroception) and the perception of our own act ofperceiving (proprioception) are simultaneously implicated in perception.Moreover, our conscious attention can be focused on one or the other of the twopoles of what is now seen as a single perceptual structure, albeit with differentpossible poles of attention (chapter 5, section 4). In other words, awareness ofself and nonself are both implicated in the structure of the perceptual act. Thissuggests that consciousness, here seen as a meaning structure at all levels of theimplication hierarchy of icon, index, and symbol, is based upon the reentrantmapping in the brain of this perceptual distinction between informationpertaining to the self (proprioception) and information pertaining to thenonself (exteroception). This means that (1) we are conscious of somephenomenon of experience which we are perceiving, performing, imagining,thinking about, or remembering and (2) we are conscious of that experience asours, i.e., as being indexically referable to our own self, as designated, forexample, by the English pronoun T. That is, we have an experience ofsomething and an experience of our self as the experiencer of that experience.Consciousness is the result of the semiotic mediation and re-structuring onhigher levels of organization of this relation. As stated earlier, this semioticmediation can take place at all levels of the implication-cum-presuppositionhierarchy of icon, index, and symbol.

Perceptual-motor, conceptual, and symbolic forms of categorization in thebrain (Edelman 1992) therefore constitute an implication hierarchy of differentsemiotic levels. The relational structure of consciousness, as described above, canbe integrated to these levels of increasing semiotic specification and consequentlyreorganized. The self-organizing character of these processes means that the braindoes not 'represent' experience in terms of pre-existing, stored categories, images,symbolic tokens, and so on. Meaning at all integrative levels emerges in andthrough the interaction between the categories in the brain and the phenomenaof experience, as mediated by a higher-order system of interpretance.

I shall now discuss this in more detail, starting with the expression stratum.First, speech sounds are iconically grounded in the physiological processes of thespeaker's body (chapter 3, section 4, pp. 118-20). In this way, speech soundsafford information which specifies the embodied states and feelings of thespeaker. In this perspective, the perceived sounds are necessarily tied to orgrounded in and provide information about the speaker's body, its feeling states,and so on. Secondly, there is information in speech sounds which indexes thespatial location of the speaker in the environment relative to the listener (Biihler1990 [1934]: 125), the personal identity of the speaker, and so on. Thirdly,speech sounds contain information which enables the listener to perceive thephonological categories, units, and structures which belong to the sound - i.e.phonological - system of a given language such as English. This kind of information constitutes the symbolic dimension of the expression stratum of spokenlanguage. Whereas the iconic and indexical information in speech sounds is, tovarying degrees, causally tied to its environmental source (the speaker) as aperceptual stimulus which it specifies, the symbolic order of speech soundsspecifies, on the other hand, the phonological units and structures of a givenphonological system. Two important implications derive from this. First, the


phonemes, syllables, and so on, in a given utterance are instantiations of thephonological categories of a given language system. They are not, therefore, tiedto specific speakers' bodily states or identities, but are potentially shareable by allspeakers of the given language. Secondly, this order of information in speechsounds symbolically construes or realizes a more abstract level of semiotic organization, viz. the lexicogrammatical units and structures on the content stratum(see also chapter 3, section 9, pp. 139-46, on the brain as somatic recognitionsystem). In Gibson's terms, we might say that this third order of information symbolically specifies the more abstract content stratum of some language system.

I shall now turn my attention to the content stratum, again with a view to analytically refracting this through the presupposition-cum-implication hierarchy oficon, index, and symbol that I proposed in chapter 3, section 4, pp. 118-20. First,the ambient stimulus information is iconically grounded in some environmentalevent which is the cause of the variants and invariants in the stimulus information picked up by the observer's perceptual systems. This information directlyspecifies that event for the observer (Gibson 1986 [1979]). Secondly, stimulusinformation in the environment can specify events which are notnecessarily present in the here-now of the stimulus information which is pickedup by the observer. Such events can be reconstructed or inferred on the basis ofthe available stimulus information, though not necessarily present to theobserver. In this case, the stimulus information indexes some environmentalvalue which may be spatially or temporally displaced with respect to theinformation picked up by the observer. Thirdly, environmental events which arespecified by ambient stimulus information, as well as those which have beenextracted from the ambient flux of environmental information, can be symbolically construed in and through the semantic categories of a given languagesystem. In this way, we respond to and orient to the world not directly - Le.perceptually - but indirectly, through the lexicogrammatical and semanticresources of a shared system of social meaning-making.

Now, it is important to point out that on both sides of this relationship - viz.expression and content - all three levels of the implication hierarchy from icon,index, to symbol are organized on the basis of degrees and kinds of difference.It would be wrong to equate the iconic level with the 'sensations' whichempiricist theories of perception and consciousness consider to be the basis ofconsciousness. Sensations, in this empiricist view, are devoid of categorial orconceptual value. However, the iconic mode is still founded on topologicalcontinuous (analogue) differentiation (topological-continuous variation) as aresource for discriminating degrees of difference rather than type of difference(Wilden 1981: 20). On the other hand, the symbolic dimension of language isfounded on typological-categorial difference, i.e. discrete digital categories ofdifferent types. This applies equally to both expression and content. This is verydifferent from the doctrine that consciousness consists of 'thoughts' and'sensations'. According to this doctrine, 'sensations' are non-conceptual and aremerely 'felt' as subjective experience on account of their impact on our sensorysystems. It is further assumed by this doctrine that organisms perceivephenomena on the basis of the 'qualities of their nerves and neurons, not withthe properties of the external world', as Gibson (1983 [1966]: 56) has eloquently


put it. In this view, pure sensations are, in turn, transmitted to the brain by thenerves and turned into thoughts and ideas. Only the latter are considered tohave conceptual value. Alternatively, the semiotic model of stratification, asdefined here, suggests that value-laden differences - topological and typological- perfuse all levels of consciousness right down to the iconic level of perceptualawareness. This is equally true for both our experience of phenomena and ourexperience of our own bodies. This observation also ties in with the intentionalcharacter of consciousness, along with the fact that the objects of consciousnessare experienced as being instances of this or that category of phenomenalexperience. Consciousness at all levels is embedded in higher-scalar systems ofinterpretance. For this reason, consciousness is not uniquely generated 'frombelow' by physical states of the brain.

The presupposition-cum-integration hierarchy of icon, index, and symbolsuggests the need for an internalist perspective on the question of consciousness.In this perspective, we are inextricably inside the system that we are observing.There is no objective viewpoint whereby we can observe and analyse the objectsystem from the outside. Nevertheless, theories of consciousness have tended torelate to the phenomenon of consciousness adaptively, i.e. as something topredict or explain. For these theorists, organization and pattern in the neuroanatomical structures and processes of the brain is the thing to be explained.These theories continue to view consciousness as if it were a system we canobserve from the outside. They fail to see that the attempt to find and explainthe seat of consciousness as something in the brain is to look in the wrong place.Both the intrinsic awareness of one's own acts of consciousness and the attribution of consciousness to others is a question of a joint, dialogic relation betweenobserver and object system. The emergent order or form of consciousnessnecessarily requires us to abandon naturalism and eliminative materialism infavour of a perspective based on meaning. Consciousness has been isolated fromits contexts on the assumption that those contexts do not impinge on the natureof consciousness itself in significant ways. Moreover, the' source of order andcoherence has been located solely in the brain; it is as if other aspects of theworld do not have their own order, and in ways which affect the nature ofconsciousness in substantial ways. If, on the other hand, the content ofconsciousness is defined on the basis of the meaning which is stored in thesystem, where meaning here refers to the internal form and structure ofconsciousness, it is clear that it is the system's internalized encounters with signsand the attempt to discover their systems of interpretance, which specifies thestructure of consciousness. The problem for the observer, who inhabits the sameecosocial scale as the individual observed, is one of having to discover andinterpret the signs which tell him or her something about the content of theother's consciousness. As we shall see in greater detail in later chapters, consciousness can be theorized as a highly specified meaning system in theperspective of the self on all integrative levels of the semiotic hierarchy of iconic,indexical, and symbolic integrative levels.


16. An Alternative Reading of Descartes in the Internalist Perspective ofInterpretive Activity

The observations I made in the previous paragraph suggest an alternativereading of Descartes's famous observations concerning the passers-by whom heobserved from his window in the street below. Here is the passage in questionfrom Descartes's SecondMeditation:

I am taken aback, nevertheless, when I reflect upon the weakness of my mind,and consider how prone it is to error. For, however I may ponder over all thisin the silence of my thoughts, words themselves hold me up, and I am almostdeceived by the terms of everyday speech. For we say that we see the same pieceof wax if it is presented to us, and not that we judge it to be the same becauseit has the same colour and the same shape. And from this I would almostconclude that I know the wax by the sight of my eyes, and not by the scrutinyof my mind. I chance to look out of the window upon the passers-by in thestreet below, I have no hesitation in saying that I see men, although their hatsand cloaks may be no more than a disguise for clockwork figures. Yet I judgethem to be truly men; and so, what I thought I saw with my eyes, I really knowthrough the sole power of judgement which resides in my mind. (Descartes1960 [1641]: 114-15)

Generally speaking, Descartes's observations have been understood in the lightof the externalist perspective (see Shanker 1998). This perspective assumes thatthe observer (Descartes) and observed (the persons in the square) do not existwithin the same system of interpretance. It is assumed that the two perspectivesare not commensurate. In the internalist perspective, on the other hand, we cansee how Descartes attempts to interpret the people he observes in the squarethrough his own interpretative activities. Moreover, it is possible in this perspective to see that Descartes is enmeshed in a universe of signs. It is these signs alonewhich provide the interpretative frameworks of his (and our) observational perspectives and viewpoints. In the externalist perspective, the referents of the signsare already given. In the internalist perspective, on the other hand, the world isconstrued through the systems of interpretance which mediate our encounterswith the world and allow us to endow it with meaning.

The first thing to note about Descartes's discussion is the focus on meaningfulpattern and order in what he observes. Thus, order can be observed in the formof men crossing the square, rather than just automatons wearing hats and cloaks.Descartes attributes this perception to a faculty of judgement whereas I wouldprefer to say that it is a case of the operation of what Salthe calls 'the constraintof informational closure' (1993: 133). This constraint depends on the observer'schoice of categories and the extent of their specification. Salthe draws onDretske's definition of information as 'the most highly specified digital contentan analog percept can have' (1993: 134). The scene which Descartes attends tois an unbounded visual percept. For this reason, it is an analogue percept. Asdiscussed in section 6 above, the experiential grammar and semantics oflanguage are one typological-categorial (digital) resource which affords the


observer a high degree of symbolic specification of the stimulus information heor she picks up in the ambient energy flux. Given that Descartes clearly possessesthe digital category 'man', the men he sees crossing the town square areperceived as men, rather than as less specified hats and cloaks which concealautomatons. The point is that the experiential semantic categories that are usedselectively by Descartes contextualize the given percept as a grounded instantiation of the given category. The fact that Descartes invokes this latter categorization demonstrates that very many different possibilities, to varying degrees ofspecification, may be implicated in the interaction between observer and object.This is, of course, the logic of the presupposition-cum-implication hierarchy oficon, index, and symbol. The most highly specified symbolic categories alwaysimplicate less specified categorizations on many different levels of specificationin the overall hierarchy of relations. This applies equally to the relations amongcategories within, say, the symbolic mode, as well as to the relations among thesymbolic, the indexical, and the iconic modes. Thus, the category 'hat and cloakwearing automatons' could function as the type-categorization of the observer'sperception. However, this category is overridden by the more specified category'men'. The fact of this overriding suggests that the former categorization is alower-level categorization which relates more closely to the observer's perceptualcategories, rather than to the symbolic categories of language.

Descartes's observations have been assimilated to, and have been seen as instigating, the modern cognitive science paradigm. According to this perspective,the mind is a field of representations which are delivered to consciousness byperception. Moreover, the radicalism of Descartes's view is upheld to lie in thenotion of perception as judgement, rather than the merely passive receiving ofsensations, as in the empiricist tradition (Seager 1999: 5). Nevertheless, it isintriguing to consider the fact that Descartes's linguistic construals of both hisown activity as perceiver as well as the phenomena he perceives - the men in thesquare - do not refer to the lower-scalar level of the ambient energy flux whichsubtends stimulus information, which we do not perceive in any case, or to theactivities of the central and peripheral nervous systems. Both ambient energyand the activities of the central and peripheral nervous systems exist on a lowerscalar level with respect to either the 'I', corresponding to Descartes as observer,or to the phenomena of experience that are perceived by the 'I'. Rather, thelexicogrammatical choices in Descartes's discourse construe phenomena on thescalar level of the world which Descartes, along with the writer and readers ofthe present text, share. This is the scalar level of the town square, the men whocross it, and the human observers who would look at these men from thewindows of their apartments. Descartes's own language locates both himself andthe phenomena observed on this same scalar level. The experiential semanticcategories in and through which Descartes construes this observer-objectrelation means that the analogue percept which Descartes sees is re-construed asa series of typological-categorial distinctions by the lexicogrammar of the clausesin Descartes's text. This corresponds to the symbolic mode in the presuppositioncum-implication hierarchy of icon, index, and symbol. These symbolic categoriesare the most highly specified categories with respect to the topologicalcontinuous (analogue) character of the visual percept, which is unbounded. The


latter is less specified, and corresponds to the iconic level. However, this does notmean that the less-specified iconic level has been replaced by the symbolic one.What it means is that the symbolic level has integrated and subsumed - i.e. 'contextually completed' - the iconic (and indexical) levels. This has the furtherimplication that, in contextually integrating to its level the iconic and indexicallevels, the symbolic level contains or implies additional levels of meaning onlower integrative levels which can be accessed by further semiotic (interpretative) labour. Such interpretative labour entails the possibility of interpreting thenon-symbolic ground of others' conscious experience from our own perspective.(See the discussion of gaze in section 9 above, pp. 201-2.)

17. The Entropic Character of Meaning

It is once again important to point out that meaning, however defined, is notinherent in each of these levels. Instead, it is derived by interpreters who haveaccess to the relevant system of interpretance. In Descartes's case, his interpretation of what he observes from his window is based on linguistic categoriesavailable in his ecosocial environment. In Halliday's discussion, as cited insection 10 above, pp. 202-5, the child's categories - e.g. the high-pitched squeak- are considerably less specified, thereby limiting the child's reaction to the noiseof the pigeons to the iconic and indexical levels. The mother's response providesthe more highly specified symbolic category which the child thus far lacks.Halliday's (1975; 1978b; 1993) research on children's 'learning how to mean'shows that language development in the child amounts to a constitutive, and notmerely contingent, increase in stored meaning along the individual's ontogenetic trajectory (see Salthe 1993: 135). This is very different from Chomsky'scharacterization of the 'underdetermined' and 'fragmented' character of theinputs which the child receives from the outside (1976: 10). The alternative tothe Chomsky-type view rests on the view that the informational, and hencemeaning-bearing, capacity of a system is entropic. This has been formulated bySalthe as follows:

If information is to be taken as entropic, it must follow the equivalents of theFirst and Second Law of Thermodynamics. That is, in a system with informational closure, (1) what can be converted into information must be found tobe conserved, and (2) information itself must increase or remain constant.(3) If macroscopic configurations are to be considered information, we mustidentify a system in which they so function. These are not insurmountableproblems. We can dispense with the first of these three criteria by noting thatenergy is what is converted into information, either in the formation of a configuration or in its discovery and classification by another system. We needmerely show that an object system and an observer exist in the same physicalsystem to satisfy the First Law as well as it is ever satisfied in thermodynamics.

To show that macroscopic information in an internalist situation followsthe second law, I have proposed the following logical argument (Salthe1990b): It is sufficient to show that observer and object system(s), Si, are


together embedded in the same complex supersystem, for the followingreasons: If we were to decompose total system information (macroscopicinformation), H, into the potential information capacity of the supersystem,Cma" and that stored in its contained subsystem(s), H, then the informationin the global system would be

H must continue to increase, or at least cannot decrease, when we explicitlyconsider the information, Ho, stored in an observer, 0, as well. Then

H =Cmax+ H', + Ho

describes the total system, in which Hocould also be observed as contributingto H by another observer in the system coordinate with 0, if there were one.(Salthe 1993: 118)

Following these important formulations of Salthe in favour of the view thatinformation, and, hence, meaning, behave entropically, it is now possible to saythat (1) the meaning which is stored along the individual agent's historicalbiographical trajectory increases as the trajectory develops; (2) meaning isproduced on the basis of energy exchanges whereby energy is dissipated along atrajectory not only as heat energy, for example, but also as information which ispotentially meaningful, relative to the observer's categories; (3) the trajectory ofconsciousness emerges, is constructed, and is maintained on the basis of theindividual's exchanges at all integrative levels in the semiotic specificationhierarchy of icon, index, and symbol with its ecosocial environment. Themeanings that are stored in the individual's Innenwelt enable meta-redundancyrelations to be constructed between these and selected aspects of the environment. The individual is, therefore, able to selectively attend to the phenomenaof experience in his or her inner and outer environments. The observer,therefore, construes the phenomena of experience according to the degree ofspecification of the signs - iconic, indexical, symbolic - to which he or she hasaccess. Moreover, the notion of meta-redundancy shows that consciousnessalways entails a patterned relationship of contextualization between internalsemiotic configurations which are stored and elaborated in the individual agentand selected environmental features.

18. Consciousness as the Contextualization of Experience in thePerspective of the Self

Insofar as language and other semiotic modalities constitute an interfacebetween the central and peripheral nervous systems of the individual organismand the ecosocial environment, such systems are a means whereby 'inside' and'outside' can 'communicate' with each other. They are the means whereby theexternal world is assimilated to the central nervous system of the organism so as


to become an integral part of the organism and its trajectory. It is in this way thathigher-scalar environmental constraints impose limits on the degrees of freedom- the many degrees of difference - of lower levels. The constraints emanatingfrom the higher level re-set the probabilities of the dynamical processes on lowerlevels so as to direct and constrain their behaviour. This re-setting of probabilities applies to all forms of categorization in the organism, viz. perceptual,conceptual, and symbolic categorization. However, it is not a question of thecategories which the organism has in its central nervous system 'representing' or'referring to' an already given external world. Instead, it is a question of selectively attending to and orienting to 'objects' through processes of differentiation. Gibson (1983 [1966]: 270) points out that these processes are learned. Theorganism learns what to attend to. In the case of the perception of objects, 'it isthe detection of distinctive features and the detection of invariants underchanging stimulation' (Gibson 1983 [1966]: 270-1). Gibson continues asfollows:

The exploratory perceptual systems typically produce transformations so thatthe invariants can be isolated . . . And the action of the .nervous system isconceived as a resonating to the stimulus information, not a storing of imagesor a connecting up of nerve cells. (Gibson 1983 [1966]: 271)

Rather than the brain producing representations of the outside world, thecategories in the central nervous system resonate with environmental stimulusinformation. This is, I argue, equally true of perceptual and conceptual categorizations, which are elaborated biologically (Edelman 1992: 108-10), and thesymbolic categories of language and other semiotic systems. The symboliccategories of a given language system are a resource emanating from the higherscalar ecosocial system. Symbolic categories allow individual consciousness toresonate with phenomena of experience which are not necessarily present in theorganism's perceptual purview. In this way, the experiential categories which arerealized in the lexicogrammar of the clause enable the construal of symbolicobjects which individuals can selectively attend to and orient to. The experientialresources oflanguage are not used to 'represent' or to 'refer to' something whichis already given, and which these categories then slice up in pre-established ways.Rather, the phenomena of experience - objects of consciousness - are construedin and through the system of categories - the value-producingdifferences - which are internal to a given semiotic system. The resonancebetween organism and environmental information of which Gibson speaks isthe process whereby phenomena of experience are made to emerge from protosemiotic vagueness. Organisms selectively contextualize the relationshipbetween themselves and the environment by means of the symbolic and othercategories which are stored in the central nervous system in and through theepigenetic exchange of information between organism and environment.Individuals can therefore converge on the phenomena of experience andselectively incorporate them into their Innenwelten.

In the Western folk-theoretical tradition, consciousness tends to be accordedthe status of an internal, subjective reality which can be 'reported' or 'repre-


sented' by linguistic tokens. In this respect, it is accorded the status of an 'object'which can be referred to and predicated ofjust like any other entity in the world.This is a realist view, which is based on the assumption that there exists a 'reallyreal' world either 'out there' or 'in here', independently of the ways in whichthese entities are contextualized by the systems of meaning of a given ecosocialsemiotic system. However, consciousness, no less than any other entity, is anentity in the supersystem transactions of some ecosocial system. It emerges andis contextualized as a phenomenon of experience in and through the systems ofcontextualizing relations of the ecosocial system. The meanings it has for the self(or for others) are meanings which are construable for the self in and throughthese same systems of contextualizing relations. Yet, consciousness is usuallyaccorded a metaphysical status as something which exists independently of thepossible ways it may be construed by a given meaning system. In ecosocialsemiotic theory, the 'reporting of' or 'referring to' conscious experience, or eveninternalized reflection on it, is always part of a larger pattern of contextualizingrelations which includes the experience itself. The notion that consciousness canbe linguistically 'reported' implies a referential theory of meaning in which theobjects that are 'referred to' are external to the (linguistic) act of referring. Inthis externalist view, consciousness is a content that exists in its own right and isexternal to the contexts in which we contextualize it. What is experienced, so tospeak, is always imminent in supersystem processes which are selectively contextualized by the meaning system. In this point of view, consciousness is, therefore,a constitutive part of those contexts, rather than being external to them. It thusmay be more appropriate to say that consciousness is, from the point of view ofthe insiders in some social situation, indexically invoked as locatable at a givensource - e.g. first-person or third-person. At the same time, it is symbolicallyconstrued in terms of the experiential categories of the lexicogrammar oflanguage. The point is that, in this view, consciousness is indexically invoked inthe given context from the point of view of the insider-participants in thesituation. The linguistic practices of construing consciousness in self and othermake no such distinction between what is external to speech and lurking'behind' speech in the mind of each individual. Rather, consciousness is indexically invoked as being internal to the speech practices themselves in the sensethat any internal phenomenon inside the 'head' only has meaning if it iscontextualized by the categories of some meaning system.

Gibson (1986 [1979]: 255) has formulated a view of non-perceptual awarenesswhich is useful here. Gibson points out how, in the older mentalist psychology,higher mental processes were opposed to sensations, and seen as the operationsof higher-order faculties in the mind. In this regard, Gibson mentions suchmental operations as remembering, thinking, conceiving, inferring, judging,expecting, knowing, imagining, dreaming, and rationalizing. According toGibson's ecological perspective, these may be explained as follows:

To daydream, dream, or imagine wishfully (or fearfully) is to be aware ofsurfaces or events that do not exist or occur and that are outside the limits ofpossibility.

These three kinds of nonperceptual awareness are not explained, I think,


by the traditional hypothesis of mental imagery. They are better explained bysome such hypothesis as this: a perceptual system that has become sensitizedto certain invariants and can extract them from the stimulus flux can alsooperate without the constraints of the stimulus flux. Information becomesfurther detached from stimulation. The adjustment loops for looking around,looking at, scanning, and focusing are then inoperative. The visual systemvisualizes. But this is still an activity of the system, not an appearance in thetheater of consciousness. (Gibson 1986 [1979]: 255-6)

The symbolic possibilities of language and depiction, to mention two semioticsystems with fully developed symbolic possibilities, mean that neither of thesetwo systems is tied by indexical necessity to entities already present in theenvironment. The lexicogrammatical and semantic distinctions on the contentstratum of language, for instance, do not refer to the ecosocial environment inany direct way. In learning the contextual meta-redundancies between linguisticforms and particular classes of environmental events, linguistic forms enter intocontextualizing relations with invariant stimulus information which is extractedfrom different kinds of environmental events (visual, kinesic, auditory, haptic,and so on). Moreover, the learning of these contextual redundancies cuts acrossthe stimulus-response dichotomy, 'for they can be vocal-motor or manual-motor'(Gibson 1986 [1979]: 281). Importantly, the learning and further elaboration ofthese redundancies between classes of linguistic and other semiotic forms andperceptual invariants is a matter of the progressive attunement of the system tofiner and finer differentiations. The attunement to these goes hand in hand withthe elaboration of a sense of consciousness along a time-bound trajectory. Thisview suggests that the linguistic and other semiotic resources for invokingconsciousness in one's self and in others function both to indexically invoke andto symbolically construe informational invariants which have been, as Gibsonputs it, detached from stimulation. In the absence of stimulation from theambient energy flux, the activity of exploring the environment for stimulusinformation has then been re-directed 'inwards' to invariants and differentiations to which the system has been sensitized through experience.

This last point further suggests that the contextualized uses of these linguisticforms may 'resonate' with perceptual invariants that have been de-located fromthe external environment and re-located in the individual's Innenwelt. In bothcases, it is still a question of the situationally-specific contextual redundanciesbetween such invariants and the semiotic forms used to index and symbolicallyconstrue them. The presupposition-cum-implication hierarchy of icon, index,and symbol is relevant here as well. Consciousness is always relative to observercategories. This is so whether the object system is external or internalphenomena of experience, or even one's own activities of consciousness (metamentation). If the categories of the nervous system become increasinglyspecified as it develops in ontogeny (Salthe 1993: 135), then it seems plausibleto suggest that the development of the organism involves the increasing specification of signs from the initial vagueness of the topological-continuous variationcharacteristic of the iconic to the highly specified typological-categorial distinctions characteristic of the symbolic (see chapter 1, section 5, pp. 23-6). This does


not mean that the symbolic mode replaces the other modes as the organismdevelops. As we have seen, the symbolic implicates the indexical, which, in turn,implicates the iconic. What it does mean is that the observer has informationstored in him or her such that an object system can be specified in terms of thefull range of possibilities afforded by the hierarchy of iconic, indexical, andsymbolic modes of semiosis.

Therefore, at all levels of organism-environment information exchanges andcross-eouplings, from the perceptual and conceptual categorization of primaryconsciousness (Edelman 1992) to the symbolic categorization of higher-orderconsciousness, the body-brain complex is embedded in higher-scalar systems ofinterpretance of varying degrees of specification. It is only through such systemsof interpretance that the organism can determine whether information which isobtained from the environment is potentially meaningful for the organism. Forexample, perception, rather than being the passive receiving of data from theoutside, occurs on the basis of a reciprocal adaptation of both organism andstimulus information to each other. A given organism is not attuned to all of theinformation that is potentially available in the stimulus flux of ambient energy.Instead, it is selectively attuned and adapted to only some of the total possibilities in its environment. In fact, it would be more appropriate to refer to thereciprocal eo-adaptation of both organism and environment (self and nonself)to each other. Thus, the release of energy provides information about anenvironmental event relative to a potential observer. This means that stimulusinformation in the environment and observer are both reciprocally conditionedby the existence of a higher-scalar context which links them together.

19. The Embedding of Consciousness in a Higher-scalar System ofInterpretance

The existence of a higher-order context means that the relationship between twointeracting entities, A and B, is not random, but occurs with a certain degree ofprobability. That is, the higher-scalar level C functions as a boundary conditionon the transaction and the signs at the A-B dyad level. In the course of theirinteraction, A and B reciprocally adapt to each other. This means that they aretransformed by the contact between them. Whereas ambient energy about indefinitely many other environmental events is not available to the organism,because it is not evolutionarily designed so as to be selectively attuned to these,the reciprocal nature of the contact between A and B means that there exists arelationship of complementarity between them. This necessarily occurs against abackground of other sources of stimulus information which are not selected bythe A-B relationship. The situation described here is a very general one, whichapplies to all biological phenomena. This suggests that what we call 'dialogue' inthe specifically human case is a more highly specified subclass of the kind ofreciprocal adaptation of A and B that is characteristic of all biological organisms.This general principle can be extended all the way right back to, for example,the trophic orientation of amoebas to light. We do not normally think ofamoebas as engaging in 'dialogue'. The 'dialogue' between amoeba and light


source would be, following Salthe (1993: 18), the vaguest and least highlyspecified relationship of reciprocal adaptation between some A and B. Thespecifically human dialogue would be, at least from our human perspective, themost highly specified case - the one for which we normally reserve the term'dialogue'. The point is, as we saw earlier (chapter 1, section 8, pp. 34-9; chapter3, section 4, pp. 118-20), that the higher, more specified levels integrate thelower, less specified ways. The latter do not transcend the former, though thecoming into existence of new higher-scalar boundary conditions will re-interpretthe lower levels in new ways.

In the human case, the arguments presented above suggest that both primaryconsciousness and higher-order consciousness are embedded in systems of interpretance of varying degrees of specification. A conscious observer is alwaysconscious of something, whatever this something might be. However, theobserver's consciousness of something can only occur on the basis of stimulusinformation, whether perceptual or non-perceptual (see above), which constitutes potential information about some 'inner' or 'outer' phenomenon. Thismeans that such stimulus information, relative to the observer and his or hercategories, is a potential sign of some phenomenon of experience. It isimportant to remember here that observer and stimulus information - i.e.potential signs - are reciprocally adapted to each other, for what the informationmeans depends on the observer and the categories stored in his or her centralnervous system. The categories stored in the observer themselves derive from ahigher-scalar system of interpretance in which both observer and what theobserver is conscious of are necessarily embedded. This shows very clearly thatthe observer is, in part, a result of the signs he or she can orient to and adapt to.By the same token, consciousness of something is not reducible to unmediatedcontact between the 'mind' and the actualities of the world. In primary consciousness, here-now samplings of the environment and their categorizationentail the adaptation of the organism to the iconic and indexical signs which itencounters and interprets. These signs mediate between the organism's activitiesin the world and those here-now aspects of it that the organism interacts with.This is a less highly specified form of semiosis with respect to the more highlyspecified symbolic forms of higher-order consciousness. Nevertheless, it remainsa form of reciprocal adaptation which is necessarily constrained by higher-orderboundary conditions which act as a system of interpretance of these iconic andindexical signs. In the human case, the contextual integration and further reorganization of these levels by the symbolic level of higher-order consciousnessvastly expands those aspects of the world with which the organism can interact.This depends on a system of symbolic possibilities which can connect andintegrate the individual body-brain complex with space-time scales beyond thehere-now one of primary consciousness.

Current theories of consciousness tend to emphasize the individual andsubjective character of conscious experience. This emphasis downplays or leavesout of the picture the way in which consciousness entails observer perspectivesand viewpoints. Moreover, such perspectives and viewpoints are shareable withothers who belong to the same system of relations. It cannot simply be a questionof a physical system - the brain - representing the outside world. The observer


interacts with the phenomena of experience in and through a given system ofinterpretance. Systems of interpretance do not only entail a system of categoriesfor construing experience and interacting with it; they also entail systems ofsocial meaning-making practices which determine how these categories aredeployed in contextually relevant ways. Theorists of consciousness have debatedthe question of the observer's intrinsic awareness of his or her conscious acts ofexperience (see Natsoulas 1998, 2000). In my view, this question closely relatesto the interaction between observer and object observed in and through a systemof interpretance. A number of levels can thus be distinguished:

1. the environmental stimulus information (the object system) relative topotential points of observation;

2. the object system in interaction with the observer;3. the observer who observes the interaction between object and observer.

The above schematization once again highlights the fundamental importance ofthe stratified nature of semiosis and its role in higher-order consciousness. Ipointed out in chapter 1 that language is a stratified system relative to an agentwho is cross-coupled to the system via the two interfaces of expression andcontent. The agent constitutes a break or an interruption in the potentiallyinfinite recursivity of the levels of relations involved. It is this robustness whichallows agents to construe models of the world, the phenomena of experience, atthe same time that such models are cross-eoupled to the bodily dynamics of theobserver-agent (chapter 1, sections 1, 2, pp. 3-11; chapter 2, section 1, pp.59-67). By the same token, semiotic stratification also entails the capacity of thesystem to be turned back on itself: there is also me ta-semiotic reflection on thesystem as an object of observation. Given my earlier argument that this alwaysentails the cross-coupling of an observing agent to the system of observation (thesystem of interpretance), it follows that what is observed in meta-semiosis is notlanguage per se, but the interaction between language and the observer-agentwho uses it. This corresponds to level 3 above.

The stratified nature of semiosis thus constitutes the permeable interfacebetween body-brain and ecosocial environment. The bodily dynamics of theagent are both kept separate from, at the same time as they enter into communication with, the phenomena of experience through the processes of symbolictransduction discussed earlier. The agents who inhabit a given ecosocial system,the phenomena of experience, and higher-scalar systems of interpretance, areincorporated into a single system of relations on account of the stratification ofsemiosis into a number of different levels of organization. It is this fact whichenables the intersection of radically different scalar levels in symbolic or higherorder consciousness. A stratified semiotic system such as language, gesture, ordepiction faces two ways. As we have seen, semiotic systems interface with boththe body of the agent and the phenomena of experience along the expressionand content strata, respectively.

Take something which is, canonically speaking, seen as cerebral, solitary, anddisembodied such as reading a book. The categories of written expression(graphemes, ete.) and of spoken expression (phonemes, ete.) emerge as figures


against a background of topological-continuous variation and vagueness. Forexample, graphemes and the larger graphological units, such as orthographicwords, orthographic sentences, ete. which these combine to form, emerge inrelation to the ground of the treated surface onto which someone has tracedthese visual invariants. The process of visual scanning of these graphic tracingson a surface employs the smallest level of the visual system, that of the fovealfield of view of the two eyes. Thus, the hemispherical field of the stationary ornear stationary head during the activity of the sampling of the arrested array ofvisual structures on the page of a book is explored by saccadic movements of theeyes (Gibson 1983 [1966]: 258-9). In this way, the body of the observer isintegrated with the specific principles of organization of the expression stratumof the written language by means of his or her active perceptual exploration ofthe visual invariants which are picked up. However, these visual invariants arethemselves internal to a still-larger system of relations on other scalar levels oforganization in a still more global system of relations.

Furthermore, the visual invariants which are picked up resonate in thereader's brain with lexicogrammatical and semantic categories on the contentstratum with which they contextually redound. That is, the specific organizational principles in operation on the expression stratum are symbolicallyreconstrued in relation to another set of organizational principles on thecontent stratum (see the discussion in chapter 3, section 9, pp. 139-46, on thebrain as somatic recognition system). It is this which permits symbolic transduction between the organism and his or her outside. Meaning is not, then,something which is isolable in just one part of the system. Rather, it is a result ofthe relationship between the agent-observer and the overall system of relationson all levels.

There is no problem as to how a sequence of graphemes can be transformedinto a larger unit such as a word, sentence, and so on. As the eyes scan thetracings on the page, they do not make discrete samples of each element(grapheme, ete.) in the series. Instead, the entire sequence as a whole constitutes a structured visual array which is unaltered by the reader's moment-bymoment scanning of it. The central nervous system detects this structure as awhole as it unfolds in time during the scanning process. It is not a question, asGibson points out, of comparing successive visual sensations with graphic orother images stored in memory:

The invariance of perception with varying samples of overlapping stimulationmay be accounted for by invariant information and by an attunement of thewhole retino-neuro-muscular system to invariant information. The development of this attunement, or the education of attention, depends on pastexperience but not on the storage of past experiences. (Gibson 1983[1966]: 262)

The point is that the reader's active scanning in order to obtain visual information produces a series of transformations and transitions. Furthermore, theproprio-sensitive nature of the visual system means that the changes specify theactivity of visual scanning and exploration, rather than visual stimuli which


passively register on the retina. Memory of previously stored graphic and otherimages is not required. There is no comparing of present stimuli with suchstored images. Instead, the graphic tracings on the page are assigned theirmeaning through a process of integration between past sampling and presentsampling.

5 The Metafunctional Character ofConsciousness: Some CorrelationsBetween the Neurobiological andSemiotic Dimensions

1. The Contextual Character of Consciousness

Linguistic and other semiotic systems are not self-contained and autonomousentities. While they have their own intrinsic principles of organization and functioning, viz. a relative functional closure with respect to their higher-scalarecosocial environments, they are also influenced and changed by the ways inwhich human beings jointly deploy the meaning-making resources afforded bysystems of social signs. Social semiotic systems exist and are maintained inrelation to higher-scalar ecosocial levels which exercise constraints or boundaryconditions on them. Language and other social semiotic resources are openboth in relation to the social meaning-making practices of social communitiesand in relation to the architecture and dynamical processes of the body-braincomplex. Moreover, the notion of embodiment, as distinct from the biologicalorganism which subtends our embodiment, has been proposed as an alternativeto the current new wave of mechanistic thinking about the body and our relationship to it, as seen in the pervasive and reductive view of the body as a bottomup assemblage of parts (organs, etc.) which can be replaced, exchanged, andeven sold on the black-market thanks to modern medical science andtechnology. In the ecosocial semiotic perspective of the present study, I haveemphasized the importance of the expression and content strata as permeableinterfaces which cross-eouple meaning-making processes to both the individual'secosocial environment - his or her Umwelt- and the body-brain complex, i.e. theindividual's Innenwelt, or invironment. The early in sights of Saussure andHjelmslev concerning the stratified nature of the linguistic sign have beentheoretically reconstituted in the ecosocial semiotic perspective to this end.Moreover, the two dimensions are not constitutively separable and must,therefore, be seen as components of a single self-organizing system of relationson diverse levels.

Modern theories of consciousness, on the other hand, have stressed theautonomy and individuality of consciousness to such an extent that they havefailed to throw off the shackles of their Cartesian solipsistic inheritance.Consciousness, which can never be autonomous in this sense, always exists in thecontext of both the lower-scalar and the higher-scalar dynamical relations andprocesses which produce and sustain it. Lukacs (1980 [1978]: 105-6) has

THE METAFUNCTIONAL CHARACTER OF CONSCIOUSNESS 237

observed that the human body can exist without consciousness, but that consciousness cannot exist without the lower-scalar biological processes that sustainconsciousness in the body-brain complex. This shows that lower-scalar physical,chemical, and biological constraints work through consciousness. By the sametoken, consciousness constrains these lower-scalar dynamical processes. Thedivision of labour between body and soul is, thus, an untenable one; it is onewhich reflects actual social divisions of labour such as the current one between amedical science and technology of the body qua organism and a religious one ofthe individual soul that is housed in the body. Lukacs's understanding of themastery of consciousness over biological processes can now be seen as referringto the ways in which higher-order or symbolic consciousness constitutes the mosthighly specified level in an integrative hierarchy containing progressively lessspecified outer levels. Thus, the individual is comprised of a number of differentintegrative levels (Salthe 1993: 70), which are constructed from the viewpoints ofdifferent (theoretical) discourses, viz. physical, chemical, biological, ecosocialsemiotic, and psychological. From the perspective of an ontology of social being,to paraphrase Lukacs, each of these levels constitutes a set of constraints, all ofwhich operate on the overall system of relations. With reference to the individualagent, higher-order consciousness is the most highly specified level because it issymbolic in nature and function. It cannot, therefore, be reduced to or explainedin terms of the constraints operating on the lower physical, chemical or biologicallevels. Instead, it contextually integrates these to its own level so as to completethem in the time-bound process of constructing and construing an individual self.Lukacs thus draws attention to the ways in which constraints pertaining to thehighest level - viz. the symbolic - operate on the lowest, least specified level ofthe physical existence of the body-as-thing obeying the physical laws of nature.

We now come back to the question of how semiosis acts on and shapes consciousness. This is not, in my view, a matter of linguistic and other semiotic formscausally acting on individual consciousness in a direct, unmediated way. Instead,the lexicogrammatical resources along with their typical patterns of use mediateand entrain consciousness to their patterns. This is so in the sense that their formsand the categorial distinctions these construe have evolved as a result of theteleological positings of individuals. Consequently, these forms are not reducibleto the biological mechanisms which sustain them in the individual. Human socialsemiotic systems have arisen in relation to social practices and orientations, ratherthan to purely biological ones. For this reason, the use of particular semioticforms requires that social agents appropriately orient to their social contexts inand through the discursive deployment of these forms, rather than to the formsperse. Semiotic forms of all kinds are mediated by higher-order meaning-makingpractices whereby agents position and orient themselves to each other and totheir contexts by means of specific co-deployments of such forms. (See Lemke1990; Thibault 1991a for earlier formulations of the relationship between socialsemiotic resource systems and social meaning-making practices.)

The question of how individuals are integrated with space-time scales that gobeyond the here-now environmental samplings that are characteristic of primaryconsciousness require us to return to the following question: what is the natureof the higher-scalar level(s) above the individual? In order to answer this


question, I propose that we need to investigate the social meaning-makingpractices in and through which linguistic and other semiotic forms are eodeployed in both social interaction and inner semiosis. Consciousness, Isuggested above, entails the recognition of meaning in our experience. Meaningdoes not refer uniquely to linguistic meaning, though this is clearly an importantdimension of symbolic or higher-order consciousness. Primary consciousness,which is based on biologically driven perceptual and conceptual categorization,is meaningful because it involves the perception of pattern in experience andthe ability to assign form to that pattern by way of conceptual categorization.Nevertheless, these patterns are not uniquely private or individual experiencesthat are locked away in the head of the individual. They are always linked tolarger-scale supersystem transactions between individuals and their environments. What are often referred to as 'inner' cognitive processes can, therefore,be seen as forms of meaningful pattern recognition on lower integrative levels inprecisely this sense. Higher-order consciousness does not mean that primaryconsciousness and its categorizations are transcended. Instead, the formerintegrates the latter so as to contextually complete it. Higher-order consciousness involves a higher level of meaningful activity whereby the recognition ofpattern and meaning in experience is linked to meanings and contexts whichare characteristic of a given culture as a whole. This indicates the need torepresent consciousness contextually, rather than seeing it as simply being'about' something which it represents. Rather than a simple pairing of consciousawareness and the object of this awareness, the approach I am developing hereassumes that there is a pattern in this relationship and that this relationship hasstructure at logically higher orders of abstraction than either the conscious selfor the observed phenomenon (object of consciousness).

In this chapter, I shall explore the contextual character of consciousness byrefracting the questions raised in this section through the metafunctional theoryof linguistic and other modalities of semiosis that has been developed insystemic-functional linguistics. In so doing, I shall argue that consciousness at alllevels is intrinsically organized in terms of a diversity of (proto)-metafunctionalparameters. These parameters help to define the various ways in whichconsciousness is integrated to both its lower- and higher-scalar environments.

2. The Metafunctions and the Shape of Consciousness

A central assumption of systemic-functional linguistic theory is that lexicogrammatical form is only describable with reference to semantics. In both formallinguistics and the information-processing model of cognition, on the other hand,syntax and neurophysiological brain processes are describable without referenceto meaning. In both cases, cognition consists in the explicit manipulation ofsymbols in the mind according to rules based on formal logic. Edelman's theoryof neural activities in the brain shows, on the other hand, that meaning isglobally organized in the brain in terms of self-organizing topological regions.Edelman's observations resonate well with the view that meaning is a multidimensional semantic space which is informed by a topological perspective


(Halliday and Matthiessen 1999: 70-1). Halliday and Matthiessen point out thatthe topological perspective allows for indeterminacy in the sense that differentregions of meaning overlap. Lexicogrammar is embodied in dynamicalattractors or constrained pathways through the multidimensional semanticspace. This observation is supported by the emergence of a lexicogrammar inthe transition from protolanguage to adult language (Halliday 1993: 96).

Lexicogrammar acts as a dynamical attractor for both bodily processes andthe phenomena of experience. This is so by virtue of its interface with both theexpression stratum and its interface with the semantics of the content stratum.Thus, lexicogrammatical neural pathways through the multidimensionalsemantic space entrain both meaning and sensori-motor activity as dialogicallyorganized and oriented meaning-making acts. The higher-order boundaryconditions of these dynamics correspond to the discourse genres and socialactivity-types of a given culture. Contextualized acts of meaning-making aremapped onto a vector space of cooperating neurons which constrains both thesensori-motor activities and the semantic features which are activated. Thissuggests that neural activities involved in meaning and consciousness can exertdownwardly causal effects on bodily processes so as to produce sociallymeaningful action. The individual agent's internalized meaning potential canthus be seen as a set of attractors which refer to the system's self-organizedinternal dynamics. However, this internalized meaning potential is neitherpre-given in the form of innate rules nor externally imposed by social norms.Instead, it is the elaborated product of the individual's transactions both with thehigher-scalar ecosocial system and the specific contexts in which meaningmaking occurs. Meaning is thus embodied in self-organizing neural networkswhich constitute a multidimensional semantic space.

The emergence of lexicogrammar in the transition from protolanguageto language is an epigenetic process based on the child's experience of andparticipation in the social meaning-making practices of the social group. Thedevelopmental progression from protolanguage to language entails that theemergence of language is an additional level of semiotic specification which,however, holistically integrates the prior protolinguistic stage to its own semioticlevel. Protolanguage is not transcended when language emerges; rather, it isrestructured by the higher integrative level at the same time as it continues toexert its own effects on the higher level (Salthe 1993: 85). The child's entrainment to the practices of the group likewise entrains his or her dynamic neurological organization. The move from the child's first symbolic acts, comprisingan expression and a meaning without a lexicogrammar, to the adult language isa catastrophic reorganization of the multidimensional semantic space and there configuring of the lexicogrammatical pathways through it such that fourdistinct regions of meaning are simultaneously attracted to these pathways.

In the adult language, the four metafunctional domains of meaning arestrongly coupled within the grammar of the clause, along with other levels oforganization below this (see chapter 3, section 12, pp. 154-8). The increasedcoupling of the four metafunctions is evidenced in the child's developmentalprogression from early protolanguage to the adult language, as demonstrated inthe longitudinal studies undertaken by Halliday (1975, 1978b, 1993), Painter


(1984), and Oldenburg (1987), as well as the work of Trevarthen. As thesestudies show, protolanguage does not exhibit this strong coupling of semanticfunctions which are intrinsic to lexicogrammatical form in the adult language.The coupling of the metafunctions occurs as the child develops a lexicogrammatical system which reorganizes the prior system, which consisted of vocalizations which realized situationally specific meanings without the presence of alexicogrammar. Language development is not caused by prior, organism-internalfactors, but by epigenetic experiences which foster the child's entrainment intothe practices of the adult system of those who surround him or her. The fact thatthese changes are accompanied by development of the frontal cortex during thiscrucial early period of language development raises the question as to how thedevelopment of the adult language in the individual, with its strong internalcoupling of metafunctional components in the grammar, contributes to theshaping of brain activity through the reentrant mapping of external linguisticactivity onto internal neural activity.

If language is intrinsically organized so as to enable individuals as complexsystems to semiotically construe, interact with, and create meaningful ties totheir ecosocial environments, then it should follow that the intrinsic metafunctional organization of language form in some way contributes to the shaping ofconsciousness itself. A number of important implications flow from this formulation. First, the notions of construing and constructing, in place of the idea ofrepresentation, indicate that the process of categorization is an active andcontextualized one. It is not a matter of acquiring a set of internal symbols whichrepresent external reality, as in the traditional cognitive science account.Secondly, the intrinsic metafunctional organization of semiotic forms furtherindicates that categorization has frequently been too narrowly defined in termsof the ideational function of language-as-thought. This tendency has been at theexpense of seeing language both as action and a means of preserving awholeness or unity of meaning in context such that individuals and themeanings they jointly make in their interactions with others are felt to be part ofa larger organizational whole which links embodied individuals to theirsurrounding milieu in textured and coherent ways.

Halliday has often argued that language and, by extension, other semioticresources have evolved and are organized as they are on account of the socialuses they have been put to over very many generations of evolutionary time (e.g.1978a: 187). In other words, our semiotic resources have eo-evolved along withthe social and biological contexts of their deployment. If consciousness emergesas the higher-order experiencing of one's own embodied experiencing, thenlanguage and other semiotic resources can be seen as having eo-evolved alongwith the emergence of consciousness as a way of mediating the experiencing ofexperiencing. It follows from this that our semiotic resources and their intrinsicdesign features are deeply implicated in the very nature of consciousness. This isnot the same as saying that all of consciousness can be reduced to questions oflanguage per se. Rather, the implication-cum-presupposition hierarchy of icon,index, and symbol shows that both primary and higher-order consciousness,both of which are modes of experiencing of experiencing, are mediated by thepossibilities of iconic, indexical, and symbolic semiosis. Experience itself shows


multilevel complexity, interactional relationism, and dynamic contingency, andthe task of our semiotic resources is to mediate these characteristics in theperspective of the self who undergoes the experience in question.

In this chapter, I shall explore the implications of the metafunctions for thestructure and functioning of consciousness on various levels of specification.I shall begin this discussion in the following section by first considering someof the vague and very general proto-metafunctional characteristics thatare discernible in the infant's earliest perceptual-motor engagements with theenvironment. As we shall see, the (proto)-metafunctional shape of consciousnessentails a contextual, rather than a solipsistic, view of consciousness.

3. Vague Contours of the Metafunctions in the Infant's Early Perceptualmotor Engagements with the Environment

Current accounts of consciousness, in reducing it to a question of physical brainstates and processes, or in failing to account for language other than as a devicefor 'reporting' or 'representing' individual, indeed soli psis tic, consciousexperience, fail to account for the fact that consciousness is itself context-bound,intrinsically dialogic, and meaningful through and through (see also Volosinov1973 [1930], 1976 [1927]). It is based on the viewpoints which agents can adopton account of the semantic phase space of meaningful alternatives that areavailable to them in a given ecosocial semiotic system. The question arises as tohow we can connect individual consciousness to the higher-scalar ecosocial system,and in ways that indicate the complex, open and non-linear dynamics of consciousness. In other words, how can we indicate its temporal and spatialgrounding, the richly patterned connectivity that may implicate networks ofrelations across many diverse space-time scales, the sense of being part of andrelated to a wider social and perceptual field of relations which includes others? Inorder to forge the semiotically mediated link between the intrinsic dynamics ofagents, the systems of meaningful alternatives available to them, and the higherscalar ecosocial environment, I shall propose that individual body-brain complexesand their dynamics are embedded in and entrained by systems of social meaningmaking practices which characterize a given society or some part of this.

Edelmari's theory of neuronal group selection has shown us how the brain, onthe basis of its ongoing sensori-motor exploration of the environment, selforganizes as a consequence of its context-bound interactions with the ecosocialand material environments within which it is embedded. In the first instance, thenewborn infant builds up perceptual-motor categories of objects, events, actions,and so on, through the multimodal reentrant mapping of his or her time-boundexperiences of seeing, listening, touching, moving, and so on (Thelen and Smith1994: 187). My re-analysis of Halliday's (1975, 1978b, 1993) account of howchildren 'learn how to mean' drew attention to the intrinsically multimodalcharacter of the dynamics of early language development (Thibault 2000a; andin this volume chapter 4, section 10, pp. 202-5). Thelen and Smith point outthat such multimodal correspondences 'are the cause of development ratherthan a product of development' (1994: 187). Furthermore, Edelman's notion of


in-built values shows how the newborn is selectively nudged along certain trajectories of multimodal perceptual-motor engagement with the world in order thathe or she is entrained into the patterns that are most salient and useful for hisor her further development. The values that Edelman speaks about arebiological and are part of our evolutionary inheritance. It is the child's spontaneous intrinsic dynamics which interact with and engage the ecosocialenvironment such that the time-bound activity of the system extends andchanges the values available to the child for further exploration and interpretation of its environment. Thelen and Smith have expressed the multimodal basisof the child's earliest experiences as follows:

... infants discover action and object categories through the cross-correlationof multimodal experiences. The disjunctive, heterogeneous experiences ofhearing and seeing and feeling and moving are all time-locked and changetogether as the infants' activity, state and actions change. We proposed thatinfants discover the category-relroant properties of objects and events invariant cues to shape, color, and texture across changes in lighting, perspective, and distance - through their multimodal and activity-dependentexperiences of objects. Likewise, they discover patterns of coordination thatprovide functional actions through movement and its perceptual consequences. These ideas turn the usual developmental stories backward:multimodal correspondences are the cause of development rather than aproductof development. (Thelen and Smith 1994: 187)

The early processes of perceptual-motor category formation are not driven bypre-coded genetic patterns which predetermine the course of development.Category formation is a dynamical process of self-organization which occurs asa consequence of the individual's time-bound interactions with his or herenvironment. Halliday's (l992a: 21) claim that meaning arises 'out of the impactof the conscious and the material, as mutually contradictory forms ofexperience' supports the claims of Thelen and Smith. The conscious domain i.e. 'in here' - refers to the internal dynamics of the child; the material- i.e. 'outthere' - refers, on the other hand, to the environment of objects, events, actions,and so on, which the child perceives and engages with through his or her activitydependent exploration of it. The child's earliest encounters with the world inthe first four months or so are essentially iconic or topological-continuous incharacter. The essential point is that the multimodal stimulus informationconcerning features of objects and events such as colour, texture, shape, and soon, that infants pick up is necessarily correlated - in and through the child's ownactive, movement-based exploration - with perceived events in the world.

Thus, typical associations are built up on the basis of a regularity in theassociations which is perceived through successive explorations over a period oftime. Edelman's theory of neuronal group selection explains how such regularmultimodal associations of events in the world are reentrantly mapped aspatterns of connections in the brain. The further point to be made here is thateven in this first, iconic-topological phase of proto-meaning-making, there is ahigher-order system of interpretance of matter, energy and information flows


which regulate the transactions between the child's internal dynamics and thestimulus information which is received about events in the world (see chapter I,section 3, pp. 11-18). The fact that perceptual-motor categories are formedthrough the selective strengthening of groups of neurons on the basis of suchmultimodal associations of stimulus information indicates that thirdness - i.e.explicit principles of interpretation and systematicity - is being constructed inthe course of the interactions between the child's internal dynamics and theenvironment.

The observations of Thelen and Smith also show, in my view, the existence ofa very vague, imprecise outline or contour of proto-metafunctional organizationin this earliest perceptual-motor phase of categorization. Thus, we have (l)proto-ideational meanings in the form of the child's multimodal experiences ofthe category-relevant properties of objects and events; (2) proto-interpersonalmeanings in the form of the child's active exploration and engagement with theworld, largely on the basis of 'movement and its perceptual consequences', aswell as through the socio-affective exchanges between child and mother; and (3)proto-textual meanings in the form of patterns of co-ordination between childand his or her surroundings that provide functional action through movement.At this early perceptual-motor stage, the child's relationship to the world is atopological-continuous one. The child's internal dynamics are probably notstrongly differentiated from the external, material world. The one is seen asbeing in a relationship of continuous variation or merging with the other. Thisdoes not mean that there is no distinction; rather, the distinction is not a sharplydefined, or typological-categorial, one.

When the child begins to move into protolanguage around the age of threeto five months, the early protolinguistic signs produced by the child - e.g. thehigh-pitched squeak in Halliday's example (chapter 4, section 10, pp. 202-5) are the child's first conscious attempts to produce signs in order to engageothers interpersonally. This is not to say that interpersonal exchange betweenchild and mother begins with the onset of these protolinguistic signs. As we haveseen, interpersonal dyads characterize the very earliest stage of infant-motherinteractions. The difference lies in the largely iconic, topological-continuouscharacter of the socio-affective exchanges that regulate the earliest stages on thebasis of continuous flows of matter, energy, and information between motherand child. The move into protolanguage is, on the other hand, indexical; theseprotolinguistic indexical signs are contiguous with the contextual values theyindex, in contrast with the continuous nature of the iconic mode whereby theintrinsic dynamics of the child and the flux of ambient stimulus informationabout the external world are only vaguely differentiated from each other ascontinuous variation, rather than categorial (digital) distinction.

Halliday (1975, 1992a, 1993) points out that the simple signs of protolanguage have Just two surfaces, interfaces between the conscious and the twofacets of the material (content purport) and expression purport), such thatmeaning consists in making the transduction between them' (l992a: 22). Thatis, a simple, situation-specific semantics is directly realized by vocal or othergestural means without, however, the mediation of a lexicogrammar. At thisstage, the proto-metafunctions are no longer vague, inchoate possibilities, barely


discernible in the many degrees of topological freedom of the iconic mode.Rather, their future, fully grammaticalized outlines can now be glimpsed in theproto-metafunctions of this indexical phase. In Nigel's high-pitched squeak, forexample, there is (l) proto-ideational meaning, 'attending to phenomenon ofexperience'; (2) proto-interpersonal meaning, 'dialogically engaging mother/demanding explanation'; (3) proto-textual meaning, 'linking internalperception to external event within sharable purview of parent-child dyadthrough use of indexical sign to selectively attend to/point to event'.

The movement from the initial iconic stage to the indexical protolinguisticone arises as a consequence of the child's ongoing, time-bound interactions withthe environment in the course of which higher-order regularities or principlesof interpretation come to be constructed. Thus, we see how the activity of thedyad progressively resets the values of the emerging system as simple, bi-stratalprotolinguistic signs appear in the system. This resetting of values shows how theperceptual-motor activity of the child's environmental explorations in the initialsystem produces positive feedback. The resulting positive feedback, in turn,leads the system far from equilibrium. As a consequence, the newly emergenthigher-order constraints of the indexical mode provide new values so that thesystem is nudged still further along its trajectory. The new constraints ofprotolanguage, again in concert with the child's internal dynamics, mean thatthe attractor pattern in which the child is located is re-structured such that theprobabilities of the system as a whole are realigned.

The move into grammar, as Halliday characterizes the transition to trulysymbolic semiosis, involves the emergence of a tri-stratal semiotic system of thekind first outlined in chapter I, section lO, pp. 46-9. The semantics is nowrealized by a lexicogrammar. The increasing dynamism of the meta-redundancyrelations between strata (see chapter 1, section 6, pp. 26-30) means that thevalues of the system are constantly being re-set from instance to instance(Halliday 1992a: 29). This can only happen because of the essentially open,dynamic character of stratified semiotic systems such as language. Such systemsare involved in constant exchanges with their environments through theactivities of the agents who use the system for their own purposes in particularcontexts. Consequently, the order - the meaning-making potential - of thesystem increases. Semiotic systems are open, far-from-equilibrium systems.Rather than settling into a state of equilibrium, far-from-equilibrium systemsmaintain themselves through constant exchanges of matter, energy, and information between the internal dynamics of the system and its external environments. Biological and social semiotic systems are both examples of this kind ofsystem. Such systems exhibit a number of fundamental characteristics. First, theyhave complexity. That is, the components which comprise the system interact invery many heterogeneous and non-linear ways. Secondly, they are dissipativestructures because they maintain their organizational complexity by obtainingmatter, energy, and information from environmental sources and by exportingor dissipating some of this energy back into their external environments.Dissipative structures are systems that exist far from equilibrium.

Dynamic open systems which both have complexity and are dissipativestructures also show another important property: they are self-organizing


systems. The child's dyadic exchanges with his or her caregivers ensure that hisor her internal dynamics are being supplied with a constant supply of potentiallymeaningful information. The emergence of language and other symbolicmodalities out of the prior proto-semiotic modes is a case in point. The systemas a whole shifts from one global principle of organization to another. Theappearance of lexicogrammar in ontogenesis is a form of emergent organizationwhich has fundamentally different organizational properties from the elementsthat constitute the prior system and cannot be predicted on the basis of the priorstate of the system. In proto-semiosis, the various meanings - proto-ideational,proto-interpersonal, proto-textual - are realized as discrete acts, which Halliday(1975, 1993) has described as 'microfunctions'. On the other hand, theemergence of lexicogrammar entails the slaving of the various functionalcomponents to a global principle of organization, which is metafunctional. Thismetafunctional principle may be considered, following Haken (1977), to be anorder parameter. This means that the system as a whole is now constrained by avery few metafunctional parameters on account of the appearance of a lexicogrammar which is 'slotted in' between the phonetic and semantic levels oforganization of the prior protolinguistic system.

This type of self-organization typifies complex systems which are engaged inconstant transactions with their semiotic and material environments. Such asystem is, as Halliday puts it, a multidimensional and highly elastic semanticspace; it has a very large and constantly evolving meaning-making potential, aswell as a large number of potential cooperative modes both with the physicalmaterial world and with other semiotic modalities (e.g. gesture, movement,depiction, and so on). In generating local points of instability, particular eopatternings of meaning-making selections can function to attract possibilitiesfrom other modalities. Such local asymmetries are amplified the further thesystem is from equilibrium. In this way, self-organizing systems are attracted to apreferred configuration out of many possible ones. When language emergesfrom protolanguage, it does so under the enslaving influence of an orderparameter, viz. a metafunctionally organized lexicogrammar. The system is thusattracted to this preferred mode. In this way, lexicogrammar is an attractor state:the system prefers the principles of organization of its state space and its semiotically salient distinctions (values) over others. Lexicogrammar as state spacerefers to the overall system of values within which the degrees of freedom of thesystem is embedded in specific contexts of use. Thus, the parameters of this statespace are constantly adjusted and redefined according to the contexts whichbring the system into contact with its environment. The emergence of lexicogrammar moves the system into an n-dimensional semiotic space (Halliday1992a: 25) such that the state space of the system is defined as an abstraction ofthe possible values of the terms in n-dimensional space, where n refers to thetotal number of terms - the system of semiotically salient differences, or valuesin the Saussurean parlance - that constitute the system.

The iconic, indexical, and symbolic modes may be seen as stable states alongthe pathway to the emergence of higher-order or symbolic consciousness. In agiven phase state, the system exists in a particular attractor landscape (Thelenand Smith 1994: 56), which is defined by the overall system of interacting


variables that define the system's dynamics and by its initial conditions. In fact,the system is never really static, but is constantly responding and adapting toenvironmental transactions. The shift from one state to another in ontogenesis- iconic to indexical to symbolic - is a qualitative leap to a new phase state suchthat the entire attractor landscape is reorganized. The overall dynamical organization of the system is transformed. The shift from iconic to indexical andsymbolic modalities of semiosis in ontogenesis progressively means that thedynamics of the body-brain complex are entrained into the global dynamicalorganization of higher-scalar ecosocial relations and processes. In this way, theneural activity of individuals is entrained to the meaning-making practices andthe semiotically salient differences of the ecosocial level of organization.Symbolic consciousness emerges out of and constrains, without ever totally transcending, the less specified perceptual-motor and conceptual forms of categorization that are characteristic of primary consciousness (Edelman 1992: chapter11). The iconic and indexical modes are always nested within and integrated bythe symbolic mode in terms of the logic of the specification hierarchy.

It is important to draw attention here to the fundamentally multimodalcharacter of all human meaning-making. The multimodal character of symbolicmodalities of semiosis is built upon and does not entirely transcend themultimodal character of the infant's earliest multimodal sensori-motor explorations of his or her environment in the process of perceptual-motor categoryformation. Moreover, the longer phylogenetic history of pre-linguistic ore-facialand manual-brachial gestures with respect to language does not mean thatlanguage transcends or otherwise stands independent of these phylogeneticallyprior forms. Rather, the historical emergence of language led to its multimodalintegration with other semiotic modalities. Rather than impeding the fullflowering of language as the ultimate expression of human rationality, thisprocess of integration has afforded the possibility of the evolution of new genresand ways of making meaning including the multimedia and hypertext genrescharacteristic of the age of the Internet.

In the next section, Gibson's ecological theory of perceptual awareness willbe examined in order to see what it can contribute to the semiotic theory ofconsciousness that I am developing here.

4. The Structure of (Self)Consciousness in Perceptual Awareness

Gibson (1986 [1979]: chapter 7) shows that the information specifying the self(proprioception) and the information specifying the perceived environment(exteroception) are not two different kinds of information. Instead, they are twopoles of attention based on the same kind of information. The perceiving of anenvironmental event means perceiving that event as occurring 'there', at acertain distance from the perceiver, and so on. Thus, Gibson shows that theperceptual awareness of objects and events in one's environment is also andalways self-awareness of one's self as the one who perceives those objects andevents in the perspective of the self (see also Battacchi 1999: 57). Reference toperceived objects and self-reference are no more than two poles of attention of


a single overarching operation which necessarily always implicates both poles.Gibson's observations show that the act of perception is always grounded in theperspective of the self qua perceiver. Thus, we see how the principle ofgrounding, which is of crucial importance in the operation of language, alsooperates at the lower integrative levels of perceptual awareness and pre-linguisticconceptual thinking. Individuals are likewise aware of the fact that theperceptual and mental states which they experience are experienced as theirown. That is, perceptual and mental states are grounded in the perspective ofthe self who has this experience. Battacchi (1998: 6) argues on this basis thatconsciousness necessarily entails not only an awareness of one's own mentalstates, but also a concomitant awareness that the mental states are one's own. Forthis reason, Battacchi proposes the term' (self) consciousness' in order to drawattention to the fact that 'consciousness is always self-consciousness' (1998: 6; seealso Battacchi 1996).

Gibson's discussion of the two poles of attention implicated in perception is auseful starting point for exploring the question of how others can access theself's perceptual awareness and states of consciousness. Let us start withperception. The first question I shall consider is how and to what extent anindividual act of perception, grounded at a particular self, can be re-groundedby another individual who is observing the first individual's act of perception andtrying to figure out what the first individual is perceiving. This happens in thecase of gaze-following by primates. Bogdan (2000: 110-13; see also chapter 4,section 9, pp. 201-2, for further discussion) discusses how primates recognizethe vectorial quality of another's gaze so that they are able to track the directionof the other's gaze. Furthermore, this involves an act of interpretation on thepart of the observer such that the observer is able to follow the gaze vector'sdirectionality in order to determine the target of the observed individual's gaze(Bogdan 2000: Ill). From the point of view of the individual who performs theact of gazing, we can say that this individual is simultaneously aware of the targetof his or her gaze qua object of attention, along with the fact the individual'sperceptual awareness of this target is grounded in the perspective of theparticular self who has the experience. This grounding of the perceptualawareness in the perspective of a self is both iconic and indexical in characterand is, therefore, tied to the individual's material interactivity with theimmediate environment in which the act of gazing occurs. Moreover, the self isaware not only that it is the selfwho has the given experience, but also the natureof the specific perceptual modality - visual perception, in this case - in which theperceived object is experienced. The two poles of this relationship may beschematized as follows:

[[PERCEIVING SELF --> [PERCEIVED OBJECT]]

In my view, the two poles of this experience are precisely what enables theexperience to be reconstructed from the perspective of an observer of the self'sactions. That is, the observer is able to re-ground the experience from his or herperspective by virtue of the contextualizing relations which the observerconstrues (interprets) between the observed selfi's gaze vector and the gazed-at-


object. In turn, the contextual relationship between these two factors is recontextualized on a higher level by the observer, i.e. from the perspective of theobserver qua selfs. This re-contextualization shows that even at the lower integrative level of perceptual awareness, both humans and non-human primates havethe ability to re-ground the perceptual experience of another and to interpret itfrom the perspective of their own self. This does not mean that the observer, forexample, is able to stand in the shoes, so to speak, of the observed self and havethe same experience. The distinction between first- and third-person accounts ofconsciousness (chapter 4, section 1, pp. 171-3) is something of a red herring.Instead, the observer's ability to re-ground another's perceptual experienceshows that even at this level, there exists the possibility of interpreting another'sexperience from one's own perspective in the way described above. In any case,both the perceptual experience of the observed self and its re-grounding by anobserver are tied to a particular perceptual purview on the immediate here-nowscale of the body in interaction with its immediate environment.

For example, the observer (self,) in our example must be able to observe and,therefore, to visually track the observed self.'s gaze vector to its target. What isimportant is that self and observer belong to the same higher-scalar system ofinterpretance. It is the integration of self and observer to this SI and itsconsequent mediation of the information deriving from them which determineswhether this information is meaningful to other selves who share the same SIand its perspectives. Thus we see that the lower integrative level of perceptualexperience and the forms of consciousness that derive from it are not randomlydifferent from one individual to another. The perception of environmentalevents from the grounded perspective of a self and the interpretation of theself's perception from the standpoint of an observer are possible because bothenvironment and the individual conspecifics who act and perceive in thatenvironment are integrated to and participate in higher-scalar relationshipswhich define and regulate possible individual-environment transactions. Thismeans that (1) the individual's sensori-motor interactions with and responses tothe environment, in spite of individual differences and variations from oneindividual to another, reflect patterns of interaction that are typical of all thoseindividuals who participate in the same environment; and (2) the perspectives ofself and observer in our example can be co-ordinated in the way describedbecause each individual's sensori-motor patterns of interaction share fundamental characteristics with those conspecifics who have been entrained to thesame kinds of self-environment transactions.

As we shall see in section 5, Damasio's neurobiological account of consciousness is based on self-object transactions. Damasio's account also begins to revealthe proto-metafunctional character of consciousness. In sections 5 and 6, I shallexamine Damasio's theory with this perspective in mind.

5. Damasio's Neurobiological Theory of Consciousness

Damasio proposes a neurobiological theory of consciousness which is based on'two intimately related problems' (1999: 9). Damasio explains these two aspectsof consciousness as follows:


The first is the problem of understanding how the brain inside the humanorganism engenders the mental patterns we call, for lack of a better term, theimages of an object. By object I mean entities as diverse as a person, a place, amelody, a toothache, a state of bliss; by image I mean a mental pattern in anyof the sensory modalities, e.g. a sound image, a tactile image, the image of astate of being. Such images convey aspects of the physical characteristics ofthe object and they may also convey the reaction of like or dislike one mayhave for an object, the plans one may formulate for it, or the web of relationships of that object among other objects. (Damasio 1999: 9)

The second problem is defined as follows:

This is the problem of how, in parallel with engendering mental patterns foran object, the brain also engenders a sense of self in the act of knowing. Tohelp me clarify what I mean by self and knowing, I urge you to check theirpresence in your own mind right now.

You are looking at this page, reading the text and constructing themeaning of my words as you go along. But concern with text and meaninghardly describes all that goes on in your mind. In parallel with representingthe printed words and displaying the conceptual knowledge required tounderstand what I wrote, your mind also displays something else, somethingsufficient to indicate, moment by moment, that you rather than anyone elseare doing the reading and the understanding of the text. The sensory imagesof what you perceive externally, and the related images you recall, occupymost of the scope of your mind, but not all of it. Besides those images thereis also this other presence that signifies you, as observer of the things imaged,owner of the things imaged, potential actor on the things imaged. There is apresence of you in a particular relationship with some object. (Damasio1999: 9-10)

Damasio's definition illustrates the principle mentioned earlier that consciousness is an evolved way of mediating the experience of experiencing. Consciousness, Damasio observes, 'from its basic levels to its most complex, is the unifiedmental pattern that brings together the object and the self' (1999: 11). Thebringing together of self and object means that the two interact in some way; selfmay be generalized as an observer system which interacts with some other system- the object - and therefore forms some kind of interactive relationship with it.The mental images which the self forms of the object constitute meaningfulinformation internal to the self. These images constitute a selective mapping ofinformation in the environments - both external and bodily - onto the observersystem's internal structure in order that the system can impose order and patternon environmental fluctuations. The interactive relation between self and objectmay be seen in terms of what Peirce called 'firstness' and 'secondness' (seechapter 1, section 8, pp. 34-9). While self in Damasio's account would be a first,the interaction between self and its object is a second. Furthermore, it is clearthat the interaction between self and object is not random, but takes place in anorganized - Damasios term is 'unified' - field of mental images (of the object)


which mediate between self and object and which enable the two to be placed insome kind of interactive relationship with one another. Again, we see that wehave self-awareness not of the self per se, but of the self-in-interaction-withsurround (i.e. the internal milieu and the external environment). That is, weloop on the whole Batesonian loop, including both efferent and afferentelements, such that the 'out there' that we construe as being different from the'in here' is re cursively mapped in the brain as an internalization of self-objecttransactions.

The system of relations described by Damasio may be seen in terms of thethree-level scalar hierarchy (chapter I, section 3, pp. 11-18). It is important topoint out here that Damasio's description of the self as 'this other presence thatsignifies you, as observer of the things imaged, ... ' does not refer to the brainper se, but to a self-referential perspective which is neurologically realized inspecific areas of the brain (Damasio 1999: 25). In terms of the three-level scalarhierarchy, interactions between self and object constitute the focal level L. Belowthis level, there is the specific neurobiological organization of the individualorganism (not the self) which predisposes the organism to certain kinds of interaction with objects in both its inner (bodily) and outer environments and toform mental images of these. This is the L-Ilevel. Above the focal level, there isa system of interpretance, which is comprised of the perceptual-motor interactions between the organism and its inner and outer environments and whichforms the basis on which mental images of environmental objects are interpreted and understood. This is the L+I level. The vaguely semiotic appeal,whether intended or not, of Damasio's designation of the self as 'this otherpresence that signifies you' (my italics) suggests that the self is realized bylower-level neurological events, but is not reducible to these.

Overall, Damasios description of the basic components of what he calls coreconsciousness shows how levels L-I and L are contextually integrated by thehigher-scalar level, L+I, at the same time as being self-reflexively connected tothe higher-scalar level which provides a theory of the lower levels (Salthe 1993:49). It is the higher level which interprets - contextually integrates - the lowerlevels. This last point is illustrated by Damasio's discussion of the importance ofmental images in the guiding of action (1999: 24). That is, mental imagesformed in core consciousness are a means for guiding the self's interactions withthe object. Damasio puts it like this:

Images allow us to choose among repertoires of previously available patternsof action and optimize the delivery of the chosen action - we can, more or lessdeliberately, more or less automatically, review mentally the images whichrepresent different options of action, different scenarios, different outcomesof action. We can pick and choose the most appropriate and reject the badones. Images also allow us to invent new actions to be applied to novelsituations and to construct plans for future actions - the ability to transformand combine images of actions and scenarios is the wellspring of creativity.(Damasio 1999: 24)


and

Consciousness generates the knowledge that images exist within theindividual who forms them, it places images in the organism's perspective byreferring those images to an integrated representation of the organism, and,in so doing, allows the manipulation of the images to the organism'sadvantage. Consciousness, when it appears in evolution, announces the dawnof individual forethought. (Damasio 1999: 24-25)

The mental images formed in the brain on the basis of the organism's interactions with its internal and external environments contextually integrateorganism and object such that the organism may selectively represent the givenobject and act in relation to it. Importantly, these image-object relations aresymbolic; they are not indexically connected to particular external environmental objects and events. Instead, the mental image symbolically interprets itsobject. For this reason, the organism is able to exercise choices in action andresponse from among a repertoire of symbolic possibilities which are internalized in its Innenwelt. Furthermore, these organism-object transactions are, as Ihave argued, integrated into a higher-order system of interpretance. In onesense, all of the basic components of consciousness occur within the brain of theindividual. However, I have shown that the logic of the three-level scalarhierarchy applies here as well. In other words, there can be no reduction toneural events per se. Core consciousness is, in the first instance, a form of protosemiosis whereby the higher-scalar level of the organism-environment transactions - i.e. the system of interpretance - is internalized as the organism'shere-now Innenwelt in the form of mental images which are grounded in theperspective of the self who observes and experiences these images. That is,models of organism-environment transactions constitute a system of interpretance in the Innenwelt such that particular interactions between self and theobjects of (self)consciousness can be interpreted. The fact that mental imagesallow the organism to choose among different potential actions implies thatmental images are potentially meaningful in some way to the organism. Thisfurther implies that a higher-scalar system of interpretance allows the organismto know and act on the here-now scale by formulating contextually appropriateinterpretants.

Mental images, Damasio claims, are created within the brain in order that theorganism may interact with objects 'actually present or recalled' (1999: 25). Thissuggests that the mental images of core consciousness, rather than simply representing something which is already pre-given, may be said to indexically invokeor otherwise symbolically constrnct their objects. Thus, mental image a isredundant with object b in the sense that they are both contexts for each other'sinterpretation (see Lemke 1984c: 36 for this formalism; see also chapter 1,section 6, pp. 26-30). Damasio had remarked (see above) that images convey'aspects' of their objects. That is, the reciprocal process of contextualization is aselective one. However, in terms of the logic of the redundancy relations, therelation between a and b itself minimally requires a further order of contextualizing relations, c, in order to be integrated to or interpreted by some higher-


scalar system ofinterpretance. That is, for a mental image to be recognized as animage of an object, there must be a further level of contextualization whichspecifies how the a/b relation is to be interpreted or recognized by the self(Lemke 1984c: 36). This would be suggested by the context-sensitive nature ofneurological processes (see chapter 6, section 4, pp. 290-5). In the here-nowenvironment of core consciousness, objects in the organism's external andinternal environments form redundancy relations with neurological events suchthat mental events are realized. These mental events are signs of environmentalobjects and events such that an image is grounded within the perspective of theself. Consciousness, in this case, would be the higher-order context in which thea/b relationship between environmental events and neurological events is interpreted as a mental image which has meaning in the perspective of the self.

Prior to the conscious core self of Damasio's account, there is the nonconscious proto-self (Damasio 1999: 173). Damasio outlines the distinctionbetween unconscious proto-self and conscious core self as follows:

As far as the brain is concerned, the organism in the hypothesis is representedby the proto-self. The key aspects of the organism addressed in the accountare those I indicated as provided in the proto-self: the state of the internalmilieu, viscera, vestibular system, and musculoskeletal system. The accountdescribes the relationship between the changing proto-self and the sensorimotor maps of the object that causes those changes. In short: As the brainforms images of an object - such as a face, a melody, a toothache, the memoryof an event - and as the images of the object affect the state of the organism,yet another level of brain structure creates a swift nonverbal account of theevents that are taking place in the varied brain regions activated as a consequence of the object-organism interaction. The mapping of the object-relatedconsequences occurs in first-order neural maps representing proto-self andobject; the account of the causal relationship between object and organism canonly be captured in second-order neural maps. Looking back, with the licenseof metaphor, one might say that the swift, second-order nonverbal accountnarrates a story: that of the organism caught in the act of representing its ownchanging state as it goes about representing something else. But the astonishing factis that the knowable entity of the catcher has just been created in the narrativeof the catching process. (Damasio 1999: 170; italics in original)

The relationship between proto-self and the sensori-motor maps of the objectsthat cause changes in the proto-self occurs in first-order neural maps. At thislevel, there is a first, weak differentiation of proto-self and object, but noawareness of this differentiation, as shown by the fact that the proto-self changesin response to environmental fluctuations that cause these changes. Damasiodefines the proto-self as 'a coherent collectionof neural patterns which map, moment bymoment, the state of the physical structure of the organism in its many dimensions' (1999:154; italics in original). That is, the proto-self experiences, but has no secondorder experience of its experiencing. Furthermore, the proto-self 'has no powersof perception and holds no knowledge' (Damasio 1999: 154). The emergence ofcore consciousness and its corollary of a conscious core self is based on second-


order neural maps which realize an 'account of the causal relationship betweenobject and organism' (1999: 170). This is where the experience of experiencingenters the picture. Damasio explains core self and core consciousness as follows:

You know that you are conscious, you feel that you are in the act of knowing,because the subtle imaged account that is now flowing in the stream of yourorganism's thoughts exhibits the knowledge that your proto-self has beenchanged by an object that has just become salient in the mind. You know youexist because the narrative exhibits you as protagonist in the act of knowing.You rise above the sea level of knowing, transiently but incessantly, as a feltcore self, renewed again and again, thanks to anything that comes fromoutside the brain into its sensory machinery or anything that comes from thebrain's memory store toward sensory, motor, or autonomic recall. You know itis you seeing because the story depicts a character - you - doing theseeing. The first basis for the conscious you is a feeling which arises in the representation of the nonconscious proto-seifin theprocess of beingmodifiedwithin anaccount which establishes the cause of the modification. The first trick behindconsciousness is the creation of this account, and its first result is the feelingof knowing. (Damasio 1999: 171-2)

The second-order neural maps of core consciousness bring in a further differentiation of the first-order reentrant loop linking proto-self to object. Thus, coreconsciousness entails a conscious experience of experiencing, as realized by thesecond-order neural maps mentioned by Damasio in the above citation. Coreconsciousness is transient; it is linked to the here-now and is non-verbal. Its timescale is fractions of seconds. Beyond the core self, there is what Damasio refersto as the autobiographical self (1999: 172-3). The autobiographical self bringshistory into the overall picture. It is built on core consciousness, but is distinguished from this by the fact that the transient experiences of the self of coreconsciousness can now be committed to memory and at the same time conceptualized (Darnasio 1999: 173). The autobiographical self has a past, as well as ananticipated future. Nevertheless, it too consists of nonverbal images. Damasionotes that this memory can be expanded and refashioned along one's lifetrajectory (1999: 173). The formation of an autobiographical self means thatfurther differentiations of the self and its relationships to objects - past, present,and future - can be made. The autobiographical self is based on a concept 'inthe form of dispositional, implicit memories contained in certain interconnected brain networks' (Damasio 1999: 174). The making explicit of thesememories at any given moment means that we move from the experience ofexperiencing (core consciousness) to the memory-based conceptualization ofexperience. This is what Damasio calls 'extended consciousness' (1999:195-200).

Darnasio's distinction between the unconscious proto-self and the consciouscore self and the autobiographical self demonstrates the logic of the specification hierarchy - both synchronically and developmentally. Thus, there is aprogression from the most vague and general qualities of the proto-self, whichhas neither perception nor knowledge, to the transient here-now knowledge that


the core self has of its mental images through to the more conceptually elaborateand memory-based images of the autobiographical self of extended consciousness. In this specification hierarchy, it is clear that the autobiographical self isboth more elaborate and more determinate than the less specified proto- andcore selves as a consequence of its memory-based conceptual ability to storeinformation as well as to individuate this information. This specification hierarchyof consciousness begins with the absence of observer self-consciousness (protoself). This is the least specified level. For this reason, it is also the level that isfurthest from our experience.

The increasingly self-referential perspectives of core consciousness and thenautobiographical consciousness, in Damasio's account, are explainable on thebasis of the increasing specification of these levels, which are closer to the perspectives familiar to the observer. As we shall see below, the metafunctionalorganization of language can be see as a more specified mode of organization ofhigher-order or symbolic consciousness in the form of, for example, conscious,linguistically mediated thinking (sections 9 'and 10, pp. 271-5). However, theforms of consciousness which occur on lower integrative levels can also be madeintelligible in terms of the metafunctional principle. This is so for two reasons.First, the principles of organization imposed by more specified levels such assymbolic (higher-order) consciousness impose constraints on lower (lessspecified) integrative levels. Secondly, statements about the organization ofhigher-order consciousness can only be made, at least implicitly, because theforms of consciousness on lower levels came into existence before the higherlevels. This is so both phylogenetically and ontogenetically. This suggests that lessspecified principles of organization on the lower levels are integrated to andreorganized on the higher levels according to more specified principles.

Importantly, selves can be conscious of mental images which are nonverbal incharacter. These images are grounded in the perspective of an intentional selfwho attends to these images in consciousness. The unconscious proto-self isbased on first-order neural patterns whereas the core self and the autobiographical self have knowledge of nonverbal second-order mental images. In coreconsciousness, these images occur in the here-now, whereas in extended consciousness they can be connected to a past and a future. The imagistic characterof both core and extended consciousness corresponds, in my view, to what Peng(1994) has called proto-meaning. Mental images are proto-semiotic, althoughthey already begin to exhibit some of the characteristics of semiosis. I believe itis right to see these mental images as proto-meanings because they are contentswhich are realized by neural events in the brain. One difference with a fullfledged semiotic system like language is that the same mental image can berealized by different patterns of neural activity within the individual on differentoccasions. Mental images, as well as higher-order nonverbal thoughts, arede-coupled from neural processes such that the content which is realized bythe neural processes becomes the object of conscious awareness and attentionrather than the neural processes themselves (Juarrero 1999: 90). While there isno conventional coupling of meaningful mental content to their means ofexpression in neural processes, there is, nevertheless, a principle of realizationwhich stratifies mental contents and their neural means of expression. Thus, the


flow of non-perceptual sensations which are felt by the proto-self may be seen asa form of pure iconicity; the transient here-now character of the perceptuallycaused images of core consciousness is akin to indexically presupposing signs;the ability to recall past images, to anticipate the future, and to elaborate mentalimages in extended consciousness suggests something like Silverstein's (1987a,1987b) notion of indexically creative or entailing signs. The latter two categoriesrequire a self-referential perspective, which would seem to be required by anyform of proto-indexical mental imagery. This is so in the sense that indexicalityis always related to and grounded at an intentional source which points to(indexes) the indexed object.

The three components of the specification hierarchy mentioned above alsoshow how proto-self, core self and autobiographical self are all embedded in animmediate supersystem which contains both (proto-)self and the object in theimmediate (internal or external) environment of (proto-)self. This shows howthe emergence of consciousness from non-eonsciousness is always guided byhigher-order supersystem transactions. Accounts of consciousness whichprivilege a bottom-up view of consciousness in terms of the organism's neurobiological dynamics may miss the point that consciousness, by definition, alwaysimplies higher-order supersystem processes and dynamics (see Salthe 1993:277-81; Lemke 1995b: 115-16). There is always a stable, higher-order regulatoryenvironment on a slower timescale which regulates the faster-scale neuraldynamics within the individual organism. Consciousness in the individualorganism and the neurobiological processes which subtend this are alwayscoupled to larger-scale self-object relations and dynamics and in ways whichintentionally direct and guide these dynamics. In the case of consciousness, selfobject transactions, as described by Damasio, are the immediate supersystem.The social character of our evolutionary ancestors, as well as the human speciesitself, means that many different individuals develop under similar environmental conditions, thereby increasing the likelihood of their undergoing similarbifurcations (Lemke 1995b: 115). The emergence of consciousness in anindividual has environmental effects that promote the same kind of bifurcationin other individuals. Organism-environment transactions might lead to theformation of neural patterns which reentrantly map these supersystem interactions to an internalized perspective. In this way, a relatively stable, higher-ordersupersystem of self-object interactions gets internalized in the organism'sInnenwelt. This would be formally equivalent to the idea of mind as an individuated system of interpretance which the self uses to attend to, interpret, andrespond to events (signs) in the self's internal and external environments.

But conscious selves are always cross-coupled to other conscious selves in stillhigher-scalar networks of both physical-material and semiotic-discursiveprocesses and relations on the ecosocial scale. According to Damasio (1999:185), language converts or translates the second-order nonverbal narrative ofimages of consciousness (core or extended) such that these 'become verballypresent in our minds' in the process of focusing on them (ibid.). I agree fullywith Damasio (e.g. 1999: 108) that self and consciousness exist prior to languageand that language does not construct consciousness as such. Damasio's coreand extended consciousness pre-exist language, as does Edelmari's primary


consciousness. The neurobiological arguments of Damasio and Edelman arecompelling in this respect. However, the suggestion that language merely has atranslating function with respect to the mental images of consciousness is, in myview, a seriously misleading notion of the function of language. I shall now turnto this question. I suggested above that consciousness is an internalization - areentrant looping - of self-object supersystem transactions. This means thatconsciousness is always the result of higher-order supersystem transactions.Consciousness came to be in the context of such higher-order environmentswhich got internally mapped as the organism's Innenwelt. It did not come intobeing on the basis of bottom-up neurological dynamics perse. Core consciousnessand extended consciousness are founded on biological and thermodynamicprocesses and transactions between the organism and its material environments.The space-time scale is the immediate here-now one. That is, individual coreconsciousness is tightly coupled to here-now material transactions; on the otherhand, extended consciousness, with its past and future, exhibits the capacity toincorporate diverse space-time scales to a much greater extent.

However, the cross-eoupling of consciousness to ecosocial semiotic processesmeans that the individual's consciousness is now linked to systems of verydifferent space-time scales (see Lemke 2000a, 2000b; Thibault 2000a). In thecase of language and other social semiotic resources used to make meaning in agiven culture, these form part of a larger-scale ecosocial system which integratesand regulates the smaller-scale subsystems of individuals to its larger-scalepatterns and dynamics. The cross-coupling of individuals to these higher-scalarecosocial relations and dynamics means that individual consciousness, inparticular what Edelman calls higher-order or symbolic consciousness, isentrained to the social practices in and through which language is deployed tomake meanings in specific contexts. Symbolic consciousness has all the characteristics of core consciousness and extended consciousness, but it is also a moresemiotically specified form of consciousness which enables us to give symbolicmeaning to experience by organizing and entraining consciousness in relation tothe always to-degrees shared social meaning-making practices of a given culture.Furthermore, the connecting of consciousness to a culture's symbolic resourcesfor making meaning means that, in symbolic consciousness, maximally diversespace-time scales may intersect in the individual (Lemke 2000a). Rather than'translating' nonverbal mental images in consciousness, language, in the processof construing these in terms of its own intrinsic categorial differentiations, crosscouples individual consciousness to the social semiotic formations of acommunity. On the other hand, to say that language 'translates' nonverbalimages in consciousness is, implicitly, to assume that language in the individualorganism is the focal level. It is to say that we make sense of language in relationto mental images arising 'from below' in the brain of the individual and whichlanguage encodes ready for decoding by someone else.

But consciousness is a relationship to phenomena on the human-scale that weshare with others. It is a question of the perspectives that the self can adopt onthe phenomena of experience and how these experiences may be meaningfullyexchanged with other selves who are able to share the same perspectives. Thisalso helps to resolve the dilemma of the third-person and first-person


approaches to consciousness. Proponents of the first approach seek to reduceconsciousness to neural events per se (chapter 4, sections 1-3, pp. 171-84). Thisapproach is reductionist precisely because it does not account for the humanscale observer perspectives of selves. It fails to see that the models of neuralevents that are made in this third-person perspective are in fact models of theobserver's relationship to the events on this level. On the other hand, firstperson perspectives, in emphasizing the uniqueness of individual consciousness,can border on solipsism in failing to account for the fact that consciousness quathe observational perspective of the self occurs on the human-scale level of otherselves which are embedded in a higher-scalar system of interpretance which contextually integrates the perspectives of different selves to its dynamics. In thisway, the perspectives of diverse selves can be dialogically co-ordinated andnegotiated in discourse. This does not mean that individuality is eliminated.Individual consciousness is robust in its system of interpretance precisely becauseit leaves signs of its individuality in its interactions with the viewpoints of otherselves who are embedded within the same system of interpretance and in termsof which individual viewpoints have their meaning (see Salthe 1993: 51 for thenotion of robustness; see also this volume, chapter 2, section 1, pp. 59-68).

6. The Proto-metafunctional Structure of Core and ExtendedConsciousness

Damasio's account of core consciousness and extended consciousness suggestssome striking parallels with the internal metafunctional organization oflanguage, as in Halliday's (1979) account. For reasons I shall discuss below, I donot think this is in any way accidental. But first I shall re-examine the principalcomponents of Damasio's theory of core and extended consciousness with a viewto seeing how they can be seen as precursors of the intrinsic metafunctionalorganization of language in higher-order or symbolic consciousness in the waythat I suggested above.

6.1 Conceptualizing the Objed ofConsciousness: Mental Images and Proto-experientialMeaningMental images are of some object either in the external environment of theindividual or within the internal milieu of the individual organism's body.Mental images are neurally constructed images of objects that are constituted asphenomena of experience for the self. Damasio points out that the brainconstructs mental images of objects in both the external environment of theindividual organism as well as of objects experienced as being within the internalbiological milieu of the organism's body. This distinction shows how the very firstlevel of consciousness - Le. core consciousness - already constitutes a distinctionbetween that which is experienced as existing 'out there' in the external worldas distinct from that which is experienced as existing 'in here' in the internalbiological milieu of the individual organism. Furthermore, the mental images ofthe objects of experience are not simply re-presentations of already givenphenomena in the inner and outer worlds. Rather, neural processes in the brain


actively and selectively construct mental images of the phenomena of experienceby attending to specific aspects of these phenomena, the emotional state of theindividual, and so on. To paraphrase Edelman, there is no pre-labelled worldwhich mental images simply reflect.

Mental images are the experiential contents of consciousness and are realizedby patterns of neural activity though without being reducible to these. Furthermore, mental images are multimodal in character: they can be formed in diversesensory-perceptual modalities such as touch, seeing, hearing, inner bodyfeelings, and so on (Damasio 1999: 9). Mental images are the ways in which consciousness constructs the world as phenomena which are experienced by the self.They may be described as the experiential contents of consciousness. The selfdoes not directly experience the object. Instead, its experience is mediated bythe mental images which the brain constructs of the objects of experience, alongwith the conceptual categorizations of these.

6.2 The Interactive Relation ofSelf to the Object: Proto-interpersonal MeaningWithout repeating the arguments discussed earlier, a second crucial aspect ofDamasio's account consists of the self's interactive relationship with the object ofconsciousness. As I shall show here, this aspect of the self-object relationship maybe seen as proto-interpersonal in character. In this perspective, self and object areboth differentiated from each other at the same time as being placed in an interactive relationship with each other (see section 3, pp. 241-6). The phenomenological experience of core consciousness is an awareness (reentrant loop) of ourdoing/being-in-the-world as an undivided, pre-polarized whole. We have selfawareness not of the self per se, but of the self-in-interaction-with-nonself (e.g.other people and objects). Furthermore, this relationship may be modulated bydifferent embodied feelings which the self has towards the object; it may also haveimplications for the specific courses of action which the self decides to undertakein relation to the object. In core consciousness, the distinction between selfand object corresponds to a different kind of differentiation with respect to theexperiential kind discussed in the previous section. In the present case, it isthe interactional distinction between the self and the 'object' (the nonself, theOther, a second) with which the self interacts and to which the self orients. Thismay be explained in relation to mental images as follows.

A mental image is a neural construal of some object or event either in theworld outside the organism's body or within the organism's internal milieu (seeabove). With respect to this image, the self stands in a particular perspectival relationship to it. The self is not a part of the whole which is construed by the mentalimage. Rather, it is distinct from the mental image; it stands in some kind of interactive relationship with it. What is important here is the relationship which theself has with the mental image or the object which the image constructs. From thispoint of view, the self can orient to the object in terms of the course of action theself intends to take with respect to the object. The self can also experiencesubjective feelings and bodily sensations in relation to the object. Furthermore,the self can acquire knowledge of the object and, therefore, incorporate theobject to a framework of beliefs and expectations which are stored in memory foractivation on other occasions in extended consciousness.


Core consciousness yields an experiencing of self-in-world where there is notyet a typological-categorial separation of self from world. Instead, there is a firstsoft polarization (weak classification) whereby self and nonself are not sharply ortotally distinguished from each other. Instead, there is a relationship betweenthe two which is based on topological-continuous variation; self and nonself arecontinuous with each other. There is topological variation without typologicaldistinction. This provides a first, most general basis for the subsequentemergence of the many more differentiated and specified relationships withincreasingly more separate others on higher integrative levels, particularly thesymbolic level. The interactive relationship between self and object in core consciousness can be seen as a less specified interactional-affective relational holismfrom which the semiotic resources of linguistically mediated interpersonalinteractional meaning ultimately derive. The interpersonal semiotic resources oflanguage symbolically constitute further typological differentiations between,say, T, 'you', and 'she', 'it', and so on, though in ways which do not totallytranscend their basis in more primitive, more holistic, more iconic forms ofself-object relations in core consciousness. Damasio (1999: 36-7) points out thatconsciousness is rooted in and modulated by inwardly directed feeling states ofthe body which precede emotion and may not always be conscious. These feelingstates of the body can, however, be displayed as socially recognizable emotions.This observation can help us to forge a link between the interpersonalinteractional semiotic resources and the feeling states of the experiencing bodyin semiosis. Damasio (1999: 79) argues that emotions are represented by asystem of neural dispositions located in a number of brain regions, notably the'subcortical nuclei of the brain stem, hypothalamus, basal forebrain, andamygdala'. The activation of these emotional dispositions, Damasio continues,leads to a number of consequences.

Following Damasio closely, these may be summarized as follows. First, a givenpattern of neural activation yields a neural representation of a particularemotion which can then become an object of consciousness awareness.Secondly, this same pattern of neural activation induces changes in body state aswell as in the state of other brain regions. It is this which creates an emotionalstate which is palpable to an observer, who is able to interpret the individual asbeing emotionally engaged with some object. Thirdly, the internal state of theorganism is altered inasmuch as it now has an emotion as a neural object ofconsciousness and at the same time senses a change in the body's feeling-stateprovided that the neural pattern of activation crosses the threshold of consciousness. Damasio further argues that the neural patterns which form thesubstrate of feeling-states of the body bring about 'two classes of biologicalchanges' (1999: 79, viz. 'changes related to body state and changes related tocognitive state' (1999: 79). Damasio explains changes related to body state asfollows:

The changes related to body state are achieved by one of two mechanisms.One involves what I call the 'body loop'. It uses both humoral signals(chemical messages conveyed via the bloodstream) and neural signals (electrochemical messages conveyed via nerve pathways). As a result of both types


of signal, the body landscape is changed and is subsequently represented insomatosensory structures of the central nervous system, from the brain stemup. The change in the representations of the body landscape can be partlyachieved by another mechanism, which I call the 'as if body loop'. In thisalternate mechanism, the representation of body-related is created directly insensory body maps, under the control of other neural sites, for instance, theprefrontal cortices. It is 'as if' the body had really been changed but it was not.(Damasio 1999: 79-80)

Changes related to cognitive state are summarized by Damasio as follows:

... (1) the induction of specific behaviors such as those aimed at generatingbonding, nurturing, exploration, and playing; (2) a change in the ongoingprocessing of body states such that the body signals filtered or allowed to pass,be selectively inhibited or enhanced, and their pleasant or unpleasant qualitymodified; and (3) a change in the mode of cognitive processing such that, forexample, the rate of production of auditory or visual images can be changed(from slow to fast or vice versa) or the focus of visual images can be changed(from sharply focused to vaguely focused); changes in rate of production orfocus are an integral part of emotions as disparate as those of sadness orelation. (Damasio 1999: 80)

It is not difficult to see in the above proposals by Damasio that the interactionalaffective holism that I mentioned above constitutes the first level in a specification hierarchy whereby embodied feeling-states and affective orientations, alongwith their corresponding cognitive states, are integrated in the brain such thatthe neural images which form the objects of consciousness are simultaneouslyconstrued as having both experiential-cognitive content and interactionalaffective values and orientations. The patterns of neural activity associated withthese dimensions are tightly cross-coupled as the self interactively relates tospecific objects of consciousness and in ways which modulate this relationship.The three factors which Damasio singles out in his discussion of changes incognitive state (see above citation) suggest that the self can (1) select from arepertoire of possible actions for relating to an object of consciousness; (2)modulate its affective orientation and feeling-state towards the object; and (3)modulate the object itself so as to alter the ways in which the self emotionallyorients to the object. Thus, self, object or the interactional relations betweenthem can all be modulated and re-shaped in ways which channel and direct theworkings of consciousness and the action trajectories which flow into theenvironment as a consequence of this.

6.3 The Grounding ofSelf in its Internal Milieu: Proto-textual MeaningThe most fundamental kind of grounding concerns the distinction betweenwhat is 'in' the structure of the organism and what is 'out' in its environment.The biological maintenance of the self is grounded in the relative structuralinvariance - the relative stability - of the body over time so that this invariance'can dispense continuity of reference across long periods of time' (Damasio


1999: 135) to the self. At the level of core consciousness, the self may be said tobe grounded in the relative stability afforded by the body, which functions as theimmediate environment of the self's time-bound trajectory. Damasio points outthat representations of the body 'come from the remarkable invariance of thestructures and operations of the body' (1999: 141). Body representations signifythis stability for the neurally realized self in the brain throughout the life of theorganism. It is the somata-sensory system and somatic signalling which functionsto send information to the brain about the state of the various parts of the body.Some of this signalling occurs via nerve fibres that convey signals from the bodyto the central nervous system; other forms of somato-sensory signalling usechemical substances carried by the bloodstream (Damasio 1999: 149). Damasiogroups somato-sensory signalling into 'three fundamental divisions', viz. theinternal milieu and visceral division; the vestibular and musculo-skeletardivision; and the fine-touch division (1999: 149). Summarizing Damasio (1999:150-3), the three divisions are as follows:

• The internal milieu and visceral division creates moment-by-momentmultiple maps of the internal milieu by means of introceptive sensingoperations which detect changes in the chemical environment of cells indifferent parts of the body. This division is also concerned with the detectionof pain via nerve pathways. It also maps the state of the smooth musclesthroughout the viscera which are under autonomic control;

• The musculoskeletal division 'conveys to the central nervous system the stateof the muscles which join moving parts of the skeleton, that is, bones' (1999:153). These muscles are under the control of the will and this division of thesomato-sensory is largely concerned with proprioception or kinaesthesis.Again, this division forms multiple maps which are 'placed at multiple levelsof the central nervous system, all the way from the spinal cord to the cerebralcortex. The vestibular system, which is responsible for mapping the body'sspatial orientation, is the highest-order mapping of somato-sensoryinformation;

• The fine-touch division 'describes the alterations which specialized sensors inthe skin go through when we make contact with another object and investigate its texture, form, weight, temperature, and so on' (Damasio 1999: 153).In contrast to the first two divisions, which are concerned with internal states,this division is concerned with external objects as specified by the hapticinformation which is picked up by the body's surface.

We can see here how at this first level of specification, the self of core consciousness is shown to be a self-referential perspective which collects some of theeffects of its own cascading (Salthe 1993: 158). In so doing, the self comes toreveal itself through the mapping of somato-sensory information from the bodilyenvironment in which it is grounded and which constitutes a second with whichthe self interacts. Following the principle of self-as-firstness, this process of selforganization builds up patterned arrangements whereby the self is experiencedas different from, yet grounded within, the milieu which is constituted by thebody. Only in this way can the self come to recognize and interact with other


selves because self-organization entails consistency of behaviour relative to aviewpoint, as embedded in a system of interpretance. The stability of the bodyas-environment-for-the-self provides, by means of somato-sensory mappings, abasis on which the self can model itself. Self-organization thus preserves atextured wholeness or unity in the core self's experience of its own body and ofthe body's experience of the outside world. This grounding of the core self inthe body is founded on a primordial sense of iconic wholeness: self grounded in,yet experienced as continuously varying from, though not categorically distinctfrom, the stable environment afforded by the body.

Interestingly, the three divisions described by Damasio illustrate a progressive(topological-continuous) differentiation from a less differentiated whole, as seenin the gradual move from the introception of the first division to the proprioception of the second division and the exteroception of the third division. As weshall see, this sense of the experiential unity or wholeness of self-in-relation-toits-body, experienced as the ground of its stability and continuity in time, is anecessary forerunner of the semiotic wholeness and unity that is embodied inthe notion of textuality. Again, we see from the proto-textual perspective of theself's iconic grounding in the body that core consciousness emerges whenthe reentrant connectivity of neural groups creates the me ta-loop that closes theloop on the 'self' pole of awareness. That is, the reentrant connectivity of neuralgroups created by mappings of somato-sensory signals leads, in core consciousness, to the closing of the loop on the self's experience of the embodied milieuin which it is irrevocably grounded. From the perspective of the specificationhierarchy, the textuality of symbolic semiosis is integrated with a hierarchy oficonic and indexical levels of proto-semiotic grounding, beginning with the coreself's grounding in the body.

6.4 The Recursive Nature ofConsciousness: Proto-logical MeaningThe ability to link 'scenes' comprising objects and events in primary consciousness to each other depends on the animal's ability to connect these objects andso on through memory to its previous value laden experience (Edelman andTononi 2000: 108). Each such remembered scene qua mental image is availableto conceptual categorization, which means that the given scene can be thoughtabout. In this way, the perceptual experience is available to thinking in the formof pre-linguistic conceptual categorizations. It is important to remember that weare not dealing with language at this stage. Carruthers (1996: 194) points outthat perceptual information about some object is held in short-term memorysuch that the information so stored is available to further reflexive thinking.That is, short-term memory is a form of record which enables each thought tobe re cursively fed back to the same short-term memory store so that eachthought is available so that it may be consciously thought about in its turn byfurther thoughts (Carruthers 1996: 194). To paraphrase Carruthers, consciousness is the reflexive making available to (conscious) thought of thoughts which,in their turn, are made available to further thought, and so on. The short-termmemory store qua record of each thought's being fed back to the same memorystore so that it is available to be thought about in turn may be seen as a precursor,at the lower integrative level in mental life I proposed by Edelman and Tononi

THE METAFUNCTlONAL CHARACTER OF CONSCIOUSNESS 263

(2000: 203-4), of the semiotic notion of text as the products and records ofsymbolic forms of meaning-making, including symbolic thinking, or innerspeech, in the case of language.

In this view, my having a conscious experience of the pen lying on the desk asI write these words is based on my ability to make this experience of the penavailable to further thinking. So, I can see the pen lying on the desk, then havethe thought, That pen is the wrong colour', then think about that thought withthe further thought, 'I must buy a blue one', and think about that furtherthought with the still-further thought, 'I'll go to the stationery shop in themorning', and so on. According to Carruthers's (1996: 195) account, which I amfollowing closely here, each of these thoughts is reflexively fed back into shortterm memory so that each thought is, in its turn, made available to furtherconscious thinking. The (invented) sequence just described may be schematizedas follows:

[CURRENT PERCEPTUAL EXPERIENCE OF PEN]/I[THAT PEN IS THEWRONG COLOUR]/I[I MUST BUY A BLUE ONEJ!I[I'LL GO TO THESTATIONERY SHOP IN THE MORNING]

What emerges in this re-constructed sequence is the way in which each thoughtthat is reflexively fed back into short-term memory so that it may, in turn, bethought about is linked to the preceding thought by principles of causal,temporal, and other modes of logical (temporal, causal, consequential, and soon) connectivity such that a series of thoughts are, for example, logically linkedto each other in an overall process of reasoning about some phenomenon. Inthis way, short-term memory serves as a kind of proto-entextualization ofconscious thinking in the sense that it (memory) enables thoughts to be linkedto each other in meaningful ways. We can see here a pre-linguistic analogue ofthe notion of logico-semantic expansion in natural language. Each thought thatis reflexively fed back into the memory store expands the meaning of theprevious thought(s) in the overall series. Moreover, the fact that the whole seriesof thoughts is held in memory suggests the parallel (on a lower integrative level,as I said before) with the unfolding development of text in natural language.Memory is the means of ensuring that the whole series can be thought about asit unfolds in time, rather than each thought disappearing forever as soon as itis thought. This also means that the reflexive addition of further thoughts tothe previously existing store has the potential to modify the significance ofpreviously held thoughts as they are consciously thought about by new thoughtsthat are fed back into the memory store. What memory does is provide a meansof integrating the temporally unfolding series of thoughts as a larger wholewhich itself, within limits, can be reflexively accessed and consciously thoughtabout. Memory is not just a store against the ravages of time, whatever thetimescale involved. Rather, it is a principle of integration whereby each thoughtthat is reflexively (recursively) deployed to think about the prior thoughts andthe relations between these becomes part of a larger whole which is consciouslyattended to and reflected on.

As we shall see in section 12, the notion of text can be seen as a further


specification and reorganization of this property on the symbolic level. Thisfurther entails that the possibilities for the integration of diverse space-timescales are massively enhanced as compared to the small-timescale short-termmemory store discussed by Carruthers.

7. Integrating the Interaction System and the Meaning SystemPerspectives on (Self)Consciousness

In my view, Damasio's neurobiological proposals concerning the central role ofbodily feeling-states and emotions in consciousness are on the threshold of amore encompassing ecosocial discourse which can show how embodied feelingstates, inner sensations, and primary emotions can be integrated with the socialmeaning-making practices of some community. Damasio's neurobiologicaltheory of consciousness, like Edelman's, is based on neurological and otherprocesses at the level of the individual biological organism. The self-object interactional relationship that is central to Damasio's theory of consciousness isconcerned with the relations between neural events in the brain, electrochemical processes linking brain to the internal milieu of the body, and so on. That is,consciousness is grounded in a complex of physical, chemical, and biologicalevents and relations within the hyper-complex system of the human body-brain.From this physical-material perspective, the body-brain participates in exchangesof matter, energy, and information with its internal and external environments.It was for this reason that I said earlier that consciousness is always founded onlarger-scale supersystem transactions. Furthermore, these exchanges extendbeyond the skin of the body out into the ecosystem which the organism inhabits.In Damasio's neurobiological perspective, consciousness is a physical processlinked by physical, chemical and biological relations. This is the view whichLemke (1984b: 113-21; 1995c: 166-75) describes as the Interaction System viewof a given ecosocial community. That is, the entire complex of physical,chemical, biological and ecological processes and relations which constitute thecommunity and which maintain its physical integrity and survival in time. Theindividual qua biological organism is a relatively small-scale entity within thislarger-scale complex of ecosocial relations.

From the perspective of the individual biological organism, the neurobiological dynamics of consciousness described by Damasio constitute the ground of ahigher-scalar Meaning System (see Lemke 1984b: 113-21; 1995c: 166-75)whereby the physical-material, chemical, and biological processes of the Interaction System have meaning for us. Meanings are immanent in and are constructed in and through the physical-material resources and matter-energy flowsof the Interaction System. Thus, we use and entrain patterns of sound, bodymovement, information in the ambient optic array, and so on, in accordancewith the recognized patterns of meaning-making in a given community, subcommunity, and so on. As Lemke explains, Interaction System processes anddynamics are selectively contextualized by the Meaning System, and at the sametime the Meaning System is embodied in and constituted by Interaction Systemprocesses and dynamics. Interaction System processes and dynamics include the


individual body-brain system and the way this is cross-coupled to other bodybrain systems and to physical-material processes in the environment, includingthe larger-scale material ecosystems in which these physical-material processesare embedded (Lemke 1995b: 118-25; Salthe 1993: 19-20).

I said earlier that current discussion of consciousness often places a lot ofemphasis on the singularity and uniqueness of individual consciousness. Fromthe Interaction System point of view where the neurobiological processes anddynamics of the individual body-brain are focal, this is legitimate - up to a point.Thus, the neurally realized mental images in the individual's brain, the self'sparticular perspective on these, the memory-based conceptualizations of theseimages and their re-elaboration in terms of past and future memories, the innerbody feelings and sensations that cross-couple with and help to entrain the self'sexperiential knowledge and understanding of mental images and so on, are allgrounded in the individual diversity of the self-organizing reentrant neuralmappings whereby a sense of self and individual consciousness is individuatedalong a historical-biographical trajectory. Importantly, this lower-scale individualdiversity also constitutes an important source of potential bifurcations at higherscalar levels of organization. The brain is constantly self-organizing its ownneural connections as well as generating new neurons (neurogenesis). Thesebrain processes are part of the Interaction System which grounds all the possiblemeanings of a community, where Meaning System processes are differentiallydistributed over the diverse body-brains in the Interaction System (see section12, pp. 277-80).

Both the constant reorganization of the brain's neural connections and theprocesses of neurogenesis mean that the enormous diversity of such connectionspossible in individual consciousness yields proto-meanings which are not necessarily typed or recognized by the contextualizing relations of the culture'smeaning systems. In other words, not all of the 'experiences of experiencing'that enter into core and extended consciousness, not all of our subjectiveexperiences of our internal milieu and our external environments, not all of ourperceptions of our inner feeling-states and bodily sensations are necessarilyrecognized as meaningful in our culture's meaning system at any given moment.Many of these experiences remain as proto-meanings within individual consciousness. But because social meaning systems are themselves constantlychanging, though on a slower timescale than those of the much faster timescaleof individual neurological processes, there also exists the potential for protomeanings constructed in individual consciousness to be assimilated to thehigher-scalar meaning system of the ecosocial system on the basis of individualand social innovations in meaning-making practices.

That is, newly emergent Interactional System processes on the scale of theindividual's neurobiological dynamics may be assimilated into the higher-scalarsystem of interpretance on the ecosocial scale. This is possible, I argue, becauseconsciousness is not concerned with neurological processes per se, but with thecontents of consciousness - the mental images and the memory-based relationsamong these - in the perspective of a self. These contents are proto-meanings,as mentioned earlier. Neurological processes realize the contents of consciousness, but are not themselves accessible to consciousness. Proto-meanings in


individual consciousness can, therefore, become objects of further elaborationand reflection, to the point where they can be connected to the symbolicmeaning-making resources of higher-order consciousness and further entrainedto its dynamics. From what we now know about the brain as a self-organizingproducer and elaborator of (proto-)meanings in response to continuallychanging contextual factors both within the individual's body-brain and in theexternal environment, I would argue that the proto-meanings of individual coreand extended consciousness belong to a specification hierarchy such that thereis a continuum of possibilities from the least specified iconic meanings of coreconsciousness to the most specified symbolic meanings of higher-order consciousness. Proto-meanings are not constructed out of nowhere; they, too, areconstructed in response to feedback loops with other systems both within thebrain and the internal milieu of the body as well as in the external environment(context of situation and context of culture). That is, they, too, are contextuallyconstrained. Positive feedback from all of these contextual factors may take thesystem far from equilibrium so that new forms of meaning and organizationemerge. Each level of consciousness in the specification hierarchy entailsincreasing differentiation of the neural dynamics realizing mental images,higher-order conceptual structures and thoughts, and symbolic meanings. Eachlevel in the hierarchy imposes its own constraints on the overall hierarchy suchthat proto-meanings may be entrained to and reorganized by the neuralattractors which embody lexicogrammar in symbolic consciousness.

It is important to point out here that the non-reducibility of proto-meaningsto neurological processes indicates that proto-meanings are realized by neuralactivity in ways which are analogous to the realization of the content stratum oflanguage (its semantics and lexicogrammar) by the expression stratum(phonology or graphology). Language is, of course, a full-fledged social semioticsystem (Halliday 1978a); proto-meanings in the brain are not. However, theprinciple of realization applies in both cases, though not in exactly the same way.The difference lies in the fact that the same mental image (content) can bevariably realized by different patterns of neurological processes (expression)whereas the semantics and lexicogrammar on the content stratum of naturallanguage are constrained by the phonological or graphological patterns oforganization on the expression stratum, and vice versa. Nevertheless, this doesnot amount to an all-or-nothing distinction. Phonemes, syllables, and otherphonological units do not exist as discrete units in the acoustic stream of speech.There is, in fact, considerable merging of the acoustic energy for a given unitwith the energy of other units (Handel 1989: 159). Moreover, the samephonemic segment can be realized by different acoustic cues in differentcontextsjust as the same acoustic cue can realize different phonemes in differentcontexts (HandelI989: 159). So, there is more variation here than what speakersand listeners are usually aware of, though this variation does not, for the mostpart, prevent speakers and listeners from picking up the relevant acoustic information from the articulatory gestures.

Doubtless, this is so because the sensori-motor and neural activities involved inthe individual's linguistic activity are entrained to the dynamics of higher-scaleattractors on the ecosocial scale such as the phonological categorizations of a


particular language system. The study of neurobiological processes in the bodybrain may be able to explain the dynamics of proto-meaning in core andextended consciousness as emergent properties of neurological processes andneuroanatomy. However, it will remain unable to explain the emergence ofsymbolic consciousness in the individual. Any theory which seriously attempts tobreak with in-built language-acquisition devices, bio-programs or languageinstincts will also need to show that language and other social meaning-makingresources exist as dynamically structured attractors at the ecosocial semiotic level.This cannot be explained as a mere bottom-up aggregation of individual bodybrains because it has its own principles of organization on the higher-scalar level.

Whereas the multiple realizability of proto-meanings by neurologicalprocesses in the brain is entrained by lower-level neural dynamics, symbolic consciousness is also entrained by the collective dynamics of entire systems ofmeanings and their associated practices at the cultural level. Proto-meanings inthe individual brain have the potential to become meaning in this semioticallymore specified sense. Because meanings are higher-scalar and collectivephenomena, their effects across many diverse space-time scales are much greaterthan are the proto-meanings that are confined to an individual's brain. This is sobecause symbolic meanings, which are cultural, are potentially accessible andsharable by large numbers of individuals such that they can affect large-scaleecosocial processes across very diverse space-time scales (Lemke 2000a: 191-3).

As we shall see in the following section, recent research on mirror neuronssuggests ways in which the individual's dialogically co-ordinated actions areentrained to the higher-scalar dynamics of the dyad which organizes self-nonselftransactions.

8. Minding the Gap between Minds: Mirror Neurons and InterpersonalMeaning

The dialogic basis of (proto)-semiosis receives support from neurologicalevidence concerning the existence of mirror neurons in both monkeys andhumans (see Rizzolatti and Arbib 1998). Mirror neurons, which are located inthe rostral part of the ventral premotor cortex of monkeys, discharge duringactive movement of the hand or mouth. Transcranial magnetic stimulation andpositron emission tomography (PET) experiments also indicate the existence ofmirror neurons in humans, including Broca's area. Rizzolatti and Arbibdiscovered that this neuronal discharge correlates with intentional action,'rather than with the individual movements that form it' (1998: 188). On thebasis of their findings, Rizzolatti and Arbib explain the existence and functioning of mirror neurons as follows:

The proposal that we and others have advanced is that their actIVIty'represents' actions. This representation can be used for imitating actions andfor understanding them. By 'understanding' we mean the capacity thatindividuals have to recognize that another individual is performing an action,and to use this information to act appropriately. According to this view, mirror


neurons represent the link between sender and receiver that Libermanpostulated in his motor theory of speech perception as the necessary prerequisite for any type of communication. (Rizzolatti and Arbib 1998: 189)

and

Our proposal is that the development of the human lateral speech circuit is aconsequence of the fact that the precursor of Broca's area was endowed,before speech appearance, with a mechanism for recognizing actions made byothers. This mechanism was the neural prerequisite for the development ofinterindividual communication and finally of speech. We thus view languagein a more general setting than one that sees speech as its complete basis.

There is obviously an enormous gap between recognizing actions andsending messages with communicative intent. We offer now a hypothesis ...on how this gap might have been bridged. Whether an individual is about toperform an action or observes another individual performing an action,premotor areas are activated. Normally, a series of mechanisms prevents theobserver from emitting a motor behavior that mimics the observed one, andthe 'actor' from initiating the action prematurely. In the case of action observation, for example, there is a strong spinal cord inhibition that selectivelyblocks the motor neurons involved in the observed action execution (L.Fadiga, pers. commun.). Sometimes, however, for example when the observedaction is of particular interest, the premotor system will allow a brief orefix ofthe movement to be exhibited. This prefix will be recognized by the otherindividual. The fact will affect both the actor and the observer. The actor willrecognize an intention in the observer, and the observer will notice that itsinvoluntary response affects the behavior of the actor. The development ofthe capacity of the observer to control his or her mirror system is crucial inorder to emit (voluntarily) a signal. When this occurs, a primitive dialoguebetween observer and actor is established. This dialogue forms the core oflanguage. The capacity to notice that one has emitted a signal and associatingit with changes of the behavior of others might or might not have developedsimultaneously. However, there is no doubt that, once established, this newassociation should have yielded enormous benefits of adaptive value for thegroup of individuals that started to make use of it, providing the selectivepressure for the extension of communicative capacities to larger groups.(Rizzolatti and Arbib 1998: 190-1)

The existence of mirror neurons indicates that there is a neurological basis tothe enacting of interpersonal relationships between addresser and addressee.The findings of Rizzolatti and Arbib show that dialogically co-ordinated intentional activity entrains neuromuscular activity. This also corresponds to thedistinction made by Peng (1994) between motor functions, which are linked tolinguistic processes, and motor activity, which is not. The point is that higherscalar semantically organized intentions entrain motor functions to theirdynamics (see also chapter 1, section 1, pp. 3--8). Moreover, this entraining is notconfined to the individual organism; instead, it results from the ways in which


individuals are integrated to the essentially dyadic supersystem structures whichregulate interpersonal transactions between individuals.

Rizzolatti and Arbib (1998: 192) somewhat confusedly apply Fillmore's casegrammar in order to derive imperative and declarative case structures, whichthey qualify as 'action structures', rather than linguistic representations. Giventhat case structures in Fillmore's account are concerned with the experientialdomain of (linguistic) meaning, it is more appropriate, from a linguistic point ofview, to relate the imperative and declarative action structures to the moodsystem of the interpersonal grammar. In Halliday's account, mood is centrallyconcerned with the enacting of dialogic, interpersonal relationships betweenaddresser and addressee. Mood is one of the interpersonal grammatical systemswhich indicate the interactional status of utterances; choice of mood integratesthe clause into an interactive context of speakers' goals and intentions. Declarative and interrogative moods express propositions. These two moods hold theproposition in their scope and modify it so as to indicate whether the proposition is being asserted by the speaker or interrogated (McGregor 1997: 214).Oblative and imperative moods express proposals for action, rather than propositions. Proposals cannot be argued about, asserted, believed, or denied in theway propositions can. In proposals, the mood operator holds a proposal in itsscope and modifies it accordingly so as to indicate how the proposal is to beinterpreted or responded to interactively, i.e. as oblatives, in which the addresseris the intentional source of the proposed action, or imperatives, in which theaddressee is selected (by the addresser) as the one who is required to carry outthe action desired by the addresser.

Different languages may express mood in different ways. English uses thesyntagmatic combination of the functions Subject and Finite (declarative andinterrogative) or their absence (imperative) to indicate the mood of the clause;other languages may use a morpheme of some kind, or intonation (McGregor1997: 219). The mood marker, however it is expressed grammatically, acts as aninterpersonal operator which holds the entire clause in its scope and modifies itaccording to the interactive meaning the speaker intends it to have. Mood is agrammatical category which has a specific interpersonal meaning, as discussedabove. The grammatical meaning of the various mood categories in a specificlanguage does not exhaust the meaning of the mood choices in specificcontexts. In specific contexts of use, the abstract mood categories interact withother lexicogrammatical categories as well as with features of the context so as toyield a context-specific semantics.

The neurological evidence provided by research on mirror neurons suggeststhat the interpersonal modification of clauses through the making of differentselections from the mood system is the construction in the interpersonallyorganized neural space of an attractor which is placed just a little ahead of boththe executed action by the addresser and its recognition by the addressee. Indiscourse, the formulating of an intentionally sourced and directed interactivemove according to mood choice thus operates as a top-down control operator(Juarrero 1999: 192) which entrains the neural and motor activity of bothaddresser and addressee in the dyad of the temporally unfolding interpersonalexchange. Thus, neural, motor, linguistic, and physical-material factors may all


become entrained to the semantic organization of the intentionally directedinteractive move. Rizzolatti and Arbib's discussion of imperative and declarativeaction structures, in spite of their confusion between experiential case grammarand the interpersonal-interactive nature of mood, shows how both addresser andaddressee are entrained by the intentional character of the act. The intentionalnature of this interpersonal semantic attractor enacts a trajectory throughsemantically organized neural space so as to ensure that the resulting actions ofboth parties to the interaction are entrained by the dynamic organization that isembodied in the semantics of the intentionally sourced interactive move.

HaIIiday's research in child language development has clearly demonstratedthat meaning is first enacted in interpersonal contexts before being construedexperientially (1993: 103). In the emerging adult language system, the two kindsof meaning gradually come together such that interpersonal selections simultaneously configure with experiential ones in the grammar of the clause. It is theconfiguring of options in meaning from a number of different metafunctionalregions that yields the semantics of a particular clause. This cannot be fullyexplained in terms of any single metafunction in isolation from the others.Intentionally sourced and directed interactive moves in discourse are, then,top-down boundary conditions which are assembled from self-organizing metafunctional constraints which act on lower-level neural and motor activity so as toproduce a particular act-token. This semantic space functions as an attractorwhich entrains bodily (neural and motor) behaviour as well as selected aspectsof the physical-material world along a particular goal-seeking pathway. It is notdifficult to see at this point how the taking up and negotiating of interpersonalmeaning in discourse by social agents is concerned with the enacting andconstrual of a dynamical system's operators whereby participants negotiate atrajectory from one point in semantic state space to another. The interpersonalenacting and negotiating of meaning thus embody top-down boundaryconditions that lead participants through specific trajectories within thissemantic space.

The fact that interpersonal and experiential selections are brought togetherin lexicogrammatical form thus provides a way of explaining how the intentionally sourced dialogic act or move constitutes a strategy for bringing someselected aspect of the world into line with one's experiential construal of it. Inimperatives, for example, it is the dynamical organization of context-dependentconstraints which seeks to make the actional outcome conditional on theintentional source of the imperative clause. The uttering (by the addresser),negotiation (by both addresser and addressee), and consequent execution of thedesired action (by the addressee) is a trajectory, whose finalization, to useBakhtiri's expression, embodies the experiential content of the intention fromwhich it (the trajectory) has its source, and which flows through the entiresystem of constraints to ensure the completion or resolution - both semiotic andmaterial - of the action trajectory.

Rizzolatti and Arbib have pointed out the reciprocally controlling nature ofmirror neurons in the recognition and execution of intentional acts (see above).The stratified nature of language as a semiotic system means that voluntarilyemitted vocal, gestural, and other motor signals on the expression stratum are

THE METAFUNCTlONAL CHARACTER OF CONSCIOUSNESS 271

constrained by the metafunctional dynamics of a self-organizing semantic spaceon the content stratum. It therefore follows that there is no need to seeintentions as separate from and prior to action. Peng's (1994) important distinction between motor functions and motor activities (see chapter 4, section 15,pp. 217-23) means that semiotically significant bodily actions or motorfunctions, such as articulation in the act of speaking, as distinct from nonintentional or involuntary body movements, entail action trajectories thatintegrate the motor functions (not activities) to the contextual, semantic, lexicogrammatical, and expression stratum levels of neuronal organization oflanguage in the brain. In this way, motor functions are entrained into the selforganizing dynamics of a metafunctionally organized semiotic space whichdirects and constrains the always embodied meaning-making trajectories ofsocial agents and their actions along a specific trajectory (section 2).

In the following two sections, I shall explore some of the ways in which innerspeech is no exception to the considerations made in this and previous sectionsof this chapter.

9. Inner Speech as Linguistically Realized Higher-order Thinking

Inner speech is linguistically realized thinking (see also Kinsbourne 2000). Thisdoes not mean that language translates prior, non-linguistic thoughts intolanguage. Rather, inner speech is linguistically constituted thinking. It is a formof language deployment which has been specialized to the inner realm.Moreover, inner speech, as Carruthers (1996: 225) points out, is a form ofconscious thinking through language. Inner speech is language which operateswithout the constraints of the external stimulus flux. This means that verbalization in the inner realm takes place without activating the motor routines that arerequired for the articulation of, for example, speech sounds. However, this doesnot mean that inner speech lacks an expression stratum. Both expression andcontent are constitutive of inner-speech speech just as they are of speech whichis implemented by vocal-tract articulatory activity as speech sounds.

Langacker is one modern linguist who recognizes the psycho-perceptualnature of internal language processes (see also Saussure 1971 [1915]: 98). Fromthe perspectives of both articulation and audition, there are cognitive routinesthat direct both the production and the reception of speech sounds. Thesecognitive routines are directed by stored acoustic images which serve tocategorize speech sounds as being instances of this or that phonological type.Langacker points out that the acoustic or auditory image, rather than thephysical sound, is primary (1987: 79). In this way, listeners have two possible waysof perceiving speech sounds. The listener may actually hear the speech sound asan acoustic event, or he or she may imagine hearing it by the activation of anacoustic image of the sound in silent verbal thought (Langacker 1987: 78-9).Analogously, the speaker may actually implement an articulatory routine andutter a given speech sound, or the speaker may imagine doing so. In the lattercase, this means that the speaker 'can mentally run through the motor routinewithout this mental activity being translated into muscular gestures' (Langacker


1987: 79). Acoustic images direct and control auditory and articulatory routinesin both inner and outer speech.

Acoustic and auditory images are invariants that have been extracted from thestimulus flux such that speakers and listeners are sensitized to these invariantsand can operate these without the constraints of sensori-motor activity. Acousticand auditory images are, then, phonological invariants which are de-coupledfrom the muscular activity involved in articulating and/or listening to speechsounds. In any case, inner speech, whether imaged in consciousness as spokenby one's self or heard as the voice of another, is, nevertheless, stratified: it hasboth expression and content. In the case of inner speech, it is the expressionplane, comprising acoustic and/or auditory images which enable the motorroutine to be activated mentally without, however, translating it into themuscular gestures responsible for the production of actual speech sounds. Innerspeech is always heard as an imagined auditory sequence for this reason. Thisshows how, in both inner and outer speech, motor routines are governed bysymbolic neural space (Peng 1994: 113-14). It is in this way that higher-orderconsciousness instantiates internalized linguistic activity (inner speech) asobjects of consciousness which can be reflected on in the perspective of a self.Just as in outer speech, the meta-redundancy relations between the expressionand content strata mean that sensori-motor routines and conceptual onesmutually recognize each other by virtue of the brain's functioning as a somaticrecognition system (chapter 3, section 13, pp. 162-8). The stratified character ofinner speech means that internalized linguistic tokens can be attended to asobjects of (symbolic) consciousness which are both 'heard' and understood inthe inner realm of imagination. Moreover, there is always a self which interactswith and relates to inner speech as object of consciousness. In other words, innerspeech is always grounded in the perspective of the self who relates to innerspeech as one's own linguistically mediated thoughts (section 11).

10. The Metafunctional Character of Inner Speech as LinguisticallyConstituted Thought

The metafunctional character of inner speech makes it clear that inner speech,like outer speech, does not simply translate non-linguistic thoughts and mentalimages into the medium of language. Language is simultaneously both 'action'and 'reflection' in both its inner and outer manifestations (see chapter 4,section 4, pp. 184-9). The constitutive role oflanguage in symbolic thinking alsomeans that it is not quite accurate to restrict the definition of thinking to the'reflection' dimension of linguistic meaning. Typically, thinking is discussed interms of its 'ideational', 'conceptual', or 'representational' characteristics. But ifinner speech is a form of semiosis which has been specialized to the inner realmof the 'mind', then we shall need to reconsider the limited notion of thinking asideation which has prevailed. As we shall see in the following paragraph, thisrequires us to take into account the full range of meanings that are characteristic of both inner and outer verbal activity. This also requires us to reconsider thestatus of less specified forms of thinking that occur on lower integrative levels


such as the conceptual thinking associated with primary consciousness(Edelman 1992: chapters 10 and 11). Rather than saying that 'thinking' onwhichever integrative level is something which is private and non- or presemiotic, it is logically more appropriate to say that the symbolic thinkingconstituted by language in inner speech is a more specified subclass of internalized semiosis that includes less specified subclasses (of semiosis) on lowerintegrative levels. In this view, all forms of thinking are at least proto-semiotic incharacter. This means that the conceptual thinking characteristic of primaryconsciousness denotes a less specified form of inner semiosis with respect to themore highly specified forms of symbolic thinking characteristic of inner speech.

The discoursal, rather than sentential, character of inner speech requires thatthe definition of semantics be extended to include all the dimensions ofmeaning which are present in any linguistically realized discursive event. In thesystemic-functional framework, language is a semantic resource for (1)construing the phenomena of experience in and through the language's experiential categories; (2) enacting dialogically co-ordinated interpersonal relationsand their associated axiological orientations; (3) construing relations ofcausality, temporality, consequence, result, and so on, between the figures whichcomprise a textually realized sequence; and (4) creating coherence both withinthe text and between the text and its context of situation. Linguistically realizedthinking in inner speech is, then, fully metafunctional in character.

Thus, linguistic utterances that are either heard as the voice of some other orarticulated as one's own voice in the inner realm can play an executive functionin planning and carrying out particular courses of action, in the way originallydescribed by Vygotsky (1986 [1934], 1987 [1934]). In this view, there is no needto oppose the view that language serves above all for the 'exchange of information and the interpersonal coordination of action' to the view that language 'isitself the primary medium of ... thought' (see Carruthers 1996: 2). InCarruthers's view, these two characterizations oflanguage correspond to what hedesignates as the 'communicative' and the 'cognitive' conceptions of language,respectively. According to Carruthers, the two views of language stand opposedto each other. However, the metafunctional character of language shows thatboth experiential reflection and interpersonal enactment are constitutivelyinseparable dimensions of linguistically realized thinking. In other words, innerspeech is an internalized semiotic process which simultaneously allows for bothconceptual (experiential) thinking and propositional (interpersonal) thinking.Insofar as conscious thinking is linguistic in nature, it can therefore be seen tosatisfy both of these requirements: it makes use of conceptual categorizationsand allows for the taking-up of modalized propositional attitudes. Moreover, itexhibits properties of textual unity and coherence in relation to both its internal(textual) organization and its contexts of utterance.

Inner speech just is higher-order or symbolic thought. Moreover, the stratifiednature of inner speech means that the utterances that occur in consciousness areheard and/or spoken in the perspective of the self who attends to them asobjects of consciousness. That is, the expression stratum of the phonologicalinvariants - the motor routines - that operates without the constraints of theexternal stimulus flux nevertheless provides an imagined perceptual basis to


inner-speech events. We imagine them as spoken or heard linguistic tokens. Inboth cases, they are grounded in the perspective of the self who attends to themas objects of conscious awareness which are immediately accessible to the self assuch (see Carruthers 1996: 238). For this reason, tokens of inner speech areobjects which are immediately given to consciousness as objects which can bereflected upon. The activation of the imaged phonological or other expressionstratum routine in inner speech provides a perceptual basis to inner speech eventhough this is a form of inner perception of phonological invariants that havebeen detached from the stimulus flux. Nevertheless, the perceptual systemswhich detect and filter phonologically salient distinctions provide immediateaccess to the content stratum (lexicogrammar and discourse semantics) which isrealized by imagined phonological routines in inner speech.

My use of the term 'imagined' in this context may suggest that there issomething less real or less concrete about the existence of these phonologicalroutines in inner speech. Nothing could be further from the truth. Innerspeech, like outer speech, is stratified. The notion of stratification is necessaryfor explaining the relationship between different orders of abstraction in theorganization of language. Lexicogrammatical and discourse semantic units andrelations on the content stratum are not comprised of phonological units on theexpression stratum. Nor is the expression stratum simply a material means forconveying content. Rather, content stratum units and relations are realized byexpression stratum ones. The realizational relationship links articulatoryprocesses of the body to discoursal meanings by virtue of the fact that motorroutines and conceptual ones mutually recognize each other. In other words, thetwo strata mutually redound with each other. Inner speech qua object ofconscious reflection must be immediately accessible to SELF. This requires, inthe first instance, a sensible dimension which is accessible to perception. Thebrain's functioning as a somatic recognition system thus provides a perceptualbasis for inner speech without the constraints of the stimulus flux. Moreover, thephonological routines that are 'imagined' in inner speech show how the bodilybasis of in ternalized semiosis is preserved. The point is that all forms of semiosisrequire some kind of expression stratum which is accessible to and/or producedby sensori-motor activity. There is no content floating free of its expression.Without an expression stratum, the brain would have no means of recognizingthe content which is realized by the expression stratum. Once again, thestratified nature of inner verbalization highlights the intimate and internalconnection between sensible bodily processes and meaning-making.

Moreover, it lends further support to the view that inner speech is a specialized internalized deployment of the same semiotic resources that are used toengage in outer discursive activity. We can see the specification hierarchy at workhere as well. We have already seen in chapter 3 (section 9, pp. 139-46) how thebrain's functioning as a somatic recognition system relates sensori-motorroutines to conceptual representations in primary consciousness. Similarly,perceptual categorization occurs on the basis of the organism's linkingmultimodal stimulus information which is picked up in its environment throughsensori-motor activity to perceptual categorizations that are built up throughreentry in the brain. Inner speech, we have seen, occurs on the basis of imaged


phonological (or other expression-stratum) routines redounding with theappropriate content-stratum ones. The basic principle is the same at all levels.What is different is the way in which each stage is more specified and moreexplicitly elaborated but, at the same time, with the prior stages not beingtranscended. Rather, later stages integrate prior ones.

A symbolic being that routinely deploys inner speech in higher-orderconsciousness is still a being which makes use of conceptual and perceptualrepresentations. Each level refers to the emergence of a level of complexity inthe development and individuation of the organism along its trajectory. Priorintegrative levels (e.g. perceptual and conceptual) refer to proto-semiotic ordersof information and categorization that the organism has elaborated along itsdevelopmental trajectory on the basis of organism-environment transactions. Bythe same token, the transition to inner verbal activity means that the individualis cross-coupled to modes of thinking that can only occur through the semanticcategories of natural language and the discourse genres in and through whichlinguistic meanings are organized as recognizable social situations. It makes nosense, then, to say that language simply 'translates' mental images into a publiclyexpressible and communicable form (cf. Damasio 1999: 107-8), for this istantamount to saying that language plays no role in the constitution of thecategories of higher-order or symbolic consciousness. Instead, the symbolicresources of language mean that consciousness is connected to space-time scaleswhich go far beyond the organismic one.

The further implication is that inner speech qua linguistically constitutedhigher-order thinking is most centrally a question of meaning. Furthermore,inner speech cannot be adequately characterized in terms of sentential formswhich express linguistically constituted thoughts. The point is, rather, thatlinguistically mediated thought is a form of contextualized meaning-making thatis specialized to the inner realm. Vygotsky (e.g. 1986 (1934]: 248) firstrecognized the semantic character of inner speech. The semantic character ofinner verbal activity requires us to see that inner speech is a form of text ordiscourse, though not without its own specific characteristics, as first studied byVygotsky. It is not comprised of sentences. Lexicogrammatical units realize theirdiscourse semantics; they do not simply comprise them. The highly condensedand elliptical features of inner speech identified by Vygotsky as a high degree ofsemantic condensation and semantic opacity (Vygotsky 19S6 (1934]: 243-8;Thibault 1995c) may be explained by the fact that the meanings expressed ininner speech are only accessible to the consciousness of the SELF in whoseperspective the given occurrence of inner speech is grounded; there is nofunctional requirement that these meanings be interpreted by others in thepublic realm. Instead, inner speech is a semiotically mediated means wherebythe SELF is able to gain conscious access to its own thinking, in the processachieving self-monitoring, self-control, and self-awareness (see also Carruthers1996: 247).


n. The Re-grounding of the Perspectives of Self and Other in SymbolicConsciousness

Symbolic consciousness is freed from the kind of indexical tie to a givenhere-now situation that is characteristic of perceptual awareness (section 3).Linguistically construed objects of symbolic consciousness always entail a SELFwho is aware that the object of consciousness is experienced by the SELF. Thenotion of SELF is defined as a self-referential perspective implicating observational perspectives or viewpoints. As we shall see below, a SELF is implicated inall occasions of semiosis, whether explicitly specified or not by linguistic orother means. The two poles of attention that Gibson identified in the act ofperception apply here as well (section 3). Independently of whether the objectof symbolic consciousness is spoken, heard, written, or read by the SELF, thelinguistic object of consciousness is always grounded in the perspective of aSELF who experiences the object of consciousness. For the same reason, theSELF can 'hear' the voices (the utterances) of others in his or her inner speech.In the case of language, the difference with respect to perceptual awareness liesin the way in which the objects of symbolic consciousness can be re-groundedin the perspective of other selves who can experience the object of consciousness from their own perspective (Deacon 1997: 427; this volume chapter 4,section 2, pp. 173-6). The fact of my speaking or writing a particular instanceof language on a given occasion does not mean that the resulting text isuniquely tied to or is uniquely grounded in perspective of the SELF who utteredit. Rather, my interlocutor (listener, reader) re-grounds it in the perspective ofhis or her self in the process of dialogically orienting to and taking up theparticular instance of language.

In this way, person deixis provides resources whereby first, second, and thirdpersons and the meanings associated with them can be grounded in the perspective of a given SELF. This means that the experience of a given object ofsymbolic consciousness is not uniquely indexically tied to the first-person perspective of the person who uttered it. Furthermore, the resources of temporalproximity deixis and modality proximity deixis in the Finite element are notuniquely tied to the perspective of the self who originally produced theparticular instance of language (see chapter 3, section 7, pp. 131-2). Again,the addressee re-grounds the temporal or modal orientation that is expressed inthe Finite element in terms of his or her own perspective as the one whoundergoes the particular symbolic experience. These interpersonal grammaticosemantic resources provide a dialogically co-ordinated frame of reference inwhich the temporal and modal (attitudinal, evaluative, and so on) orientationsof diverse selves can be co-ordinated across different space-time scales. Bothtemporal and modal proximity deixis do not simply index the addresser'stemporal perspective or modal evaluation of some proposition; they also work toorganize the addressee's orientation as well. The same applies to the spatialproximity and other forms of deixis that operate in the nominal group. Here,too, the addressee re-grounds the given object of consciousness in the perspective of his own self rather than in that of the person who first uttered orotherwise produced the utterance.


Person deixis and finiteness ground the symbolic object of consciousness in adialogically co-ordinated frame of reference. Person deixis enables meaningspertaining not only to the first-person perspective of SELF, but also meaningspertaining to second and third persons to be made relevant to and integrated tothe first-person observational perspectives of SELF. The resources of temporalproximity deixis and attitudinal proximity deixis in the Finite element dialogically co-ordinate addresser and addressee in terms of the temporal or modalgrounding of the proposition. This means that addressees can re-ground thetemporal or modal perspective specified by the Finite in terms of their ownperspective. Both of these resources show how, from the point of view of theinterpersonal metafunction, there is a dialogic complementarity between theperspectives of SELF and NONSELF. This means that the SELF's awarenessthat it is the SELF who is consciousness of the given object can be re-groundedin the perspective of some other SELF. The dialogic complementarity of SELFand NONSELF in the exchange process refers to the way in which the interpersonal semantic structure of a given dialogic move, as spoken, written orsigned by some addresser, results in a corresponding change in the addressee.Thus, the addressee is internally modified in ways which result in the dialogic coordination of deictic perspectives. The information which is supplied by persondeixis and the Finite are principles of order whereby cross-individual dialogicprocesses integrate individuals to higher-order dialogically co-ordinated framesof reference. At the same time, they also provide individuals with resources forbeing conscious of being conscious of something in the perspective of a SELF asa locus of (inter)action and intentionality.

12. Text, Social Meaning-making Practices and Higher-order Consciousness

The artefactual nature of text-as-material-object can be seen as a further development and specification of the function of short-term memory that I discussedin section 6.4 in connection with the work of Carruthers. The kind of reflexivethinking described by Carruthers as necessary for consciousness is confined tothe organismic scale of short-term memory. On the other hand, linguisticallymediated discursive events, along with their entextualized products and records,mean that consciousness itself is extended across diverse space-time scalesbeyond the organismic one in its immediate environment. We saw earlier that inCarruthers's account short-term memory enables each thought to be recursivelyfed back to the same short-term memory store. In this way, each thought isavailable so that it may be consciously thought about in its turn by furtherthoughts. The recursive nature of this process is homologous with the logicalmetafunction in language. The recursivity described by Carruthers (see section6.4, pp. 262-4) is a less specified version of the same kinds of logico-semanticresources that construe relations of causality, temporal sequence, and so on,between clausally realized figures in a given sequence. The logico-semanticresources of expansion enable clausally realized figures to be related to eachother as sequences of figures (chapter 4, section 6, pp. 195-8). The relations oflogico-semantic interdependency that are created between figures provide a


means whereby each figure of the figures in an unfolding sequence or even thesequence as a whole can be consciously reflected on by further figures as theyemerge.

Unlike the perceptual information referred to by Carruthers, an entextualized sequence oflinguistic figures is not necessarily limited to the small temporalscale of short-term memory. Carruthers's notion of reflexive thinking shows howeven perceptual information on the immediate here-now scale of the body'sinteractions with its material environment can be implicated in space-time scalesthat go beyond the here-now. The fact that thoughts can be consciously thoughtabout by other thoughts means that the thoughts and the sequences of thoughtsso thought about are embedded in larger space-time scales. The artefactualcharacter of linguistic text represents a massive expansion and elaboration ofthis possibility. Text can be viewed as material object or artefact; it can also beviewed as a semantic process of making or operationalizing meanings in somecontext. The dual status of text as material and semiotic entity means that (1)texts can be used, manipulated, transported, stored, and so on, on theimmediate here-now scale as objects; and (2) their embedding in specificmeaning-making practices in which their semantic potential is activatedconnects the here-now object (the letter from a family member that I hold in myhand) to far larger semiotic scales which are condensed in the lexicogrammatical and discourse patterns of organization of the text as a unit of meaning.

Textually mediated interactions between individuals expand the possibilitiesfor more and more indirect kinds of interactions with other individuals andinstitutions on different space-time scales. The proto-interpersonal relationscharacteristic of the early socio-affective dyads between mother and infant arethus expanded to include an increasingly differentiated array of interactionsbetween increasingly differentiated first and second persons. Subsequently, thisincludes third persons who are made the focus as the dyad expands outwards tointegrate with the perspectives of such persons. It is not difficult to see here howthis increased expansion outwards to include third persons of various kinds is, intime, extended to include non-human objects of varying degrees of abstractionas semiotically mediated objects of symbolic consciousness, including thoseconstituted by purely linguistic means.

We have considered how metafunctional constraints which are intrinsic to theorganization of linguistic and other semiotic forms enable meanings to beenacted and construed in and through semiotic forms of all kinds. Moreover, Ihave considered how the metafunctional basis of linguistic and other semioticforms has eo-evolved with human beings and the ecosocial forms of organizationin which human activity is embedded and by which it is both constrained andenabled. It is this hook-up between individual human agents and the higherorder boundary conditions constituted by the ecosocial system which enable theemergence in individuals of complex, individual properties and viewpoints onthe basis of the self-organizing neural networks which are formed along a timebound trajectory in the individual brain. Higher-order or symbolic consciousnessis the result of the self-organization of symbolic space whereby human individuals qua agents can carry out intentional symbolic activity in their interactionswith others. The systemic possibilities of lexicogrammar and other semiotic


modalities constitute, as Halliday (1993) has pointed out, a multidimensionaland highly elastic semantic space which enables human beings to construeexperience, to enact social-interpersonal relations with others, and to constructoften extremely complex acts of internally coherent discourse. The meaningpotential of language, depiction, and so on, constitutes a phase space wherebyindividuals are provided with systems of choices - meaningful alternatives - formaking meanings in specific contexts.

The notions of scalar heterogeneity and complex dynamic open systems thatinteract with their ecosocial environments under conditions far from equilibrium show us how social agents build both history and context into their owninternal dynamics. Our symbolic meaning-making resources allow us to extendourselves beyond the here-now scale of our immediate surrounding milieu so asto selectively interact with diverse, often very different space-time scales, as wellas to internalize these interactions so that they become part of the individual'sInnenwelt. In this way, we interact with not just a series of immediate contexts ofsituations, but also the context of culture which the former mediate and in whichthey are embedded. Thus, symbolic - not iconic and indexical- semiosis enablesus to interact with and internalize scalar heterogeneity. It is our being situated atthis intersection of maximally diverse scales, mediated by our symbolic resources,which in turn acts on our internal dynamics and modifies them.

The further question that arises here concerns the constraints on theparticular forms of meaningful action that can be carried out by agents inparticular forms of social organization. If the systemically organized meaningmaking resources of language and other semiotic modalities both enable andconstrain the kinds of meanings that can be made, then we need to investigatehow variation in the relationship between semiotic forms and the socialmeaning-making practices in which agents are positioned leads to variation inthe ways in which consciousness is organized. The higher-order systems ofmeaning-making practices of a given culture constrain both the agent'strajectory and his or her internal dynamics to the global probability landscape e.g. the coding orientations - of the culture as a whole. Thus, individuals areentrained to the overall dynamical organization of a particular coding orientation (Bernstein 1971, 1990) or habitus (Bourdieu (1977 [1972]) such that, inlogogenesis, the probabilities regulating the individual's access to, participationin, and understanding of specific texts, discourse genres, and ways of makingmeaning are pre-set. As Hasan (1992b: 520-1) points out, variability in the waysin which different categories of social agents are positioned in and orient to thesocial structure leads to socially significant differences in the ways in whichdifferent categories of individuals in a given community orient to and deploy intheir daily life the meaning-making resources of the community. This furthersuggests that the semiotically mediated nature of consciousness will lead tovariant forms of consciousness according to the variable ways in which individuals are positioned by the differential distribution of power and control and,hence, to the semiotic and material resources of the society.

The slower rate of the higher-scalar attractor space (e.g. the coding orientation which an individual belonging to a particular social group habituallyaccesses) regulates lower-level individual dynamics by entraining its much faster


dynamics to the slower ones on the higher level. This suggests that the very fastdynamics of neural activity in the brain are entrained to the higher-scalarconstraints of particular coding orientations through the creation of attractorsthat reorganize the patterns of firing of individual neurons such that these areconstrained by the attractor space that the coding orientation constitutes. AsBenjamin Lee Whorf long ago understood, though without the benefit ofmodern neural science, higher-order consciousness is shaped and entrained bywhat he referred to as 'a whole large order of experience' (1956a: 137). That is,it is shaped by higher-scalar dynamical processes on the ecosocial level.Moreover, Whorf also understood that language is a more specified integrativelayer of something far more general, which 'may generalize down not to anysuch universal as "Language", but to something better - called "sublinguistic" or"superlinguistlc" - and not ALTOGETHER unlike, even if much unlike, what wenow call "mental" (Whorf 1956b: 239).

6 Brain, Meaning, and Consciousness

1. Biological and Socio-cultural Factors Form a Single System of ComplexlyRelated and Interacting Factors: Putting Time and Activity Back into thePicture

The science of language has existed in both the Western and Eastern traditionsfor at least the past two or three millennia. Linguists have been able to makestatements about the phonological, graphological, lexicogrammatical, semantic,and discourse aspects of language because there exist neurological and sensorimotor processes that make the level(s) of organization about which linguisticstatements are made possible. The discourses of biology are, of course,concerned with the former level. It does not follow from this line of reasoningthat statements about language qua semiotic system must necessarily be madewith reference to or reduced to statements about biology. The fact that severalmillennia of thinking about and theorizing about language has taken place inthe absence of our modern understanding of biology shows that such a nonreductive discourse about language is eminently possible. Nevertheless, thehistorical order in which various scientific discourses have appeared should notdistract us from the specific logical problem I am addressing here. Much of ourreceived wisdom about language emerged well before the onset of a crediblebiological discourse. This fact has tended to exacerbate the trend towards talkingabout language as a metaphysical object that is dissociated from both nature andour bodies. The evidence for this trend abounds in many of the theoreticaldiscourses about language that have predominated in the twentieth century. Theresulting dissociation has tended towards fragmented and arbitrary accounts oflanguage which do not connect with the biological dimension of our being.

In chapters 4 and 5, we saw that the biological and socio-cultural dimensionsof the specificity of the human species need not be dichotomized as two separatecausal domains. In recent years, neurobiology has made considerable progressin showing that it makes no sense to oppose the human capacity for linguisticallymediated meaning-making and higher-order symbolic 'thought' to ourbiological being. This has often led to the tendency to view the latter assomething lofty and 'spiritual' and therefore beyond biology whereas the latteris seen as belonging to the crassly physical-material realm. In reality, the formerbelong to our biological make-up, and at the same time our biology has evolvedso that we are the socio-cultural and symbolic beings that we are. Language in all


of its facets is intrinsic to our biological make-up. There is continuity with otherspecies, as the ape-language research of Savage-Rumbaugh et al. (1998) shows,though this in no way implies that human biological specificity is reducible tothat of other species. The more fundamental point is that all forms of biologicalorganization are founded on elementary forms of semiosis whereby organisms ofall kinds selectively interpret and adjust to their specific version of reality in theprocess of constituting their Umwelten. This fact is evident in the infant-caretakersemiosis characteristic of primary intersubjectivity.

The reciprocal contributions of both caretaker and infant to the dyadscharacteristic of primary intersubjectivity (Trevarthen 1987, 1992; Halliday 1975,1993) are based on forms of matter, energy, and information exchange bothbetween the two members of the dyad and between the dyad and the higherorder ecosocial system in which the dyad is embedded. Semiosis, includinglinguistically mediated semiosis, emerges as a consequence of these processes ofexchange. The fact that language is emergent from such forms of exchangemeans that it is intrinsically dialogical and social. Therefore, the biologicalorganization of the human species is inherently dialogic from the outset. Theelementary forms of semiosis that characterize the caretaker-infant dyad inprimary intersubjectivity are what make the emergence of language in individuals possible. Doubtless, the genetically inherited value wash predisposes thenewborn to participate in the elementary forms of dialogic communication withcaretakers that Trevarthen has identified as primary intersubjectivity. In turn,this participation makes possible the infant's accessing and selecting of morecomplex forms of communication along his or her trajectory.

Phylogenetically, it is the emergence of language from prior forms ofmeaning-making which makes possible the genetic predisposition to language.Rather than an innate linguistic competence which grows into language in theindividual in a pre-programmed way, it is newly emergent socio-cultural forms ofmeaning-making which act on and select the biological substrate so as to modifyand entrain it along a specific cultural trajectory so that language and communicative ecosystem both eo-develop and co-individuate in the individual (Bateson1980; Deacon 1998 [1997]; Lemke 2000a; Thibault 2000a; 2003b). As Wheeler(1988: 15) points out, 'a mind with no programming is no mind'. The mind isprogrammed by the networks of meaning-making in which it participates. It isthe dialogic engagement of world by organism whereby self interrogates andreceives response from the world in the exchange process that enables meaningto be exchanged between the two (see also Wertsch 1991). In the process ofinterpreting the world through exchange, we also affirm our continuity withothers linked in the same networks of exchange. This is possible because of thesystematic nature of our meaning-making resources. Chomsky's nativisthypothesis, by contrast, sees the programming as being built into the geneticinheritance of the individual from the outset in the form of hard-wired formalrules in the brain.

Chomsky's (1965) theory attempts to make explicit the formal properties thatdefine the range of human languages. It is in this sense a characterization of theuniversal properties of human language. Competence excludes meaning andthe relationships among language users, their language, and context. The latter

BRAIN, MEANING, AND CONSCIOUSNESS 283

refers to performance. Linguistic competence is seen as an abstract and formalknowledge which sits above the general cognitive system comprising memory,attention, real-time perception, and so on. Chomsky claims that competence isbiologically constrained and characterized by knowledge of abstract linguisticrules which are genetically determined. The brain is thus seen as a store for alarge number of formal linguistic rules that are hard-wired in the brain alongwith the means for operating on these according to predetermined logicalcriteria. Two problems with this theory are the failure (l) to provide convincingbiological criteria for showing how abstract linguistic rules exist in the brain and(2) to demonstrate how competence is manifested in real-time discourse activity.The arguments to be developed below present a very different view.

In chapter 4, section 10, pp. 202-5, I argue for the need to refocus onlanguage development as an epigenetic process along a developmentaltrajectory which is typical of members of a given species. The human organism- the body-brain complex - is adapted to eo-develop and co-individuate togetherwith language by virtue of its - the organism's - being embedded in a specificecosocial system. Moreover, the body-brain complex is itself constructed andmodified by the matter, energy, and information-meaning exchanges it engagesin along its trajectory. Language, rather than being innate in the child from theoutset, has 'universal' properties and functions on account ofthe species-specificgenotypic commonalities and type-specific developmental pathways such thatsome kinds of information and structure rather than others are selected fromthe developmental and cultural context as being relevant to the development ofthe child in ways which are typical of the species (see also Cowley 2001). ThePinker-type distinction between 'what an organism learns' and 'what a braininnately knows' has no place in this conception (Pinker 1994,1997). Moreover,an individual's language potential is stored in the brain not as hard-wiredabstract rules, divorced from time and context, but as memory-governednetworks of associations which are acquired and adaptively modified along theindividual's time-bound trajectory. There are important developmental andindividuating implications of this claim.

2. Re-interpreting Flohr's Brain-based Theory of Phenomenal Awareness: AThree-level Hierarchy View of the Emergence of Proto-meaning in the Brain

Hans Flohrs research is dedicated to overcoming the prevailing dualismbetween physical versus phenomenal brain states. Flohr (1991: 248-9) outlinestwo classes of inputs which are received by the cerebral cortex as follows: (1)specific afferents from sense organs which are relayed through specific thalamicnuclei before reaching the primary cortical projection areas; and (2) nonspecific afferents which reach the cortex from the mesencephalic reticularformation, where the term reticular formation applies to 'a network of neuronsand nerve fibers extending from the caudal medulla to the diencephalon' (1991:249). Such neurons form 'a widely distributed network of synaptic contacts'(1991: 249). Moreover, it is now established that this formation 'is organized intodifferent nuclei with specific afferent and efferent connections' (1991: 249).


Flohr explains the projection of these specific and non-specific afferents withinthe cerebral cortex as follows:

Within the cerebral cortex, primary auditory, somatosensory and visual fieldseach project to adjacent unimodal association areas which, in turn, project tocontiguous secondary unimodal association fields. It is assumed that thesesuccessive areas are concerned with progressively higher-order abstractionswithin a modality. The unimodal association areas project to a number ofpolymodal sensory areas lying in the cingulate gyrus, parietal, temporal andfrontal lobes. Their function is vaguely described as crossmodal associationand synthesis. The polymodal association areas project to the inferior parietallobe which has been termed a 'supramodal' area. Polymodal and supramodalregions have connections to the limbic system; these connections provide theanatomical substrate by which motivational states influence corticalprocessing of sensory stimuli. (Flohr 1991: 249)

Flohr points out that previously consciousness had been understood asdepending on the convergence of specific and non-specific afferrent systems,'both of which contain complex feedback loops and are interconnected atvarious levels' (1991: 250). However, Flohr also points out that this understanding remains 'vague' after several decades of research 'into the role of the brainstem in the generation and maintenance of consciousness' (1991: 250).Moreover, it remains unable to explain both 'the difference in the subsequentcortical processing of sensory data distinguishing conscious and unconsciousbrains' and 'the specific nature of those brain processes which result from thisconvergence and whose occurrence is a precondition for the occurrence ofconsciousness' (1991: 250).

The central problem that remains for Flohr is a two-fold one, viz. (1) whathappens beyond this point?; and (2) which specific physiological states andprocesses result from this convergence of (specific and non-specific) afferentsignals such that these might be 'constitutive of phenomenal experiences'?(1991: 250). In order to formulate a hypothesis which may provide a solution tothe problem just outlined, Flohr first distinguishes two categories of possibleneural processes, which are the result of the interaction between specific andnon-specific afferents (1991: 251):

1. 'conventional' synaptic transmission, i.e. signal propagation in neural netswith rigid structure and given synaptic weights;2. activity-dependent self-organization of neural nets, i.e. changes in synapticconnections which are induced by synaptic activation and dependent on thedegree of activation. (Flohr 1991: 251)

According to Flohr's hypothesis, the occurrence of phenomenal states in thebrain is dependent upon processes that pertain to the second group of factors.On this basis, he proposes that phenomenal states depend on the rate of 'activitydependent synaptic changes' and the rate of formation of 'neural assemblies'(1991: 251).


The concept of neural cell assemblies, as Flohr (1991: 251) points out, wasfirst developed by D. O. Hebb (1949). Hebb's conception is also central in thework of Edelman (1987). Hebb proposed that '... repeated stimulation ofspecific receptors will lead slowly to the formation of an 'assembly' of associationarea cells which can act briefly as a closed system after stimulation has ceased .. .'(quoted in Flohr 1991: 251). Hebb had proposed that such an assembly is thesimplest form of 'representative process' of an image or idea. Flohr continues:

The basis of this self-organization of neural nets is the presence of plasticsynapses, now termed Hebb synapses. Hebb assumed that synapses on aneuron that are active while the neuron discharges will be strengthened,whereas inactive synapses will be weakened. Synapses from differing inputsthat are active at the same time on the same neuron will be reinforced andselected over others.

When applied to nets of spatially distributed groups of neurons with nonspecific, random interconnections, these so-called Hebb rules lead to arelative stabilization and association of neurons firing in a correlated fashion.An assembly of preferentially connected, coherently active cells is formed. Ifcoincident activity is induced in some neurons of such a net by a patternedinput, an assembly will be formed because the synchronous activation selectively modulates the pathways connecting these neurons. The assemblydetects and encodes the coherent properties of the stimulus pattern such thata representation of that pattern is generated. Once the assemblies have beenformed, they would function as detectors of the same or similar input patternsexpressing the detection of coherent features by co-ordinating their activities.It is easy to envisage that the output of such assemblies could in turn be usedas input to other modifiable nets which then would reorganize their structureas a function of this input. Iteration of such processes would generate moreand more abstract metarepresentations ...

These nets are able to create topographical representations in which therelationships between elements of the stimulus pattern become encoded intospatio-temporal patterns of activities. Moreover, in such nets abstractions fromprimary sensory data at increasingly higher levels of generalization areformed and stored. Synaptic plasticity is the means by which neural netsdetect and represent coincident properties and, hence, the nomologicalrelations existing between elements of the input pattern. (Flohr 1991: 251-2;italics in original)

Flohr's proposals can readily be assimilated to the three-level hierarchy, asfollows:

L+1: encoding of the properties of the stimulus pattern such that a representation of that pattern is generated;

L: application to nets of spatially distributed neurons leads to stabilizationand association of neurons firing in a correlated fashion to form anassembly of globally coherent cells;


L-l: strengthening and/or weakening of synapses on different units whileneuron discharges as induced by patterned stimulus pattern information(inputs) .

This shows how the strengthening of neural discharges on level L-l constitutesinformation which is reorganized on level L as a global cortical effect in the formofan 'assembly' of 'coherently active cells' which are, in turn, interpreted at levelL+1 as being a representation of a stimulus pattern. The fact that assemblies,once formed, can function as 'detectors' of the same or similar patterns overlonger periods of time shows that they function as filters or buffers such that onlythe salient information from neuronal discharges on level L-l is recognized onlevel L. Moreover, the fact that such self-organizing nets create topological representations on level L+1 spatio-ternporal patterns of activity shows how the representations created on this level constitute a system of interpretance comprisingrepresentational categories. For this reason, globally coherent assemblies of cellson level L are appropriately interpreted as representations of this or that representational category.

A critically important component of Flohr's argument is the postulatedexistence of rapid synaptic changes as against the previously dominantassumption that synaptic weights remain constant in the performing of neuralnets. Flohr argues that the NMDA-receptor channel complex acts as themolecular agent which brings about rapid synaptic changes. I shall quote Flohrat some length on this rather technical question before venturing some remarksof my own:

The essence of Hebb's postulate is that a synapse is strengthened if thepresynaptic activation of the synapse is coincident with the firing of the postsynaptic neuron. This poses two difficulties:(a) some mechanism of the synapse must respond to coincident pre- andpostsynaptic activate and initiate temporary changes in the weights of thesynapse;(b) in the case of Malsburg-type synapses this mechanism must be fast enoughto allow rapid changes[... ]One (but probably not the only one) such mechanism that could fulfil bothconditions has been described. Much recent data indicate that the NMDAreceptor channel complex may serve as the molecular mechanism for Hebbtype synapses. This channel is voltage-dependent ... , i.e, it opens under twoconditions: (a) presynaptic activity that releases a transmitter that binds to thereceptor, and (b) sufficient postsynaptic depolarization. At membranepotentials around resting potential, NMDA channels are blocked bymagnesium ions; if the membrane is depolarized, to say -30 to -20 mY, theblock is removed and the channel is switched on. The NMDA reccptor-linkedchannel is permeable to Na', K' and Ca". Thus, NMDA receptors provide amechanism to (a) amplify synaptic currents, and (b) increase Ca" influx ifaparticular level of postsynaptic depolarization is present. Ca" is hypothesizedto trigger a number of Ca"-dependent molecular changes inside the post-


synaptic terminal that modify the efficacy of synapses (other than those gatedby NMDA). (Flohr 1991: 253)

Postsynaptic depolarization through the effects of the NMDA-receptor meansthat the afferent stimulus patterns which impinge upon the neurons belongingto a coherent population are indexical signs of inner and outer environmentalevents. However, the amplification of synaptic currents, including the firingproperties of the postsynaptic neuron, on detection of correlated signalsconverging on a single cell, leads to the formation of positive feedback loopssuch that neurons near the threshold level will 'persist as a co-operative,coherently active subpopulation' (1991: 253). It is this depolarization whichleads to the formation of coherent assemblies. Depolarization of the postsynapseis the translation via the coherent cell assembly of the indexical signs - theafferent inputs - related to some environmental event into symbolic signs whichtranslate the indexical sign from the inner perspective of the coherent cellassemblies which are formed in the cerebral cortex. In this way, indexical signsof the Umweltare translated into subpopulations of neurons. The latter, followingSalthe (1993: 176), are symbolic because they do not stand in any necessaryrelationship to the afferent inputs relating to environmental events. The fact thatthese same assemblies can then function as 'detectors' over longer timescales, asdiscussed above, constitutes the basis on which an inner perspective whichconstructs models of its Umwelt is built up. By contrast, the afferent input quaindex of environmental event occurs on the shorter timescale of the here-nowoccurrence and pick-up of that event. It is on the basis of such models that asystem of interpretance emerges. The phenomena of consciousness are attendedto and experienced from the inner perspective of a self.

Flohr (1991: 255) draws attention to the rate of activation of a cell population:unconscious brain structures exhibit a lower rate; conscious ones a much fasterrate of assembly, which reaches and crosses a critical threshold. This emphasis onthe rate of assembly activation follows an important principle of scalar hierarchytheory, viz. processes on lower levels in the hierarchy have faster temporal cycleswhile those on higher levels have slower ones (Salthe 1993: 46). The faster rateof assembly of cell populations in the formation of a conscious experience goeshand in hand with the fact that the conscious experience itself is experienced ona slower timescale which corresponds to that of the observer perspectives andcategories of the person who undergoes the conscious (phenomenal)experience. In accordance with the logic of scalar-hierarchy theory (Salthe 1993:36-52), we could say that a phenomenal experience is a higher-scalarphenomenon on level L+1 (with respect to the lower-scalar entities andprocesses on levels L-l and L) which has a 'cogent moment' that lasts longerthan any of the lower-scalar processes that sustain it. The cogent moments ofconscious experience therefore reflect timescales pertaining to the higher-scalarentities of the observer's perspective rather than the lower scale of, for example,the synaptic discharges on level L-l, which are not accessible to consciousness(section 9, pp. 310-13). Moreover, each level in the three-level hierarchy exhibitsproperties which are specific to that level such that relations and processes ondifferent scalar levels are, to some extent, ontologically isolated from each other


(Salthe 1993: 46). It is this factor of ontological isolation which enables us tounderstand how phenomenal consciousness can emerge from physical brainstates and processes on different scalar levels without being reducible to thelower-scalar relations and processes.

A further important aspect of Flohr's hypothesis concerns the possibility that'iteration' of such processes can lead to the formation of increasingly abstractmeta-representations (see above quote). Thus, phenomenal awareness ofstimulus patterns deriving from both the external world, the body's surface, andthe body's internal milieu can be integrated to and, therefore, reorganized asmeaningful information for conceptual and symbolic consciousness, as outlinedin Figure 6.1. Furthermore, awareness is 'the result of the system's capacity toactively generate representations and meta-representations' (1991: 255) suchthat self-referential and meta-cognitive activities leading to the formation of an

Figure 6.1: Reentrant loop of conscious experience, showing relationship between selfand object of experience

Efferent outputs transduced as action ('attending to')

Pick-up of environmental information transduced as afferent inputs

'In here' =L+l: EXPERIENCER-IN-HEREattends toREPRESENTATION OFSTIMULUS PATTERNL: GLOBAL CELLASSEMBLYL-l: WEIGHTEDSYNAPTICDISCHARGES

'Out there' =ENVIRONMENTALINFORMATION ABOUTINNER AND OUTEREVENTS


inner perspective will emerge (1991: 258). It will be recalled at this point thatphenomenal awareness ultimately derives from the patterned activation andmodulation of receptor cells pertaining to the various perceptual systems. It issuch patterns of stimulus information, derived from inner and outer environmental sources, which induce the strengthening or weakening of synapticdischarges on level L-l of the proposed three-level hierarchy in terms of which Iinterpreted above Flohr's proposals concerning phenomenal awareness. Flohr isnot concerned, in the article I have discussed here, with conceptual andsymbolic forms of consciousness. Phenomenal awareness is always awareness ofan organized field of experience of objects, events, and so on, both within ourbodies and in the world outside our bodies. This draws attention to the fact thatthe afferent stimulus inputs derive from sources outside the cerebral cortex. As

I see it, Flohr's notion of phenomenal awareness can be described as 'consciousexperience of our experiencing', i.e. 'before' meaning in the socially conventional sense necessarily enters the picture. Crudely, this can be interpreted as areentrant loop whereby the conscious experiencer, in the form of a higher-scalarme ta-representation of itself as the one who has the phenomenal awareness orundergoes the conscious experience of the given phenomenon, is linked to the'object' of this experience roughly along the lines shown in Figure 6.1.

The purpose of Figure 6.1 is to show how conscious awareness of the kindtheorized by Flohr is always embedded in a higher-scalar information-richenvironment which constitutes the boundary conditions of phenomenalawareness itself.

3. Brain, Meaning, and Symbolic Consciousness

Proto-meanings are made up of neural activities in the form of neurophysiological entities called impulses which enable communication to take place betweenneurons by means of bio-chemically based neurotransmitters. The protomeanings that are stored in the brain combine with brain waves from other brainregions where phonological systems and structures are stored. The brain wavescorresponding to the latter should not be confused with the actual sounds thatare produced by the vocal-tract activity in phonation. Peng (2000: 194) pointsout that some distance has to be travelled, 'because of the cortico-striato-pallidothalamo-cortical-loop', before the brain waves corresponding to acoustic imagesare transduced into physical sounds. Peng points out that neural impulsescorresponding to proto-meanings are 'catalytically mapped onto, that is, boundto, an acoustic image' (2000: 195) such that 'in production, the impulse can nolonger be amorphous' (Peng 2000: 195).

It follows, from the way in which neural impulses that are analogous to thecontent and expression planes in the CNS are catalytically mapped onto eachother only to be separated prior to articulation, that linguistic content as such isnot transmitted from speaker to hearer. The only thing which is 'transmitted' isthe physical speech sounds produced in vocal-tract gestural activity andpropagated through the air as sound waves. Only the neural impulses corresponding to the expression stratum are transmitted via the loop to the vocal apparatus


for speech production. The impulses corresponding to the meaning - thecontent - remain, as Peng (2000: 195) shows, in the speaker's brain. Meanings donot travel through the air to the hearer. Instead, the hearer must re-construct inhis or her own brain the relationship between the two strata by (1) recognizingthe speech sounds which are picked up by the auditory system; and (2)combining Of binding in his or her brain the neural impulses deriving from thebrain regions concerned with proto-meaning (content) and expression. Thehearer performs this process of re-construction on the basis of the fact that he orshe has a language system stored in long-term memory which is sufficiently similarto that of the speaker that the same kinds of patterned relations between neuralimpulses associated with expression and those associated with content are made.This presumes that the neural and physiological dynamics of the central andperipheral nervous systems of speaker and hearer have been entrained to thesame higher-scalar ecosocial system and its meanings even if their individualtrajectories through this can never be identical in every respect.

In the previous section, I considered how Flohr's proposals represent animportant statement concerning the way in which afferent stimulus patternsderiving from multimodal sources either within or outside the body are represented as conscious experience of phenomena by the brain. The reorganizationacross scales of lower-level physical brain processes led to their translation intohigher-scalar representations of input stimulus patterns which have meaning forthe self who is the conscious observer of these from an inner perspective, itselfthe result of higher-scalar meta-representations in the brain. The states ofphenomenal awareness which form the basis of Flohr's discussion are largelyperceptual in character, whether these derive from external environmentalevents or from within the internal milieu of the body. However, this distinctiondoes not alter the fact that afferent inputs derive from physical stimulus information which is reorganized as conscious experience of phenomena in the waydiscussed in the previous section. Furthermore, the individual's experience ofthe given phenomenon is always mediated by a system of interpretance in termsof which the phenomena of which one is aware are experienced as phenomenacorresponding to this or that category of experience. This shows that consciousness is semiotically mediated at all levels. It is about a meaningful relationbetween conscious experiencer and phenomenon experienced. In the case ofphenomenal awareness, as discussed by Flohr, it is important to point out that'meaning' should be taken to refer to something less specified and thereforemore general than the more specified symbolic categories which are found inecosocial semiotic systems. The question now arises as to how Flohr's discussionmay help us to understand symbolic or higher-order consciousness and the roleof symbolic modes of semiosis such as language in this.

4. The Embodiment of the Material and the Conscious Modes inExpression and Content

Flohr's discussion of phenomenal awareness highlights the way in which stimuluspatterns, derived from acoustic, photic, and other forms of energy, stimulate and


modulate the receptor cells such that nerves are stimulated to fire and transmitsensory information to the brain in the form of nerve firings. In this way, thequalia of phenomenal awareness - i.e., the elementary conscious experiences ofbasic sensations such as the colour blue, a sweet taste, or low pitch - areregistered on consciousness. This is sensory perception. But this is not, as Gibson(1983 [1966], 1986 [1979]) has shown, what perception is normally all about.Perception is of events which occur in an environmental context which theorganism actively and adaptively explores and orients to. When we listen tospeech sounds, we do not perceive raw acoustic sensations such as low pitch.Instead, we perceive a speech event. In other words, we perceive the syllables,rhythmic patterns, and intonation contours characteristic of speech events in aparticular language. To be sure, our perception of this event is supported byacoustic energy in which many fine-grained distinctions not normally available toconsciousness can be identified by physical measuring devices such as the spectrogram. The flow of energy measured by the spectrogram is continuouslyvarying at the very short timescale of sensory perception. Many of these distinctions are not salient for the perception of the speech event and are filtered outby the particular phonological system in operation (see chapter 3, section 2, pp.109-16, for a full discussion). Event perception is oriented to the pick-up ofinformation about the meaning which the auditory event affords the listener. Inthe case of speech events, we perceive information about (1) the processes ofvocal-tract articulation of the speaker who is the source and producer of theevent and (2) the ways in which the phonological units and relations that areconstrued in the sound stream qua expression stratum may be cross-coupled withlexicogrammatical and semantic units and relations on the content stratum. It isimportant to remember here that content-stratum systems and structures areneither contained in nor transmitted by the expression stratum. I shall comeback to this point. What I want to emphasize here is that all forms of semiosis,e.g. the act of speaking under consideration here, entail the perception ofsemiotic events and/or products (object-texts) qua expression stratum. Theperception of the speech event as expression stratum means that listeners orientto and perceive phonological categories and relations on different scales oforganization in the speech event (see chapter 2, section 9, pp. 83-6). This is notsaying anything at all new. Nevertheless, it is important to reassert thisfundamental point so that we can address the relevance of this to symbolic consciousness. The main point is that the expression stratum is itself a phenomenonof conscious awareness. Language users do not perceive raw acoustic sensations.Instead, speech sounds are phenomena of experience which are organized andperceived as instantiations of the phonological categories of a particularlanguage system. The perception of a speech event therefore involves the introjection into the domain of the conscious of the vocal-auditory information whichis picked up as the embodied experience of vocal-tract or other gestural activity.Halliday has discussed the overall semogenic, or meaning-creating, process inrelation to the two strata of expression and content as follows:

In the act of meaning, then, the two modes of experience [viz. the materialand the conscious, ~JT] through the projection of the one by the other,


become fused and transformed into something that is new and different fromeither. We can think of this as creating a 'plane of content' in the Hjelmslevian sense. If we look at this process dynamically, it is meaning-creating, orsemogenic. Ifwe look at it synoptically, as a relation construed by this process,it is semantic; and it appears as an interface (our original notion of semanticsas 'interlevel' was relevant here), one 'face' being the phenomena ofexperience. We often refer to these phenomena collectively as 'the material',as if the only form ofexperience was what is 'out there'. But this is misleading.Our experience is at once both material and conscious; and it is the contradiction between the material and the conscious that gives these phenomenatheir semogenic potential ...

But in order for meaning to be created there has also to be a secondinterface, a transformation back into the material, or (again, rather) into thephenomenal - this time in its manifestation in the meaning subject's ownbody: as physiological processes of articulation or gesture. This is thephonetic/kinetic interface; the 'expression plane', in Hjelmslev's terms. Sincethere can be no meaning without expression (meaning is intlnSubjectiveactivity, not subjective), the act is 'doubly articulated', in Martinet's terminology: it is the transduction of the phenomenal back into the phenomenalvia these two interfaces of content and expression. (Transduction nottransformation, because as Lamb (1964) pointed out many years ago intransformation the original is lost, ceases to exist. And again I am suggestingthat we should conceive of it as phenomenal rather than material, since boththe 'outer' faces, that of the content substance on the one hand and that ofthe expression substance on the other, embody both the material and theconscious modes of being.) (Halliday 1992a: 21-2; italics in original)

Both content and expression embody the material and the conscious modes.Many theories of meaning have lost sight of this material-phenomenaldimension. In doing so, they miss out on the fact that meaning-making requiresan intersubjectively shared form of bodily expression. Meaning-making is not thetransmission of an abstract content form in the form of coded ideas or thoughtsfrom one 'mind' to another. Expression is not a mere vehicle or carrier ofcontent, but is itself phenomenal. Meaning is not transmitted in this or any otherway.

The embodiment of both the material and the conscious modes may beexplained as follows. Speech sounds qua physical and biological processes arebased on topological-continuous variation. Phonological systems and structures,on the other hand, operate in terms of discrete typological-eategorial distinctions on various levels of organization. The physical and biological processestend to be confined to the organismic scale and its material interactions with itsimmediate environment. However, the phonological systems and structures of aparticular language system are ecosocial categories. Bodily processes of the firstkind are epigenetically entrained to and selected by the dynamics of the higherscalar ecosocial level by virtue of the individual's participation in the meaningmaking practices of that community. From this point of view, the body quaphonetic/kinetic interface, to use Halliday's term, is a cultural tool which


parUClpates in meaning-making practices that can extend across very manydiverse space-time scales. This is the principle of scalar heterogeneity asdeveloped by Lemke (2000a; see also Thibault 2000a). It is in this way that thecategories of symbolic consciousness and its artefacts ramify across and integratewith many different scales beyond that of the physical-material processes of thebody in interaction with its immediate environment.

Lexicogrammatical and discourse systems and structures on the contentstratum similarly operate in terms of typological-categorial distinctions for themost part. Again, the phenomena of experience in the material domain of thebody's inner milieu and the body's environment are topological-continuoussuch that the contradiction between the material and the conscious manifestsitself here as well. As I said above, these categories of content are not transmitted from one person's brain to another person's brain (e.g. speaker to listener),but are reconstructed in the individual's (the listener's) brain. The no-Iessembodied brain and its neural dynamics is also a tool which the individualactivates and puts to use for the purposes of social meaning-making. As discussedin section 7 below, pp. 300-1, this is possible because of the meta-redundancy orcontextualizing relations in operation. The cross-coupling of the two strata inthe brain is possible because of the contextualizing relations that specify whichexpressions combine with which contents, and in which contexts. Rather thanthe transmission of information or thoughts from one mind to another, I wouldexplain what happens in the following terms. It is the use in determinate socialmeaning-making practices of expression-stratum artefacts in the form of (1)bodily (gestural) activity which is projected into and/or introjected from theecosocial environment or (2) material textual artefacts, that provides an intersubjectively shared material basis whereby the two strata can be cross-coupled inindividual brains (see below) such that the transduction of the phenomena ofexperience back into the phenomenal can take place. Individuals makemeanings by harnessing and entraining their body-brain dynamics to ecosocialones. Intersubjective vectors of expression integrate them both materially andsemiotically to this common purpose.

Figure 6.2 provides a schematic representation of the relations betweenexpression and content with respect to the central nervous system (CNS) andthe peripheral nervous system (PNS).

Phonological units and relations have both indexical and symbolic properties,which means that they are more typologically specified than are the topologicalcontinuous differentiations characteristic of the sensory perception of theacoustic information in the speech event (see above). Moreover, the expressionstratum is an interface which connects the central nervous system to the externalenvironment of the individual via the peripheral nervous system (see above).It is an interface in which sensori-motor routines have been entrained to andrecategorized as the phonological routines of a particular language system.The more specified nature of phonological expression vis-a-vis less specifiedsensori-motor routines on lower integrative levels means that individuals are incontact with a greatly expanded Umwelt. Speech events which are picked up andattended to by a listener impinge on the listener as more than just indexes ofnecessary features of their correlated environmental events. In this way, the


Figure 6.2: Content and expression and their relations to the central and peripheralnervous systems; adapted from Peng (1994: 105)

Context of Culture

Context of Situation

CNS Discourse CONTENT STRATUM

Lexicogrammar

Phonology, Graphology

EXPRESSION STRATUM

PNS Phonology, Graphology

speech event indexes features of the articulatory processes which produce it. Tobe sure, the logic of the specification hierarchy shows that speech events haveiconic, indexical, and symbolic properties. This aspect is discussed in chapter 3,section 4, pp. 118-20.

However, the main point to be emphasized for now is that the transduction ofthe speech sounds into nerve impulses which are transmitted to the centralnervous system via the bulbo-cortical pathways results in their cross-coupling tolexicogrammatical and discourse levels of organization on the content stratumin the listener's brain. This cross-coupling is possible because the stimuluspattern which enters the cerebral cortex consists of a stable correlation ofarticulatory, perceptual, and visual inputs which are reentrantly mapped at pro-gressively higher levels of abstraction to form a predictable language-specificphonological category. Coherent assemblies of cells are formed in the brain soas to represent a given phonological category. It is important to bear in mindhere that expression-stratum categories such as phonological ones are alsostored in the cerebral cortex. They do not belong to the peripheral nervoussystems of sensori-motor activity per se. The presence of stable repertoires oflanguage-specific phonological categories in the cerebral cortex means thatstable assemblies of neuronal populations are epigenetically selected andstrengthened in both logogenesis and ontogenesis as representations of phono-logical categories in the CNS.

As we saw in section 2, neuronal nets with a high rate of assembly formationdo not only produce complex, qualitatively different representations; they alsoproduce progressively higher-level meta-representations. It is in this way, Isuggest, that lower-level phonological repertoires are cross-coupled in thecerebral cortex with higher-level lexicogrammatical and discourse units andrelations on the content stratum. Expression and content are cross-coupled inthis way on the basis of the principle of semiotic alternation developed by Lemke(1999; see in the present volume chapter 1, section 4, pp. 18-22). Thus, the cell


assembly (the neuronal repertoire) which represents a particular phonologicalcategory contains information which enables it to be translated into contentstratum units and relations by virtue of the cross-coupling dynamics involved.This is possible because of the brain's capacity to recategorize sensori-motor andphonological routines as conceptual and lexicogrammatical ones, respectively(see chapter 3, section 9, pp. 139-46). The term 'cross-coupling' serves tohighlight the dynamic, time-bound nature of the processes involved. Expressionand content are said to be cross-coupled to each other because their respectiveprocesses simultaneously unfold in time and act on and affect each other. (In thefollowing section, I shall consider the selective and two-way nature of the crosscoupling dynamics involved in terms of Lernke's theory of meta-redundancy.)

5. Periodicity and the Intrinsic Temporal Organization of the ExpressionStratum of Speech on Diverse Scalar Levels

Speech events are characterized by periodicity patterns at different levels of hierarchical organization on the expression stratum of phonological organization.Temporality is, therefore, intrinsic to the structural organization of such events.This means that time cannot be viewed as an abstract constant which is externallyapplied to such events. The timescale on which speech events are perceived assuch is relative rather than absolute. Absolute or clock time only applies to theenergy flows measured by the spectrogram. Relative time simply means that theduration of a given event is compared to that ofother events rather than to someabsolute external standard. Relative time is context-dependent and intrinsic tothe rhythms of the organism in relation to his or her environment. Phonology isemergent from vocal-tract gestural activity, which is essentially neuromuscular incharacter. These (and other) gestures are comprised of cycles of periodicprocesses on various levels of organization.

In systemic-functional phonological theory, for example, three vocal-tractcycles have been proposed (Halliday 1992b, 2000; Cleirigh 1998: 37-40). Thethree cycles correspond to the phonological ranks of syllable, foot, and tonegroup in systemic theory. The first cycle involves the processes of opening andclosing the vocal tract through articulators so as dampen phonation in theproduction of syllables. This cycle can be modelled as the oscillation betweendifferent energy states of the neuromuscular systems involved (Cleirigh 1998:38). In the syllable, the constituents of the Rhyme have the potential to bemoraic, i.e. to function as a timing unit, whereas the constituents of the Onsetare non-moraic. On this scale, the articulatory cycle consists of activating andde-activating syllable weight or moraicity. On the next-higher scale of activity,vocal-tract activity is associated with variations in loudness and duration so as toproduce stress. This leads to the production of a rhythmic cycle, which iscomprised of activating and de-activating stress. The third cycle is thatconcerned with major pitch shifts so as to create tonic prominence or tonicity.The processes involved on this scalar level produce intonation through acontinual process of activating and de-activating tonicity.

The temporal structure of the speech event is based on periodicities at


different scales of organization which provide internally predictable principlesof time measurement based on the principle of regularity. Moreover, theseperiodicity patterns and their intrinsic temporal organization are inherent inphonological organization. On their respective scalar levels, both in- and out-ofphase properties of moraicity, rhythm, and tonicity are semiotically salienttemporal events, Le. they are temporal differences which make a difference. It ismy view that an important function of these events is to entrain the listener tothe perception of the vocal-auditory event as meaningful (see above). It is on thebasis of such entraining that the always time-bound cross-coupling processesbetween expression and content takes place. Periodicity patterning in speechand other semiotic modalities entrains the listener (and the speaker) to semiotically salient temporal structures in the overall temporal flux. This temporalstructure consists of smaller units nested within larger units from the syllable, say,to a whole discourse event, or even to diverse discourse events on differentoccasions. Moreover, the fact that time-bound phonological periodicities alsointegrate with and demarcate lexicogrammatical and discourse levels of organization on the content stratum highlights their role in the cross-couplingdynamics referred to above.

Consistent with Gibson's ecological theory of perception, the periodicities ofsemiosis and their pick-up qua semiotically salient environmental events are commensurate with the temporal structures which the human organism is adaptedto perceive on the scale on which meaning-making takes place. In this way,bodily(neuromuscular) periodicities are entrained to higher-scalar ecosocialmeanings. On the other hand, external clock time has scales from attosecondsto, say, 10100 years hence of the Dark Era which cosmologists predict as theultimate fate of our expanding universe, when little more than waste photons,neutrinos, electrons, and positrons will remain. This refers to the external andabsolute measurement of time by extrinsic measuring devices that I mentionedabove. But that is not the temporal scale to which the human organism is biologically adapted in its ecosocial environment. The temporal scale of meaningmaking and perception is intrinsic to the organism in its environment; it is alsorelative, in that one event is measured through comparison with other suchevents rather than against some absolute standard external to these events.

6. The Creation of Symbolic Objects of Consciousness in Semantic NeuralSpace

Flohr also draws attention to Hebb's notion that 'plastic neural nets can functionas coincidence detectors, and that the eo-activation of certain neuron populations by stimulus patterns creates a representation of these patterns' (1991: 258).In this way, intentional contents can be the cause of physical brain processes, e.g.representations can activate stimulus patterns. This is coherent with the two-wayand reciprocal nature of the cross-coupling dynamics whereby expression andcontent selectively contextualize one another. It is interesting in this regard tonote that inner speech is associated with imagined auditory routines which aresilent and which direct the temporal flow of inner dialogue. These imagined


routines are cognitive routines realized by the activation of neuronal populations. They do not depend on the production of audible speech sounds, thoughthey also direct these in the way discussed above. My reason for bringing thispoint up is two-fold.

First, it demonstrates the independent existence of both phonological,graphological, sign, and other language-specific expression-stratum units andrelations in the central nervous system. This does not change the fact that thesehave to be epigenetically selected in ontogenesis and experientially strengthened in logogenesis in order to get into the CNS in the first place. Secondly, theexistence of both semiotic strata - expression and content - together withthe reciprocal nature of the contextualizing relations which cross-link them inthe CNS further suggests that the contents of symbolic neural space can 'cause'physical-brain and sensori-motor processes by entraining these lower-levelrelations along a specific action trajectory which loops from the organism outinto its ecosocial environment (see chapter 1, section 1, pp. 3-8). The objects ofconscious attention are capable of eliciting action trajectories with respect tothat 'object', whether this is a material object in the immediate perceptual fieldof awareness of the individual or a symbolic object which has been generated bythe cross-coupling of expression and content in symbolic neural space. Flohralso makes the point that objects selected for conscious attention 'are alsodifferent subjectively' and that 'this occurs in those domains in which theassembly turnover is enhanced by thalamo-cortical activation' (1991: 257).

But it is time to rein in this discussion and come to some kind of statement asto what objects of symbolic consciousness are. Whether in the CNS of theindividual in the form of silent inner speech, or in the form of the perceptualpick-up of stimulus information from speech events along with their transduction into nerve impulses which are transmitted to the cerebral cortex of thelistener, the creation of objects of symbolic consciousness arises through thecross-coupling of neuronal populations corresponding to expression- andcontent-strata repertoires and routines in the brain of the individual. To take asimple example, the lexicogrammatical resources of clause grammar experientially construe a semantic figure comprising a configuration of a Process and Participant(s). The semantic figure so construed is just such an object of symbolicconsciousness which can be attended to and focused on in symbolic neural spacefrom the inner perspective of the self who so attends to it. This applies equallyto all modalities of natural language - speech, writing, sign, Braille, and innerspeech - irrespective of the specific means of transmission from peripheral tocentral nervous system. The example just given refers to the content stratum,though expression-stratum entities can also be attended to in the same way.

A clause, say, is never just an experiential construal of some phenomenon ofexperience. It is also a way of acting on it and of intersubjectively orienting to it.This is the interpersonal dimension of meaning. Symbolic consciousnesstherefore means attending to its objects in both ways, simultaneously. Now, theobjects so attended to - i.e. experientially construed and interpersonally enacted- are phenomena of experience which have been categorized by the symbolicresources of whatever semiotic system is in operation. These phenomena ofexperience may be actually perceived material objects, as I said before, or they


may be purely imaginary or hypothetical ones which can be imagined insymbolic neural space. They can also be recorded in the textual records andproducts whereby such objects can be re-constructed in different brains alongthe lines I sketched out above. In both cases, the individual's neural dynamicsmust cross-couple expression and content to produce the appropriate symbolicrepresentation in his or brain. Consider the following example:

(1) look at the mess Paul (2) you're gonna have to clean up in here(3) it's a pig sty

The three clauses which comprise this speaker's utterance are addressed to herhusband (Paul). In this example, the 'mess' referred to in clause 1 is taken to begiven in the perceptual purview of both of the participants in the exchange,which occurs in the living room of their home. However, there is an importantdifference between the perception of some phenomenon and its linguisticconstrual as an object of symbolic consciousness, as this example makes clear. Toillustrate this point, I shall consider in some detail the first clause in the abovepassage. The metafunctional analysis of this clause is presented in Table 6.1.

Table 6.1: Metafunctional analysis of the imperative clause (you) look at the mess

Metafunction (you) look at the mess Paul

Experiential Behaver Process: Range:Behavioural affected

Interpersonal Imperative: Vocativeproposalfor action

Textual Theme/Given Rheme/New

The metafunctional analysis of the clause look at the mess reveals different facetsof its functioning in simultaneously construing and enacting a symbolic act ofconsciousness. The experiential analysis shows that the speaker of the clauseconstrues the given phenomenon of experience as a semantic figure. In thiscase, it is categorized as a behavioural process of the subtype concerned withconscious action (Halliday 1994 [1985]: 138-40; Halliday and Matthiessen 1999:136). The clause analyses the overall semantic figure as comprising threesemantic elements: Participant: Behaver"Process: Behavioural: Conscious Action11 Participant: Range. The semantic category of behavioural process often refersto processes that are on the borderline between inner consciousness and outeractions and events in the material world. This is so in the sense that theseprocesses do not occur in the external world of material actions and events, butare exterior actions of the body which, however, manifest or index consciousintent on the part of the person who carries out the behaviour.


In this case, the process 'look at' construes a conscious act of visuallyattending to something. The Behaver is the participant who performs thisprocess and the Range is that which is so attended to. In construing her interlocutor in this way, the speaker of the clause integrates him into a particularexperiential frame of reference, as supplied by the experiential category whichthe clause instantiates. In this way, he is experientially construed as a consciousparticipant who is required to visually attend to the designated Range. Now, thenominal group the mess, which is used to construe a particular referent situationwithin the perceptual purview of the two participants, also functions tocategorize some facet of the situation as an instance of the semantic category ofThing designated as 'mess' by the Thing element in the nominal group. Thepoint is that 'mess' is a semantic category which belongs to the English language.It is not inherent in the referent situation as such. In using this particularselection, the speaker creates a symbolic object of consciousness which in itselfcan be attended to as a way of understanding the given referent situation. As Isaid before, the mess is Range in its clause such that the particular experientialframe chosen here integrates the implicit 'you' as the Behaver who attends tothis symbolic object of consciousness in the way discussed above.

The observations made in the preceding paragraph go hand in hand with theinterpersonal semantics of the clause. Thus, the selection of imperative moodserves to modalize the addressee as the one who is required to carry out the actionspecified in the clause. The addressee is the implied grammatical Subject, and assuch is selected as the entity in which the speaker invests a particular interpersonal take. In this example, it takes the form of a proposed action which she seeksto negotiate with her interlocutor (her husband, Paul) as the one who isdesignated as being responsible for carrying it out. In this way, the interpersonalmeaning of the clause integrates or attempts to negotiate the integration of theaddressee to the same evaluative stance on the proposed action. The selection ofimperative mood therefore enacts a particular orientational stance towards theaction. Moreover, the noun mess implicates a particular evaluative or attitudinalstance on the part of the speaker. Once again, this constitutes an attempt tointegrate the addressee to the same stance. That is, she seeks to direct attentionto the phenomenon of concern from a particular interested point of view inwhich a negative evaluation is implicit. Both of these interpersonal facets of theclause's overall meaning also implicate the speaker's ability, in proposing thecourse of action in question and in taking up the value stance that she does, toput herself in the position of the other (the addressee) who is being asked toadopt the course of action and therefore to accept its associated evaluative stance.This can only happen through the resources of symbolic meaning-making.

The textual organization of the clause in terms of Theme-Rheme andGiven-New shows (1) that 'you' as Theme is the local point of departure for thedevelopment of a message about the Theme in the Rheme and (2) 'you' is takenas the Given in relation to which the New is presented. In this particularexample, this reflects the way in which the speaker presumes the 'you' as thecontextually given item in relation to which the new contextual information ispresented. The textual organization of the clause structures the message intoitems of consciousness of varying degrees of interest or significance. In the


present case, the 'you' whose consciousness is being acted on is being asked tofocus on the New information, in particular the mess, which occurs in finalposition in this clause.

I have chosen this particular example because it demonstrates a link betweena perceptual purview at the same time as bringing out the radical differencebetween symbolic consciousness and consciousness which is tied to theperceptual awareness of phenomena. Perceptual awareness is tied to the pick-upof environmental information which impinges on the organism as indexes ofnecessary correlations with environmental events. As Gibson has shown,perception serves to orient and guide the behaviour of the organism in itsecological environment relative to the environmental events that are perceived.But the meaning-making resources of symbolic consciousness enable individualsto select from a symbolic system of possibilities which is not inexorably tied to thehere-now perceptual purview by indexical necessities. Thus, speakers of aparticular language can make choices from the resources of a linguistic system,for instance, such that they can create varying possibilities of experiential categorization, varying interpersonal and evaluative takes, and varying ways forpresenting and organizing information in discourse such that a given item ratherthan some other is presented as, say, theme in the clause. Moreover, the metafunctional hypothesis which I have assumed here shows that all these diverseoperations occur simultaneously on account of the way in which a diversity ofsemantic functions is mapped onto the one lexicogrammatical form.

7. Contextualizing Relations, the Principle of Meta-redundancy, and theBrain as ContextuaIizing Tool

The cross-coupling of expression and content in the individual's neuraldynamics depends on higher-scalar systems of contextualizing relations whichfunction as a system of interpretance within the brain. One candidate forexplaining the abstract and general principles which cross-link the two domainsis that of meta-redundancy, developed by Lemke (1984c) on the basis of someearlier proposals by Bateson (1973b). Bateson developed the theory ofredundancy in order to show how increasingly higher orders of contextualization can be modelled and related to each other as a hierarchy of logical types.All semiotic systems are based on the principle of meta-redundancy. Thisprinciple is, fundamentally, a contextual one. In a given semiotic system, we cansay that the semantic contents (linguistic or non-linguistic, it does not matter forthe principle is the same) a, b, c are realized by the expressions x, y, z. Thismeans that there is a redundancy relation between the terms on either side ofthe content/expression divide. As Lemke (1984c: 35) points out, (meta-)redundancy relations are two-wayor symmetrical ones in the sense that two partsof the same message are redundant with each other in the sense that the information carried by one makes it possible to predict some or all of the informationcarried by the other. Halliday (1992a: 2) has aptly illustrated this by showing thetwo-way, or dialectical, nature of realization, i.e. the relationship which is expressible by the term 'redounds with' is implicated in the two-way nature of the real-


izatory relationship between expression and content. For example, expression'realizes' content and content 'is realized by' expression. Thus, given themeaning a, we can predict sound, gesture, graphological feature x, and givenfeature x, we can predict meaning a. (For further discussion of these principlessee also Thibault 1991a: chapter 4.)

The individual's ability to predict, with a more than even chance of beingcorrect, the nature of the redundancy relations involved is always a probabilistic,or stochastic, process. Such first-order redundancy relations are, in turn,dependent on some still-higher-order contextual, or me ta-redundancy,relation(s). This can be explained as follows. The expressions x, y, z realize themeanings a, b, c, while the meanings a, b, c realize the still-higher-ordercontextual relations p, q, r. Now this relationship, as expressed by the concept ofredundancy, is not a linear or causal one, as Halliday (1992a: 24) points out.Thus, it is not the case that 'a realizes p' or that 'p is realized by a' (see chapter1, section 6, pp. 26-30). Rather, there is a meta-redundancy relation such that p,q, r does not redound with a, b, c, but with the redundancy of a, b, c with x, y, z.Following Bateson, the slash formalism, i.e. 'I', is used to indicate the orders ofthe redundancy relations involved. The single slash, the double-slash, and so on,indicate the various orders of contextualizing or meta-redundancy relations, atincreasingly higher orders of abstraction, as in our simplified example. Thus:x, y, z I a, b, c I I p, q, r.

Given that there is no directionality or causality implied, this may be readeither way, according to the specific analytical or situational requirements andcontingencies. In other words, p, q, r redounds with the redundancy of a, b, cwith x, y, z; or, x, y, z redounds with the redundancy of a, b, c with P: q, r. Eachof the three orders or levels of relations in our example may correspond to thedifferent levels of organization (strata) in some semiotic system. One suchexample would be the tri-stratal system of language, comprising phonologyIgraphology, lexicogrammar, and semantics. There is neither directionality norcausality because redundancy is a principle which implies no separation of thelevels involved into separate ontological domains. Instead, it expresses howpatterns of relations emerge and are construed in and through the hierarchiesof contextualizing relations involved.

In Flohr's description of 'activity-dependent self-organization of neural nets',the same general principle of redundancy operates. I have attempted to approximate the meta-redundancy relations as follows:

stimulus pattern I assembly formation I I representation ofspatio-temporal pattern

The generation of increasingly higher orders of abstraction from the 'stimuluspattern' may have important consequences for theories of human development.The question then arises as to how such patterns of neural activity in the brainmap onto emergent semiotic activity. This also raises the question as to how suchhigher-order social semiotic processes and dynamics act on and shape brainfunction and activity. This requires a theory of semiosis and the place of thebrain in such a theory which is truly dialogic in character (see Thibault 2000a).


8. A Definition of Consciousness in Terms of the Semiosis that OccursWithin the Brain

Consciousness of phenomenal states, or conscious experience of phenomena,does not require either mind-brain dualism or physical reductionism for itsexplanation. Both conscious experience and self-consciousness awareness orknowing that one's self is undergoing that experience are emergent propertiesof higher-scalar global patterns of neural organization that realize semanticproperties. Furthermore, emergent semantic properties which are realized byglobal patterns of neural organization constrain both the possibilities forconscious experience of phenomena as well as the possibilities for action inresponse to these phenomena by entraining lower-scalar sensori-motor dynamicsto their semantics (juarrerc 1999: 172). In this way, our conscious experiences ofphenomena and our responses to them are semantically constrained by the selforganizing properties of the higher-scalar system of interpretance in the brain.

Consciousness cannot be reduced to or viewed as identical with physical brainprocesses. It was Descartes who first understood that consciousness is notfounded on the opposition between raw sensations and thoughts. Descartes'sdualistic solution was to postulate that mind and brain are entirely separateentities. However, Seager (1999: chapter 1) has cogently argued that, withoutaccepting Cartesian dualism, which is clearly false, there is a more important andpositive lesson to be learned from Descartes's philosophy of mind. That is, consciousness is linked to 'ideas', which are basically representational. Therefore,every conscious experience is an experience of a representation ofan idea in themind. I shall argue below that this essential insight about the nature ofconsciousness can be entirely reconstituted within the ecosocial semioticconceptual framework of the present study. In Descartes's view, neural firings, inmodern terminology, realize what Descartes called ideas. Ideas can be about theoutside world as perceived through our perceptual systems, our inner and outerbodies, as well as imagined ideas elaborated by the mind. Consciousness, in thisview, is consciousness of the conceptual representations - the ideas - elaboratedby the mind. We have conscious experiences of such representations and not thephysical brain processes which realize these. Mind, in this view, is a system ofmeanings which are realized by, though not reducible to physical brainprocesses. I am using the term 'meaning' here in an extended sense to cover allsemiotic processes that are encompassed by the semiotic hierarchy of iconic,indexical, and symbolic meanings. In this view, conscious experience encompasses meanings based on sensori-motor activity, pre-linguistic conceptualroutines, and the higher-scalar symbolic resources typically associated withculture (e.g. language, depiction, and so on).

Seager (1999: 188-9) points out the conceptual character of our consciousness of phenomena. That is, our consciousness of the perceived world, forinstance, is always consciousness of bicycles, people, clouds, trees, and so on.Consciousness is perfused by the perceptual and conceptual categorizationscharacteristic of Edelman's primary consciousness. The interaction betweenmemory and the neuronal groups that constitute the systems of perceptual andconceptual categorizations within the brain means that the categorizing activity


of global mappings of neurons constitutes a system of interpretance wherebyconscious experience is always consciousness of something as something. InSeager's (1999: 184) terms, we are always conscious of something in the sensethat the intentional object of our consciousness is an object which is presentedto our consciousness 'under a certain aspect'. This property of entering consciousness under a certain aspect is further extended, refined, and specified bythe systems of linguistic categorization of a given language system such that wecan, for example, experientially construe a given phenomenon of experience asan instance of a particular experiential category. These phenomena ofexperience do not have to be given to perception, but may be remembered,imagined, or hypothesized phenomena which are symbolically construed asintentional objects that can be made available to consciousness.

Consciousness is semiotic interpretation of phenomena. Descartes's 'ideas'are meanings, however elementary, which cannot be reduced to physical interaction between organism and some physical stimulus such as the light waveswhich register on the retina of the eye or the sound waves propagated from asource which register on the auditory nerves. At the lowest-scalar level ofperceptual activity, Gibson (1983 [1966]: 40-1) has identified the receptor cellswhich respond, for example, to mechanical, chemical or light energy. The interaction between the organism's receptor cells and the energy responded to isphysical interaction. But Gibson identifies a scalar hierarchy, comprisingreceptors, organs, and perceptual systems, which is involved in the perception ofsome environmental event. Organs, Gibson (1983 [1966]: 41) writes, are'functional, not anatomical, units' in a hierarchy of units and functions assummarized in Table 0.2.

Gibson's point is that organs and perceptual systems are higher-scalar levelswhich mediate the lower-scalar level of the physical interaction between receptorcells and the energy which is applied to these. Organs and perceptual systems areadaptive; they are not passive mechanisms which simply receive a physicalstimulus. Rather, they actively and adaptively respond to and modulate information about their environment. In this way, the organism as a whole, rather thanthe single receptor cell, responds to environmental information as signs of environmental events. That is, the organism is responding on a higher-scalar levelthan the physical stimulus which is applied to the receptor cells. Organs,perceptual systems and the cross-modal information from these which isreentrantly mapped in the brain function as a higher-scalar system of interpretance. Lower-scalar physical stimuli are reinterpreted at higher levels as (iconicand indexical) signs of events both in the external world and within our bodieswhich are available to conscious experience. Conscious experience is, then,intrinsically semiotic in nature because it involves interpreting informationwhich is picked up by our perceptual systems as signs of experiencedphenomena in relation to which the organism can organize a required responseat a higher-scalar level.

The processes involved here are physical processes involving interactionsbetween receptors, organs, perceptual systems, the CNS, and stimulus energy inthe environment. Yet, consciousness is not reducible to physical processes per se.After all, we have no conscious awareness or experience of the neural processes


Table 6.2: Scalar hierarchy ofanatomical structures ofperception and relativefunctions; summarized from Gibson (1983 [1966J: 40-3)

Anatomical Unit

Receptors: cells andunits

Organs

Systems (of Organs)

FWlction

Microscopic structures, e.g. the photosensitive rods andcones of the retina, usually in groups of such cells.Receptors are connected by afferent neurons to the CNS.They register the sequence or arrangement of stimulation,that is, information, not the energy of stimulation.

An organ of sensitivity is a structure containing manyreceptors that can adjust so as to modify the input fromcells and units. An organ has muscles and these aresupplied with efferent fibres from the CNS. Organs of sensitivityexist in a hierarchy: lower organs are subordinatedto higher ones.

A system has very many levels of function ranging fromlower-level ones such as the single eye with adjustable lensfor sharpening the retinal image and the pupil foradjusting intensity; the eye with attached muscles is a stillhigher-order system, stabilized in the head relative to theenvironment and with the help of the inner ear, it canscan the environment; the two eyes together make a dualsystem of still-higher-order, such that they converge fornear objects and diverge for far ones; the two-eyes-withhead-and-body system, in cooperation with posturalequilibrium and locomotion, for moving around inenvironment and looking at things.

which support, sustain, and realize conscious experiences in the brain. Wehave no conscious experience of the stimulus energy to which the receptorcells respond. Instead, conscious experience of phenomena involves a reinterpretation of information across scalar levels (Lemke 1999). Thus, themicroscopic energy fluctuations to which the receptors respond are re-interpreted on higher-scalar levels as potentially meaningful information about(inner or outer) environmental events. Furthermore, such information constitutes the basis for a variety of responses or action repertoires in the sense thatthe same information can be variably interpreted and responded to according todiffering contextual factors at higher scales. In the case of symbolic interpretation, which is mediated by a system of interpretance on the ecosocial scale wheresocio-cultural practices enter the picture, information from lower levels is constrained by social conventions and practices emanating from the ecosocial scale.Conscious experience of perceptual and other phenomena involves a semioticprocess of interpretation by integrating contextual factors on different scales.There is nothing deterministic about the processes involved. Instead, there arecomplex inter-level relations among factors on many different scales whichcombine to produce emergent higher-order conscious experiences which can beused by the organism in adaptive ways.


Descartes's view that nerve actions yield ideas in the mind and that it is theseideas which enter consciousness can be understood as meanings which are theemergent result of the inter-level interactions across many different scalar levels.For instance, firings in neural networks are reorganized (re-interpreted) as moreglobal cortical effects. These effects are, in their turn, reorganized as higherscalar sensori-motor categories such as a perceptual experience of somephenomenon or the phonological categories of a given language. Thesecategories - perceptual or phonological - in turn constitute an attractor spacewhich guides and monitors motor action along a trajectory. The co-ordinatedaction produced by vocal-tract and other gestural activity in response to environmental events is an example of this. In turn, gestural activity is re-interpretedat the ecosocial scale as symbolic activity in and through the categories derivingfrom the paradigmatic contrast sets of a given system of interpretance.

In Peircean terms, environmental information is a representamen, R.However, the R does not directly determine its object, X, which is a given environmental event. R is a sign of X according to the way it is mediated by a higherscalar SI. That is, R is a sign of X - an interpretant sign - according to the way inwhich the SI in operation interprets Rs. Is R food, therefore reason for attack, ora predator, therefore a reason for flight? Such interpretations will depend on theappropriate higher-scalar system of interpretance - the system of perceptionaction categories stored in the organism, though ultimately derived from itstransactions with its higher-scalar ecosocial environment. The ecosocial environment (semiotically) mediates lower-scalar signs such that signs of environmentalevents are appropriately interpreted and responded to. Descartes's 'idea' is thena representamen, R, about some object, X, i.e. some event in the world or withinthe body of the perceiver. The higher-scalar SI assigns an appropriate interpretation to the relationship between Rand X. Conscious experience ofphenomena is experience of grounded instances of categories of phenomena,rather than of pure sensations or a continuous flux of stimulus energy. When Ilook up from my computer while writing this to observe from my window, I seechildren cycling in the street below, trees in the garden, apartment buildings, apark on the other side of the street, the hills of Bologna in the distance, a coupleof cranes towering above a nearby construction site, an aeroplane taking off inthe distance, the planet Venus making its appearance in the early evening sky,and so on. I do not see an indistinct flux of visual sensations. Instead, the objectsof my perceptual experience - the objects of consciousness - are interpreted bya higher-order system of interpretance as instantiations of categories whichderive from and which are meaningful relative to this higher-scalar SI. In otherwords, meaning, though not necessarily symbolic or discursively organized (e.g.linguistic) meaning, constitutes the perceptual, the conceptual, and also thesymbolic categories in and through which consciousness is itself constituted.This observation accords, in some ways, with Descartes's claim that perception isjudgement. That is, seeing is judging in the sense that seeing is not a matter ofraw (visual) sensations impacting on the eye. Rather, perception involves interpretative judgements about the perceived phenomena which are construed asobjects of consciousness (Seager 1999: 5; see also this volume, chapter 4, section16). In our terms, it involves higher-scalar processes of interpretation which


assign meaning to lower-scalar phenomena according to the categories of the SIwhich is in operation.

The system of perceptual categories which is stored in the organism's brain isa higher-scale SI at level L+1 whereby the stimulus information which is pickedup by the perceptual systems on the lower level L is as Lemke (1999) points outdoubly construed. First, the stimulus information which is picked up andconstrued as a representamen is a given pattern of material (matter-energy)interactions at level L-l. Secondly, this pattern of material interactions isconstrued by the SI on level L+1 as having a meaning (an interpretant) which isan object of consciousness on the scalar level of the SI. That is, on the levelwhere phenomena are construed as objects of conscious experience andawareness. The material interactions on level L-l do not in themselves count asphenomena of experience. To be a phenomenon of conscious experience, it hasto be an interpreted phenomenon - one that is contextually closed by an irreversible act of amplification such as is made possible by the categories derivingfrom the higher-scalar SI on level L+l (see Wheeler 1988: 10). Only on this leveldoes the object have relevance as a perceptual or other object of consciousexperience. To be sure, when I look out of my window to view the scene beforeme, light energy registers on the receptors in the retina of my eyes. However, Iam not conscious of this light energy or the effect it has on my receptors.Instead, I have a conscious experience of children riding their bicycles in thestreet below, and so on (see above). The light that reaches the receptors in theretina of my eyes is reinterpreted as environmental information that indexes anenvironmental event, viz. children riding their bikes, ete. The stimulus information about the environment which is picked up by my perceptual systems ismeaningful for a given SI because it is a sign of some phenomenon ofexperience on the scale of the SI. This is the scale, for example, of the speciesspecific organism-environment interactions studied by Gibson (1986 [1979]:8-12).

Lemke (1999: 3) has re-interpreted Peirce's claim that the interpretant in itsturn becomes a representamen that 'determines some still further interpretant',which in Peirce's work constitutes the basis for the notion ofa chain of indefinitesemiosis, as the basis for an 'indefinite (open-ended) hierarchy of scale levels ofsystems of interpretance' (Lemke 1999: 3). Such an open-ended hierarchy canbe used to conceptualize the relations between conscious experience ofperceptual, conceptual, and symbolic categories of experience, respectively.Each of these three levels is describable in terms of the three-level hierarchy suchthat in each case the L+1 scalar level of the SI where the interpretant ofperceptual categories is located becomes, in turn, the representamen of thenext-highest triplicate of scalar levels. Figure 6.3 provides a schematization of theways in which consciousness of perceptual phenomena is integrated to consciousness of conceptual phenomena, which is integrated to consciousness ofsymbolic phenomena in accordance with the logic of specification-hierarchythinking.

Seager (1999: 6) has pointed out that we are not conscious of the physicalbrain states and processes which support and in some sense generate consciousexperience and awareness. Instead, we are conscious of the representations


Figure 6.3: open-ended hierarchy of semiotic triplicates of levels showing the integration ofperceptual, conceptual, and symbolic phenomena of conscious experience

Hierarchy of scalar levels,showing triplicates of levelscomprising systems ofperceptual, conceptual, andsymbolic interpretance

Symbolic-discursive

L+l SI comprising ecosocial semiotic systems of meaning-making resources and conventions where symbolic(semantic) interpretants are located as objects ofsymbolic consciousness

L Phonology/graphology and lexicogrammar

L-l Conceptual and sensori-motor routines

Conceptual-mental

L+l SI comprising system of conceptual categories whereobject of conceptual experience in the form of mentalimages elaborated in the brain on the basis of therecategorization of sensori-motor routines

L Sensori-motor routines

L-l Object of perceptual experience

Perceptual

L+l SI comprising perceptual categories and values whereobject of perceptual experience is located

L Stimulus information as sign of environmental event

L-l Light, mechanical, etc. energy modulating nerveimpulses in receptors

which physical brain processes produce. To take an example referred to bySeager, we are conscious of our experience of the colour red rather than theneural activity within our brain which generates this experience in our consciousness. In Seager's terms, which are adapted from Descartes's distinctionbetween formal and objective reality, we are conscious of what is representedrather than the formal means of its representation. Seager (1999: 175-6)considers this aspect of consciousness puzzling in comparison with all other representational systems in which we are simultaneously aware of both the formalmeans of representation and the represented content. In my view, this distinc-


tion between formal means and represented content is itself unnecessary whenseen in the light of the semiotic triplicates of levels discussed above. Consciousawareness therefore requires the connection of physical brain processes that arebelow the threshold of our awareness to meanings that are recognized andresponded to on a higher-scalar system of interpretance within the overall systemof relations involved. Consciousness means connecting to a meaning, whetherthis be iconic, indexical, or symbolic, or even all three simultaneously. Vaguesensations, mental images, perceptual awareness, and language all entail thebringing into awareness of neural meanings that are themselves brought abouton lower-scalar levels by the activations of neural maps and the communicationsthat occur between these. The crossing of the threshold into consciousnessawareness - i.e. from a prior state of neural activation below that threshold means that lower-scalar neural activities are semiotically reorganized on higherscalar levels such that they enter into contextual redundancies with emergentsigns - iconic, indexical, symbolic. It is these signs which enable the lower-scalaractivity to emerge in consciousness as an awareness of something, be this asensation (iconic), or an elementary association between two phenomena(index), a mental image or a clause (symbolic), and so on (see also Beaugrande1997: 114).

In this view, discrete neural firings are reorganized as globally coherentphenomena at a higher-scalar level than the individual firings to produce aparticular perceptual, conceptual, or symbolic (e.g. linguistic) categorization ofsome event, whether this occurs within the brain, in the body, or in the outsideworld. Meaningful patterns are emergent across global neural activity such thatthey are co-ordinated on higher scales of globally organized neural activity onlarger timescales to produce the emergent phenomena of conscious experience.The process is a stochastic or probabilistic one insofar as particular configurations of neuronal groups that subtend a specific categorization are the mostlikely ones to be activated for various instances of a particular event.

In terms of Lernke's (1999) Principle ofAlternation, we can say that the transformation from discrete, local neural firings to global configurations on higherscales and the transformation of such global configurations to objects ofconscious experience within the brain constitutes a semiotic transformation, asLemke says, 'of the information content of lower levels as signs for higher levels'(1999; italics in original). Consciousness of the objects of experience is the resultof dynamical processes across many different scales of organization within thebrain such that consciousness is an emergent level of organization relative to theobservational scales of the self. The objects of our conscious experience andawareness are objects which have information value for the higher-scalar systemof interpretance, comprised of value-laden memory in interaction with globallyorganized neural groups that function to activate perceptual, conceptual, andsymbolic categorizations. The systems of interpretance within the brain arethemselves built up over time on the basis of the cross-modal sampling andreentrant mapping of the body-brain's transactions with its inner and outerenvironments. Each of these levels of categorization is characterized by its ownsystems of types and at the same time the 'lower' levels are integrated to andreorganized by the higher levels such that perception, for example, may be


suffused with conceptual and symbolic categorizations. This does not alter thefact that different selection pressures act on different individuals differentlyalong their individuating trajectories such that the neuronal configurations ofdifferent individuals exhibit considerable diversity.

By the same token, such diversity is also entrained to the higher-scalarrelations and dynamics of a given ecosocial semiotic system. We are conscious ofthe phenomena which the higher-scalar systems of interpretance in our brainrecognize as differences which make a difference, i.e. which have informationalsemiotic value to that system. For this reason, many lower-scalar neural processesare no more than points in an overall phase space leading into the same higherscalar attractor. These lower-scalar processes include, of course, the configurations of neuronal activity which, to use Seager's term, 'represent' the contents ofour conscious experience. It is the attractors and their basins on higher levelsthat specify, semiotically, what can be recognized as an object of consciousness atthe same time as our interpretation of a given object of consciousness under agiven aspect - perceptual, conceptual, or symbolic - leaves open the manydegrees of topological freedom of its matter-energy flows. The view presentedhere obviates the need for a 'formal reality' which 'represents' the contents ofconscious experience because it is now clear that 'what is represented' is a resultof the semiotic closure that arises from the dynamics of the higher-scalarsystem(s) of interpretance within the brain. Lower-scalar neural firings, forexample, do not formally express higher-scalar content; instead, they constituteit by being entrained to the dynamics of the more global levels of organizationwhere conscious experience emerges in the perspective of the self.

It is the expression stratum of semiosis which provides the means whereby thesystems of categorization within the individual's brain may be communicated toother brains. This does not occur on the basis of any kind of transfer of codedthoughts or ideas from one brain to another. The dynamics of the body-braincomplex as a whole is entrained to the dynamics of the higher-scalar ecosocialsystem. For this reason, both the systems of conceptual-semantic discriminationsand sensori-motor ones are recategorized as the lexicogrammatical and semanticand phonological distinctions of the language system which is distributed acrossthe participants in a particular social network on different space-time scales. Inthis way, the cross-eoupling of sensori-motor activity to lexicogrammatical anddiscourse patterns of organization in the production of meaning-makingactivities and the texts that may be derived from these activities allows for bothembodied sensori-motor repertoires and conceptual repertoires stored in thebrain to be re categorized as phonological and lexicogrammatical and discoursesystems and structures which are semiotically salient in a particular ecosocialsemiotic system or some part of this. These systems of distinctions on theexpression and content levels of semiotic organization enter into the hierarchiesof contextualization relations - Lernke's meta-redundancy relations - such thatsensori-motor activity acquires meaning beyond that of mere body movement orvocalizing per se. The assigning of meaning to such activity occurs because aparticular sensori-motor routine or body movement is construed as having aregular pattern which redounds with some higher-order cultural or discursiveformation or genre convention.


The fact that such higher-order me ta-redundancy relations are distributedacross and hence recognized as semiotically salient in a given community meansthat the participants in the meaning-making and text-creating activities of thatcommunity have access to or participate in meaning-making practices wherebythey define and recognize the regular patterns in meaning-making activity andtextual artefacts. Rather than encoding thoughts for transmission from onemind to another, these regular patterns are semiotic affordances which allowparticipants to re-interpret these patterns according to systems of interpretancewithin their brains. Such processes of interpretation conform, to degrees, to therelatively stable hierarchies of contextualizing relations that are operationalwithin a given community. Therefore, perceivable patterns of embodied activityand their entextualizations can be re-contextualized in some other individual'sbrain as semantic or other semiotically salient categories which make adifference in both understanding and action. A semiotically salient difference isa value-laden difference in the sense that it makes a difference by virtue of itssystematic relations with other differences within the overall meaning system.This suggests that Edelman's notion of value-category memory can be extendedto the ecosocial scale of organization (Thibault 2000a). This is so in the sensethat the value-categories or semiotically salient differences that operate on spacetime scales considerably greater than either the individuallifespan trajectories orthe much smaller-scale logogenetic or meaning-making trajectories constitute atransindividual, cross-scalar ecosocial memory to which individuals areentrained.

9. The Three-level Hierarchy, Neuronal Activity, and the Emergence of theSelf-perspective

The self-perspective is not produced by neuronal activity. Instead, it is the selforganizing activity of the neuronal circuitry itself. Self-organizing systems arecomplex systems which have a history and individuality. This means that themany degrees of freedom of an individual's neuronal activity permit the construction of a continuity of individuality - a self-perspective - from one changeof-state in the system to another. Cerebral neurons are massively interconnectedin ways which allow for very many degrees of freedom. It is these degrees offreedom which build time and individuality into the structure and organizationof the cerebrum. Like all complex systems, an individual's neural networks canbe in states which are different from the states of other individuals. This ispossible because the individual's neural activity has a history of activity and interconnections which can never be exactly the same as another individual's. For thisreason, the self-perspective described by Flohr is an emergent property of neuralactivity itself. In other words, an individual perspective on one's own neuralactivity arises as a consequence of being in a particular brain state. It is in thissense that we can say that consciousness is a result of the formation of highlyspecified neural systems of which an individual's self-organizing neural networksare a particular case. As Kinsbourne (1988: 246) points out, neurons are characterized by the property that when they interact they produce awareness.


Consciousness is an emergent property of the interactions among neurons in aneural network. The same argument applies to the sense of a self who organizesand interacts with the self's own stream of consciousness.

Patterns of neural activity undergo constant change. A given change in neuralactivity may correspond to a particular brain state. A brain state is a change inthe overall neural network in response to a given input (Kinsbourne 1988:242).The system of neural activity is able to constitute the 'before' and 'after' ofa given brain state in ways which enable a continuity of individuality to be constructed. The individual character of self-organizing neural activity is a consequence of neural activity's network characteristics as well as the operation ofexcitation-inhibition balances which can either separate the cortical activity of aparticular mental operation from the cortical activity of other mental operationsor integrate it with other cortical activity. The self-organizing activity of theindividual's neural activity has a history of the ways in which neurons are interconnected on the basis of their network characteristics. Furthermore, mentaloperations and the cortical activity that instantiates these can be separated fromor integrated with other mental operations through the modulation of theexcitation-inhibition balance.

Both the network characteristics of cerebral activity and the excitationinhibition balance can be used to explain consciousness in terms of the threelevel hierarchy. This is a minimal assumption; it is not intended to explain thedetailed neural architecture of the brain. Rather, it can help to sort out therelations between the various levels involved. According to Salthe's three-levelparadigm, the focal level L is an emergent phenomenon which is constituted bythe interactions among variables on some characteristic scale. Units on level Lare constituted by the interactions among lower-scalar units on level L-1. For thesake of the present argument, let us say that level L-I refers to the interconnections among cerebral neurons whereby communication occurs betweenneurons. Level L units are mental operations which are constituted by the interactions among neurons on level L-l. Communications among neurons on levelL-I take place on the basis of impulses that are carried by the fibre tracts thatreciprocally link grey and white matter. The resulting cortico-cortical connections give rise to the cortical activity that instantiates a particular materialoperation on level L. We have seen that mental operations on level L areconstituted by smaller-scale, faster interactions at level L-l among the cerebralneurons at that lower level.

Now, mental operations on our putative level L can, as I pointed out abovewith reference to the work of Kinsbourne, be insulated from or integrated withother mental operations by virtue of the excitation-inhibition balances inoperation. Kinsbourne argues that 'to the extent a mental operation is selectivelyinsulated, it is automatic, and outside awareness, to the extent it is co-ordinatedwith other mental operations, it contributes to the "integrative field ofawareness'''. By the same token, not all of the possible communications whichmight occur on level L-l do actually occur. The interactions which producemental operations on level L are constrained by boundary conditions which aredetermined at the higher-scalar level L+1. Mental operations on level L areinstantiated by areas of cortical activation. Moreover, 'each cortical area


represents input within the constraints of its neural specialization and currentstate of interrelations with other areas (including activating and inhibitoryinfluences). Some representations are available as contributions to the field ofconscious awareness and, if attended, enter it' (Kinsbourne 1988: 240). LevelL+I imposes contextual constraints on level-L interactions. Further, only thoseinteractions on level L will be meta-stable on the appropriate timescale if theyconform to the constraints imposed by level L+1. In this perspective, the interrelations between different areas of neurally specialized cortical activity,including the modulation of the excitatory influences between areas, leads tothe integration on level L+I of diverse mental operations to the integrative fieldof awareness identified by Kinsbourne.

Rather than information processing, which would presuppose a one-to-onemapping between representations and outputs, the three-level hierarchy modelshows that the realizatory relationship between neuronal activity on level L-I andmental operations on level L is a many-to-one relation. Representation implies redescription, which leaves no room for the kind of reorganization across scalarlevels of a neuronal network that is fundamental in the perspective of the threelevel hierarchy. Re-description entails that discrete variants on one level aremapped to discrete variants on the next-higher level, or, alternatively, thatcontinuous variation is mapped to continuous variation, again at the next-higherlevel. In each case, there is re-description, but not reorganization, across levels.However, Lemke points out that the significance of emergent new levels of intermediate organization lies in the ways in which each higher level has many possiblerealizations at lower levels. That is, there are many more degrees of freedom onthe lower level than are salient on the higher level. In this way, a many-to-onemapping from level L-I to level L allows for the many degrees of freedom of theinformation on level L-I to be filtered or buffered. Neuronal interconnectivity onlevel L-I is activated as more global, larger-scale cortical activity realizing mentaloperations on level L. Furthermore, as we saw above, a mental operation on levelL can be integrated to and interpreted as a more global neural network whichembodies a conscious experience on level L+1. Level L+I is the higher-scalar,slower timescale of neuronal activity which integrates the cortical activity ofdiverse mental operations on the level below to its dynamics.

Lernke's Principle of Alternation suggests that discrete neural firings on levelL-I are reorganized as more global, continuous mental operations as one movesfrom lower to higher levels in the hierarchy. Information on level L-I is transformed as signs for higher levels such that both many-to-one classifications (levelL-I to level L) and one-to-many context-dependent reinterpretations (level L tolevel L+I) are possible. Consciousness is an emergent level of organizationwhereby level L entities are re-interpreted on level L+I as signs in the perspective of the self of microstates of the system at level L-I. When neural activity onlevel L-I is a consequence of interaction with the environment at level L-I,higher levels of the system respond to them as signs at level L of phenomena inthe environment at level L+I. Consciousness of phenomena in the environmentof the self is a higher-scalar attractor state of the global neural network whichmay not, as Lemke points out, have any direct material-causal relationship tolevel-L-I interactions.


The three-level hierarchy is incompatible with both dualism and reductionism. In the first instance, consciousness and the sense of a self, viewed asemergent properties of neuronal networks, do not transcend materialism in theway that mind-body dualism suggests. Instead, consciousness, mind, and self arehigher, larger-scale meaningful patterns of organization of lower, smaller-scaleneurophysiological processes. There is no suggestion that the higher-scaleprocesses have an existence independent of the lower-scale ones. Moreover, thehigher, larger-scale processes are emergent phenomena whose dynamicalprocesses and principles of organization cannot be reduced to or causallyexplained in terms of the lower, smaller-scale ones. In the second instance, thethree-level hierarchy view does not support the reductionist view that consciousness, mind, and self are reducible to and explainable in terms of lower-scaleneuronal processes per se as being the more fundamental causal level. Reductionism precludes emergence because it entails that the lower level constitutesthe more fundamental level and that everything else is epiphenomenal to this.

The perspective afforded by the three-level hierarchy view as presented heredoes not, of course, resolve all of the problems of consciousness. For example,the present account does not provide a detailed account of the structure andorganization of the cerebrum and its role in explaining the brain basis of consciousness. However, it does provide a perspective in which the tight linkagesacross the different scalar levels are seen as linking our material interactions withour inner and outer environments to our categories and ways of makingmeaning such that the body-brain complex is linked to the larger scales ofecosocial processes and their dynamics, including the always constructedtrajectory of the self (chapter 4, section 3, pp. 176--84) (see Lemke 2000a: 193).More generally, it also helps us to see human consciousness as a highly specificcase, at least from our own viewpoint, of more general types of complex selforganizing systems that exhibit properties of emergent organization, individuality, and developmental and individuating trajectories. Moreover, it shows thatpersons are embedded in and participate in ecosocial practices by virtue of theways in which they construe meaningful pattern in their material interactionswith their environment across diverse scalar levels. In this way, the distributednature of the neural networks of consciousness in the individual's brain can belinked to its higher-scalar boundary conditions and its lower-scalar initiatingconditions in a conceptually more unified way.

Epilogue

Towards a New Way of Thinking and Talking about Meaning-making

This book isjust the beginning of an attempt to rethink meaning-making activityfrom the perspective of the body-brain system - the signifying body - embeddedin its ecosocial semiotic environment. The key terms in this endeavour includeactivity, agency, body-brain system, contextualization, cross-eoupling, differentiation, dynamic open systems, ecosocial environment, embodiment, individuation,meaning-making, metafunctions, multimodality, process, scalar hierarchy, selforganization, specification hierarchy, system, timescales, topological, trajectory,typological, and value.

As a result of the way of thinking that has been partially explored in this book,many previous habits of thinking in linguistics and semiotics now seemsuperseded and are no longer useful tools for thinking. These includethe following dualisms: biology/society, behaviour/cognition, brain/mind,communication/cognition, form/meaning, mind/body, meaning/matter,nature/nurture, organism/environment, perception/semiosis, signifier/ signified, and structure/function.

Some of the fundamental premises that I have presented in this book may besummarized as follows.

The Semiotic Activity of the Signifying Body Integrates Diverse Space-timeScales

The notion of the 'signifying body' brings together the idea of a unitary bodybrain system, our socially constructed sense of an embodied identity-in-time, andthe ecosocial semiotic environment. The body-brain system regulates matter,energy, and information flows between itself and its ecosocial semiotic environment, and at the same time that the latter reorganizes body-brain dynamics andfunctions.

Semiotic activity requires no prior representation of signs within some preexisting system. The available resources - semiotic and material - which affordmeaning-making are not all simply stored inside the individual's head. Instead,they exist in the ecosocial semiotic environment in the form of, for example,

EPILOGUE 315

other persons, artefacts, tools, texts, technologies, and so on, that can beintegrated to and exploited in meaning-making activity. The structure ofsemiotic activity self-organizes in relation to its ecosocial environment. There isno single locus of control in the form of internal programs, competences, orplans. Rather, control is distributed along the entire trajectory that loopsbetween organism and its ecosocial environment.

Three-level hierarchy thinking shows that lower-scalar biological initiatingconditions on level L-l, e.g. the many degrees of topological freedom of thevocal tract, or of hand-arm activity, provide both affordances and their ownmaterial constraints, without entailing a central program that controls thesystem's activity. A higher-scalar system of interpretance in the ecosocial environment (level Lt l ) imposes boundary conditions on the kinds of organismenvironment transactions that can occur on the intermediate level L.

Meaning-making activity is a result of the dynamical interplay and integrationof processes on many different levels. The structure of the activity is shaped bythe way it interfaces with the body-brain system, on the one hand, and with theecosocial semiotic environment, on the other. In this way, the signifying body iscontextually integrated, through the mediating activity of semiosis, with diversespace-time scales that go far beyond the body's immediate material interactivitywith its here-now environment.

Brain Activity Regulates Body-environment Relations and, at the same time,Body-environment Relations Organize and Shape Body-brain Systems andFunctions

Neural activity regulates body-environment relations by producing sensorimotor outputs whereby responses to environmental events are produced andprojected back into the environment. The infant's high-pitched squeak inHalliday's (1993) example, as discussed in chapter 4, section 10, pp. 202-5, isa response in this sense. The infant's proto-linguistic squeak is a response to anevent - the scattering of the pigeons - on the here-now scale. However, thereis nothing which necessarily connects it to that event. In this sense, the infant'svocalization can be said to stand in a proto-symbolic relation to the environmental event to which it responds. It is only the connectedness of the infant'svocal-tract activity to the stimulus information (the sound) that is produced byhis vocal-tract activity and which can be picked up by others that is, strictlyspeaking, indexical.

On the other hand, the relation between the infant's vocalization and theenvironmental event - the pigeons - to which it constitutes a response, is protosymbolic from the point of view of observers. This is so because the transductionin an observer system (e.g. the mother) of the infant's sensori-motor activity asinformation about an event (the pigeons) that is not necessarily (indexically)tied to the infant's own body affords both dialogic engagement with - e.g.responding to and being responded to - the nonself and jointly attending tosome phenomenon of experience. Thus, we see that neural activity regulatesbody-environment relations in ways that are not tied to external events in a fixed


relation of indexical necessity. This allows both for creativity and for thedecreasing dependence of brain-regulated body-environment relations on theimmediate here-now scale.

Body-environment Relations and Body-brain Systems and Functions areShaped by Metafunctional Principles of Semiotic Organization

At the same time, body-environment relations contribute to the reorganizationof body-brain functions along progressively metafunctional lines as the infant'ssemiotic activity is increasingly hitched to and entrained by the models ofdiscourse and society that are provided by more senior partners in meaningmaking. The vague outlines of the metafunctions are therefore evident in theinfant's vocalization and can be related, as we saw in chapter 5, to the shape ofconsciousness itself.

Brain Activity is Contextually Integrated to and Directly Participates inDiscourse, Rather Than Causing it

Distributed patterns of neural activity within the networks of connections in thebrain are contextually integrated to particular kinds of meanings and events inthe ecosocial environment. A particular pattern of neural activity on its scalarlevel is integrated to patterns of activity on other scalar levels in the formation ofstable categories of meaning and experience. In this view, all is activity andprocess on many different scalar levels of organization. There is no need for aninner program or homunculus which programs the activity. Nor is there anyultimate, lower-level substance on which the higher levels are supervenient(Bickhard and Campbe1l2000: 329-32).

Alterity is a Primitive Intrinsic Value that Motivates Self-nonself Relationsand Meaning-making Activity

The dialogical orientation of the self to the nonself is intrinsic to our biologicalinheritance from the outset. Alterity is therefore a fundamental organizingprinciple of the self right from the outset. In this sense, it is a biologicallyintrinsic value which selectively nudges the infant along some developmentalpathways rather than others (chapter 1, section 11, pp. 49-54). It is a principlewhich is both 'inside' and 'outside' the individual by virtue of the fact thatmeaning-making is a distributed activity between body-brain systems and theirecosocial semiotic environments on diverse scalar levels of spatio-temporal andsemiotic organization.

EPILOGUE

(Self)Consciousness is a Dialogical and Semiotically Mediated Relationbetween Self and Nonself

317

(Self-)consciousness is necessarily and irreducibly a dialogical and semioticallymediated relation between self and nonself. The inner, self-referential perspective of the self is constructed in and through a complex chain of contextualizingrelations which integrates processes on very many different space-time scales,e.g. from the assembly formation of nerve nets in the brain to the generating ofmeta-perspectives on the dialogical refraction of the self through the perspectiveof the other in discourse, as in Volosinov's (1976: 86) definition of 'selfawareness' or 'self-consciousness'.

The assembling of nerve nets is based on patterns of neural activation. In thebrain, as the theories of Edelman, Flohr and Kinsbourne show, patterns ofneural activation are related to each other in complex hierarchies. To use theBatesonian parlance, there are patterns of patterns of patterns of patterns ... ofneural activity. These patterns form complex hierarchies of contextualizingrelations both within the brain and with the diverse scalar levels of organizationthat link brain, body, and ecosocial semiotic environment.

Flohr (1991: 258) points out that the formation of a self-referential innerperspective is a consequence of the formation of such hierarchies of patterns ofpatterns ... of neural activity to very high orders of complexity. Furthermore, theresultant self-referential perspective integrates to its own perspective the brain'srepresentation of the various parts of the body, including its internal milieu,along with the meaning-making activity of the body, such that the selfjust is thisintegration of the body, its meanings, and its experiences, on all levels of thespecification hierarchy, to the self-referential perspective that is realized insymbolic neural space in the brain.

Towards a Process Ontology of Social Being-in-doing

The 'arrow of time' of irreversible processes refers to concrete, lived historicaltime. This is the time of all complex hierarchically organized, dynamic, adaptiveand open goal-seeking systems. This conception now makes it possible, historically speaking, for the first time, to construct what Lukacs (1990: 111-12) hasdesignated 'a unitary ontology' of ecosocial semiotic processes and, hence, ofour identities-in-time. The ontological bases of this ontology are the physicalmaterial and semiotic-discursive cross-couplings and the irreversible timedependent processes and activities on many scalar levels of organization whichconstitute them. The self-organization of self and nonself and, therefore, ofmeaning, embodiment, and consciousness, is an irreversible historical andecosocial semiotic process in this most fundamental way.

References

Abercrombie, David 1967. Elements of General Phonetics. Edinburgh: Edinburgh UniversityPress.

Andersen, Peter Bagh 2000. 'Genres as self-organising systems'. In Peter Bogh Andersen,Claus Emmeche, Niels Ole Finnemann, and Peder Voetmann Christiansen (eds),Downward Causation: Minds, Bodies and Matter, 214--60. Aarhus: Aarhus UniversityPress.

Armstrong, David F., Stokoe, WiIliam C., and Wilcox, Sherman E. 1995. Gesture and theNature of Language. Cambridge: Cambridge University Press.

Astington, J. W, Harris, P., and Olson, D. R. 1988 (eds) , Developing Theories of Mind.Cambridge: Cambridge University Press.

Bakhtin, Mikhail M. 1981 [1975]. 'Discourse in the novel'. In Michael Holquist (ed.), TheDialogic Imagination:FourEssays, 259-422. Austin: University of Texas Press.

Bakhtin, Mikhail M. 1986. 'The problem of speech genres'. In Caryl Emerson and MichaelHolquist (eds), Speech Genres & OtherLateEssays, 60-102. Trans. Vern W McGee. Austin:University of Texas Press.

Bakhtin, Mikhail M. 1990 [1924]. 'The problem of content, material, and form in verbalart'. In Michael Holquist and Vadim Liapunov (eds), Art and Answerability. EarlyPhilosophical Essays byM. M. Bakhtin, Trans. Vadim Liapunov, 257-325. Austin: University ofTexas Press.

Baron-Cohen, S. 1989a. 'Perceptual role-taking and protodeclarative pointing in autism'.BritishJournal ofDevelopmental Psychology 7: 113-27.

Baron-Cohen, S. 1989b. 'The autistic child's theory of mind: a case of specific developmental delay'. Journal of ChildPsychology and Psychiatry 30: 285-98.

Bates, Elizabeth 1979 (with the collaboration of Laura Benigni, Inge Bretherton, LuigiaCamaioni, Vicki Carlson, Karlana Carpen, and Marcia Rossner). The Emergence ofSymbols: Cognition and Communication in Infancy.NewYork and London: Academic Press.

Bateson, Gregory 1973a. 'The cybernetics of "self": a theory of alcoholism'. In GregoryBateson, Steps to an Ecology of Mind, 280-308. London and New York: Granada.

Bateson, Gregory 1973b. 'Redundancy and coding'. In Gregory Bateson, Steps to an EcologyofMind, 387-401. London and New York: Granada.

Bateson, Gregory 1973c. 'Problems in cetacean and other mammalian communication'. InGregory Bateson, Steps to an Ecology of Mind, 334-48. London and New York: Granada.

Bateson, Gregory 1973d. 'Style, grace, and information in primitive art'. In GregoryBateson, Steps to an Ecology of Mind, 101-25. London and New York: Granada.

Bateson, Gregory 1973e. 'The role of somatic change in evolution'. In Gregory Bateson,Steps to an Ecology of Mind, 316-33. London and New York: Granada.

Bateson, Gregory 1980 [1979]. Mind and Nature: A Necessary Unity. London: Fontana.Bateson, Gregory 1987 [1951]. 'Information and codification: a philosophical approach'.

REFERENCES 319

In Jurgen Ruesch and Gregory Bateson, Communication: The Social Matrix ofPsychiatry,168-211. London: Norton.

Battacchi, Marco W. 1996. 'Conscience de soi et connaissance de soi dans l'ontogenese'.Enfance 2: 156---64.

Battacchi, Marco W. 1998. 'Self-knowledge and self-consciousness'. Dipartimento diPsicologia, Universita degli Studi di Bologna: Mimeo.

Battacchi, Marco W. 1999. 'Consciousness, awareness, and experience'. Teorie & ModelliIV(3): 55-63.

Battacchi, Marco, Battistelli, Piergiorgio, and Celani, Giorgio 1998. Lo Sviluppo del PensieroMetarappresentativo e della Coscienza: Milan: Franco Angeli.

Beaugrande, Robert de 1997. New Foundations for a Science of Text and Discourse: Cognition,Communication and the Freedom of Access to Knowledge and Society. Norwood, NJ: Ablex.

Benson,James, Fries, Peter, Greaves, William, Iwamoto, Kazuyoshi, Savage-Rumbaugh,Sue, and Taglialatela, Jared. 2002. 'Confrontation and support in bonobo-humandiscourse'. Functions ofLanguage 9 (1): 1-38.

Benson, James, Fries, Peter, Greaves, William, Iwamoto, Kazuyoshi, Savage-Rumbaugh,Sue, and Taglialatela,Jared. 2004. 'Language, apes, and meaning-making'. In GeoffreyWilliams and Annabelle Lukin (eds), Language Development: Functional Perspectives onSpecies and Individuals, 91-111, London and New York: Continuum.

Bemstein, Basil 1971. Class, Codesand Control, VolumeI: TheoreticalStudies Towards a SociologyofLanguage. London and Boston: Routiedge & Kegan Paul.

Bernstein, Basil 1990. Class, Codes and Control, Volume JY.. The Structuring of PedagogicDiscourse. London and New York: Routiedge.

Bernstein, N. 1967. Coordination and Regulation ofMovements. New York: Pergamon Press.Berthoz, Alain 1997. Le Sens du Mouvement. Paris: Editions Odile Jacob.Bhaskar, Roy 1979. The Possibility of Naturalism: A Philosophical Critique of the Contemporary

Human Sciences. Brighton, Sussex: Harvester.Bickhard, Mark H. and Campbell, Donald T. 2000. 'Emergence'. In Peter B0gh Andersen,

Claus Emrneche, Niels Ole Finnemann, and Peder Voetmann Christiansen (eds),Downward Causation: Minds, Bodies and Matter, 322--48. Aarhus: Aarhus UniversityPress.

Bogdan, RaduJ. 1987. 'Mind, content and information'. Synthese 70: 205-77.Bogdan, Radu J. 2000. Minding Minds: Evolving a Reflexive Mind by Interpreting Others.

Cambridge, MA and London: MIT Press.Bohr, Nielsen 1948. 'On the notions of causality and complementarity'. Dialectica 7(8):

313-14.Bolinger, Dwight 1985. 'The inherent iconism of intonation'. In John Haiman (ed.),

Iconicity in Syntax, 97-108. Amsterdam and Philadelphia: John Benjamins.Bouissac, Paul 1998. 'Converging parallels: semiotics and psychology in evolutionary per

spective'. Theory & Psychology 8(6): 731-53.Bourdieu, Pierre 1977 [1972]. Outline ofa Theory ofPractice. Trans. Richard Nice. Cambridge

and Melbourne: Cambridge University Press.Braten, S. 1992. 'The virtual other in infants' minds and social feelings'. In A. Heen Wold

(ed.), The Dialogical Alternative: Towards a Theory of Language and Mind, 77-97. Oslo:Scandinavian University Press.

Browman, Catherine and Goldstein, Louis 1991. 'Gestural structures: distinctiveness,phonological processes, and historical change'. In Ignatius G. Mattingly and MichaelStuddert-Kennedy (eds), Modularity and the Motor Theory of Speech Perception, 313-38.Hillsdale, NJ: Lawrence Erlbaum.

Browman, Catherine P. and Goldstein, Louis 1995. 'Dynamics and articulatory phonology'.In Robert F. Port and Timothy van Gelder (eds), Mind as Motion: Explorations in theDynamics of Cognition, 175-93. Cambridge, MA and London: MIT Press.

Buhler, Karl 1990 [1934]. Theory of Language: The Representational Function of Language.Trans. Donald Fraser Goodwin. Amsterdam/Philadelphia: Benjamins.

320 REFERENCES

Carruthers, Peter 1996. Language, Thought and Consciousness: An Essay in PhilosophicalPsychology. Cambridge and New York: Cambridge University Press.

Catford, J. C. 1998. A Practical Introduction to Phonetics. Oxford: Clarendon.Chomsky, Noam 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.Chomsky, Noam 1976. Reflections on Language. London: FontanaClark, E. and Clark, H. 1977. Psychology and Language. New York: Harcourt Brace.Cleirigh, Chris 1998. A Selectionist Modelof The Genesis of PhonicTexture: Systemic Phonology &

Universal Darwinism. Department of Linguistics, University of Sydney: UnpublishedPh.D. thesis.

Cole, Michael 1996. Cultural Psychology: A Once and FutureDiscipline. Cambridge, MA andLondon: Harvard University Press.

Coleridge, Samuel Taylor 1967. Biographia Literaria. London: Dent.Corina, David P., Vaid, ]yotsna, and Bellugi, Ursula 1992. 'The linguistic basis of left

hemisphere specialization'. Science 255: 1258-60.Coulthard, Malcolm 1992. 'The significance of intonation in discourse'. In Malcolm

Coulthard (ed.), Advances in Spoken Discourse Analysis, 35-49. London and New York:Routledge.

Cowley, Stephen J. 1998. 'Of timing, turn-taking, and conversations'. Journal of PsycholinguisticResearch 27(5): 541-71.

Cowley, Stephen J. 2001. 'The baby, the bathwater, and the "language instinct" debate'.LanguageSciences 23: 69-91.

Cowley, Stephen J. 2002. 'Why brains matter: an integrational perspective on The SymbolicSpecies'. LanguageSciences 24: 73-95.

Cowley, Stephen J., Moodley, S. and Fiori-Cowley, Agnese. In press. 'Grounding signs ofculture: primary intersubjectivity in social semiosis'. Mind, Culture, and Activity11.2.

Cremonini, Giorgio and Frasnedi, Fabrizio 1986. Nell'Universo delSenso. Bologna: Il Mulino.Damasio, Antonio 1999. The Feeling of What Happens: Body, Emotion and the Making of Con

sciousness. London: William Heinemann.Davidse, Kristin 1991. 'Categories of Experiential Grammar'. Ph.D. thesis. Leuven:

Department of Linguistics, Katholieke Universiteit.Davidse, Kristin 1992. 'A semiotic approach to relational clauses'. open Papers in Systemic

Linguistics (OPSL) 6: 99-131.Davidse, Kristin 1997. 'The Subject-Object versus the Agent-Patient asymmetry'. Paper

presented at the congress 'Objects, grammatical relations and semantics', University ofGhent, 23-24 May 1997.

Davidson, RJ. 1984. 'Hemispheric asymmetry and emotion'. In K. Scherer and P. Ekman(eds), Approaches toEmotion, 39-57. Hillsdale, NJ: Erlbaum.

Davidson, R J. 1992. 'Anterior cerebral asymmetry and the nature of emotion'. Brain andCognition 20: 125-51.

Deacon, Terrence 1998 [1997]. The Symbolic Species: The Co-evolution of Language and theHuman Brain. London and New York: Penguin.

Descartes, Rene 1960 (1641]. 'Meditations'. In Discourse on Method and Other Writings,101-69. Trans. Arthur Wollaston. Harmondsworth: Penguin.

Doty, R W. 1984. 'Some thoughts and some experiments on memory'. In L. R Squire andN. Butters (eds), Neuropsychology of Memory. New York: Guilford.

Dretske, F. 1972. 'Contrastive statements'. Philosophical Review 81: 411-37.Edelman, Gerald M. 1987. Neural Darwinism: The Theory of Neuronal Group Selection. New

York: Basic Books.Edelman, Gerald M. 1989. The Remembered Present: A Biological Theory of Consciousness. New

York: Basic Books.Edelman, Gerald M. 1992. BrightAir, Brilliant Fire: On the Matter of the Mind. London and

New York: Penguin.Edelman, Gerald M. and Tononi, Giulio 2000. A Universe of Consciousness: How Matter

Becomes Imagination. New York: Basic Books.Finnemann, Niels Ole 2000. 'Rule-based and rule-generating systems'. In Peter B0gh

REFERENCES 321

Andersen, Claus Emmeche, Niels Ole Finnemann, and Peder Voetmann Christiansen(eds), Downward Causation: Minds, Bodies and Matter, 278-302. Aarhus: AarhusUniversity Press.

Firth 1957a [1948]. 'The semantics oflinguistic science'. In Papers in Linguistics 1934-1951,139-47. London and New York: Oxford University Press.

Firth,]. R. 1957b [1948]. 'Sounds and prosodies'. In Papers in Linguistics 1934-1951,121-38. London and New York: Oxford University Press.

Firth,]. R. 1957c [1951]. 'General linguistics and descriptive grammar'. In Papers in Linguistics 1934-1951, 216-28. London and New York: Oxford University Press.

Flohr, Hans 1991. 'Brain processes and phenomenal consciousness: a new and specifichypothesis'. Theory & Psychology 1(2): 45-262.

Fowler, Carol A. 1986. 'An event approach to the study of speech perception from a directrealist perspective'. Journal ofPhonetics 14: 3-28.

Fowler, Carol A. and Rosenblum, Lawrence D. 1991. 'The perception of phonetic gestures'.In Ignatius G. Mattingly and Michael Studdert-Kennedy (eds) , Modularity and the MotorTheory ofSpeechPerception, 33-59. Hillsdale, NJ: Lawrence Erlbaum.

Gee, James Paul 1992. The Social Mind: Language, Ideology, and Social Practice. New York andLondon: Bergin & Garvey.

Gendlin, Eugene 1962. Experiencing and the Creation of Meaning: A Philosophical and Psychological Approach to the Subjective. Evanston, IL: Northwestern University Press.

Gibson, James.J. 1983 [1966]. The Senses Considered as Perceptual Systems. Westport, CT:Greenwood.

Gibson, James]. 1986 [1979]. The Ecological Approach to Visual Perception. Hillsdale, NJ andLondon: Lawrence Erlbaum.

Goldfield, Eugene C. 1993. 'Dynamic systems in development: action systems'. In Linda B.Smith and Esther Thelen (eds), A Dynamic Systems Approach to Development: Applications,51-70. Cambridge, MA and London: MIT Press.

Goodwin, Charles 1996a. 'Practices of color classification'. Ninchi Kagaku (CognitiveStudies: Bulletin of the Japanese Cognitive Science Society). 3(2): 62-82. [Alsopublished as Goodwin, Charles 2000. 'Practices of color classification'. Mind Culture andActivity 7(1&2): 19-36.]

Goodwin, Charles 1996b. 'Transparent vision'. In Elinor Ochs, Emanuel A. Schegloff, andSandra Thompson (eds), Interaction and Grammar, 370-404. Cambridge: CambridgeUniversity Press.

Grene, Marjorie. 1990. 'Perception and human reality'. In Roy Bhaskar (ed.), Hard and HisCritics:Essass in Honour ofRom Harre with his Commentary on Them, 17-22. Oxford: BasilBlackwell.

Crusser, 0.]. 1991. 'Cortical representation of head movement in space and some psychophysical considerations'. In A. Berthoz, P. P.Vidal, and W. Graf (eds) , The Head-NeckSensory-MotorSystem, 497-509. Oxford and New York: Oxford University Press.

Haken, H. 1977. Synergetics: An Introduction. Heidelberg: Springer-Verlag.Halliday, M. A. K. 1967a. A Course in Spoken English: Intonation. Oxford: Oxford University

Press.Halliday, M. A. K. 1967b. 'Notes on transitivity and theme in English 1'. Journal of Linguis

tics 3(1): 37-81.Halliday, M. A. K. 1967c. 'Notes on transitivity and theme in English II'.Journal ofLinguis

tics 3(2): 199-244.Halliday, M. A. K. 1968. 'Notes on transitivity and theme in English III'. Journal ofLinguis

tics 4(2): 179-215.Halliday, M. A. K. 1975. Learning How to Mean: Explorations in the Development of Language.

London: Edward Arnold.Halliday, M. A. K. 1978a. Language as Social Semiotic: The Social Interpretation ofLanguage and

Meaning. London: Edward Arnold.Halliday, M. A. K. 1978b. 'Meaning and the construction of reality in early childhood'. In

322 REFERENCES

Herbert L. Pick and Elliot Saltzman (eds), Modes ofPerceiving and ProcessingInformation.67-96. Hillsdale, NJ: Erlbaum.

Halliday, M. A. K. 1979. 'Modes of meaning and modes of expression: types of grammatical structure and their determination by different semantic functions'. In D.J.Allerton,Edward Carney, and David Holdcroft (eds), Function and Context in Linguistic Analysis:A Festschriftfor William Haas, 57-79. Cambridge: Cambridge University Press.

Halliday, M. A. K. 1984. 'Listening to Nigel: Conversations of a very small child'. Unpublished manuscript, University of Sydney, Linguistics Department, Sydney, Australia.

Halliday, M. A. K. 1988. 'On the ineffability of grammatical categories'. InJames D. Benson,Michael J. Cummings, and William S. Greaves (eds), Linguistics in a Systemic Perspective,27-51. Amsterdam/Philadelphia: Benjamins.

Halliday, M. A. K. 1991. 'Towards probabilistic interpretations'. In Eija Ventola (ed.),Functional and SystemicLinguistics: ApjmJachesand uses, 39-61. Berlin: Mouton de Gruyter.

Halliday, M. A. K. 1992a. 'How do you mean?'. In Martin Davies and Louise Ravelli (eds),Advances in SystemicLinguistics: Recent Theory and Practice, 20-35. London and New York:Pinter.

Halliday, M. A. K. 1992b. 'A systemic interpretation of Peking syllable finals'. In P. Tench(ed.), Studies in SystemicPhonology, 98-121. London and New York: Pinter.

Halliday, M. A. K. 1993. 'Towards a language-based theory of learning'. Linguistics andEducation 5: 93-116.

Halliday, M. A. K. 1994 [1985]. Introduction to Functional Grammar. 2nd edn. London andMelbourne: Arnold.

Halliday, M. A. K. 1995. 'On language in relation to the evolution of human consciousness'.In S. AlIen (ed.), Of Thoughts and Words: Proceedings ofNobel Symposium 92: 'The RelationBetween Language and Mind'. Stockholm, 8-12 August 1994, 45-84. River Edge, NJ:Imperial College Press.

Halliday, M. A. K. 2000. 'Phonology past and present: a personal retrospect'. Folia Linguistica XXXN(I-2): 101-11.

Halliday, M. A. K. and Hasan, Ruqaiya 1976. Cohesion in English. London: Longman.Halliday, M. A. K. and Matthiessen, Christian 1999. Construing Experience through Meaning: A

Language-based Approach to Cognition. London and New York: Cassell.Handel, Stephen 1989. Listening: An Introduction to the Perception of Auditory Events.

Cambridge, MA and London: MIT Press.Harre, Rom 1979. Social Being: A Theory for Social Psychology. Oxford: Blackwell.Harre, Rom 1983. Personal Being: A Theory for Individual Psychology. Oxford: Blackwell.Harre, Rom 1990. 'Exploring the human Umwelt'. In Roy Bhaskar (ed.), Harre and His

Critics: Essays in Honour of Rom Harre with his Commentary on Them, 297-364. Oxford:Basil Blackwell.

Harre, Rom and Gillett, Grant 1994. The Discursive Mind. Thousand Oaks, CA and NewDelhi: Sage.

Harris, Roy 1995a. Language, Signs, and Communication. London and New York: Routledge.Harris, Roy 1995b. Signs of Writing. London and New York: Routledge.Hasan, Ruqaiya 1980. 'The texture ofa text'. In M. A. K. Halliday and Ruqaiya Hasan, Text

and Context: Aspects of Language in a Social-semiotic Perspective, 43-59. Sophia University,Tokyo, The Graduate School of Languages and Linguistics, Linguistic Institute forInternational Communication: Working Papers in Linguistics, No. 6.

Hasan, Ruqaiya 1992a. 'Rationality in everyday talk: from process to system'. InJan Svartik(ed.), Directions in Corpus Linguistics: Proceedings of Nobel Symposium 82 Stockholm, 4-8August 1991, 257-307. Berlin and New York: Mouton de Gruyter.

Hasan, Ruqaiya 1992b. 'Speech genre, semiotic mediation and the development of thehigher mental functions'. Language Sciences14(4): 489-528.

Hasan, Ruqaiya 1996. 'Semantic networks: a tool for the analysis of meaning'. In CarmelCloran, David Butt, and GeoffWilliams (eds), Ways ofSaying: Ways of Meaning, 104-31.London and New York: Cassell.

REFERENCES 323

Hasan, Ruqaiya 1999. 'Speaking with reference to context'. In Mohsen Ghadessy (ed.), Textand Context in Functional Linguistics, 219-328. Amsterdam/Philadelphia: JohnBenjamins.

Hauser, Marc D., Chomsky, Noam, and Fitch, W. Tecumseh 2002. 'The faculty of language:what is it, who has it, and how did it evolve?'. Science Vol. 298, 22 November 2002:1569-78.

Hebb, D. O. 1949. The Organization of Behavior. New York: Wiley.Hjelmslev, Louis 1954. 'La stratification du langage'. Word 10(2-3): 163-88.Hjelmslev, Louis 1961 [1943]. Prolegomena to a Theory of Language. Trans. Francis J.

Whitfield. Rev. English edn. Madison, Milwaukee and London: University of WisconsinPress.

1edema, Roderick A. M. 1997. 'Interactional dynamics and social change: planning as morphogenesis'. University of Sydney, Department of Linguistics: Unpublished Ph.D. thesis.

Jakobson, Roman and Santilli, Kathy 1980. Brain and Language: Cerebral Hemispheres andLinguistic Structures in Mutual Light. Columbus, OH: Slavica.

Johnson, M. H. and Morton,J. 1991. Biology and Cognitive Development: The Case ofFace Recognition. Oxford: Blackwell.

Johnston, Trevor 1992. 'The realization of the linguistic metafunctions in a sign language'.SocialSemiotics 2(1): 1-43.

Juarrero, Alicia 1999. Dynamicsin Action:Intentional Behavioras a Complex System. Cambridge,MA and London: M1T Press.

Kauffman, S. 1993. The Originsof Order: Self-organization and Selection in Evolution. New Yorkand Oxford: Oxford University Press.

Kaye, Kenneth. 1984 [1982]. The Mental and SocialLife of Babies: How Parents Create Persons.London: Methuen.

Kelso,J. A. S., Saltzman, E. L., and Tuller, B. 1986. 'The dynamical perspective on speechproduction: data and theory'. Journal of Phonetics 14: 29-59.

Kinsbourne, Marcel 1988. 'Integrated field theory of consciousness'. In A. J. Marcel andE. Bisiach (eds), Consciousness in Contemporary Science, 239-56. Oxford: Clarendon.

Kinsbourne, Marcel 2000. 'Inner speech and the inner life'. Brain and Language71: 120-3.Kinsbourne, Marcel 2001. 'Unity and diversity in the human brain: evidence from injury'.

In Gerald M. Edelman and Jean-Pierre Changeux (eds), The Brain, 233-56. NewBrunswick, NJ and London: Transaction Publishers.

Kosslyn, S. M., Chabris, C. F., Marsolek, C. J., Jacobs, R. A., and Koenig, O. 1995. 'Oncomputational evidence for different types of spatial relations encoding: reply toCook'. Journal ofExperimental Psychology 21(2): 423-31.

Kress, Gunther and Van Leeuwen, Theo 1996. ReadingImages: The Grammarof VisualDesign.London and New York: Routledge.

Lakoff, George andJohnson, Mark 1999. Philosophy in the Flesh: The Embodied Mind and itsChallenge to Western Thought. New York: Basic Books.

Lamb, Sydney M. 1966. 'Epilegomena to a theory of language'. Romance Philology 19(4):53]-73.

Langacker, Ronald W. 1987. Foundations of Cognitive Grammar. Vol. I. Theoretical Prerequisites.Stanford, CA: Stanford University Press.

Langacker, Ronald W. 1991. Foundations of Cognitive Grammar. Volume Il. Descriptive Application. Stanford, CA: Stanford University Press.

Latour, Bruno 1986. 'The powers of association'. In John Law (ed.), Power, Action and Beliej,264-80. London and Boston: Routledge & Kegan Paul.

Latour, Bruno 1996a. 'On interobjectivity'. Mind, Culture, and Activity 3( 4): 228-45.Latour, Bruno ]996b. 'Pursuing the discussion of interobjectivity'. Mind, Culture, and

Activity 3(4): 266-9.Lave,Jean 1997. 'The culture of acquisition and the practice of understanding'. In David

Kirshner and James A. Whitson (eds), Situated Cognition: Social, Semiotic, and PsychologicalPerspectives, ] 7-35. Mahwah, NJ: Lawrence Erlbaum.

324 REFERENCES

Laver, John 1970. 'The production of speech'. In John Lyons (ed.), New Horizons inLinguistics, 53-75. Harmondsworth: Penguin.

Lemke, Jay L. 1983. 'Thematic analysis: systems, structures, and strategies'. Recherches Sbniotiques/Semiotic Inquiry (RSSI) 3(2): 159-87.

Lemke, Jay L. 1984a. 'Action, context, and meaning'. In J. L. Lernke, Semiotics andEducation, 63-93. Victoria University, Toronto: Toronto Semiotic Circle Monographs,Working Papers and Prepublications, No. 2.

Lemke, Jay L. 1984b. 'Making trouble'. In J. L. Lemke, Semiotics and Education, 94-149.Victoria University, Toronto: Toronto Semiotic Circle Monographs, Working Papersand Prepublications, No. 2.

Lemke, Jay L. 1984c. 'The formal analysis of instruction'. In J. L. Lemke, Semiotics andEducation, 23-62. Victoria University, Toronto: Toronto Semiotic Circle Monographs,Working Papers and Prepublications, No. 2.

Lernke, Jay L. 1984d. 'Towards a model of the instructional process'. In J. L. Lemke,Semiotics and Education, 6-22. Victoria University, Toronto: Toronto Semiotic CircleMonographs, Working Papers and Prepublications, No. 2.

Lemke, Jay L. 1985. 'Ideology, intertextuality, and the notion of register'. In James D.Benson and William S. Greaves (eds) , Systemic Perspectives on Discourse, Volume I, 275-94.Norwood, NJ: Ablex.

Lemke, Jay L. 1990. TalkingScience: Language, Learning, and Values. Norwood, NJ: Ablex.Lemke, Jay L. 1995a. 'The social construction of the material subject'. In Jay L. Lemke,

TextualPolitics: Discourse and Social Dynamics, 80-99. London and Bristol, PA: Taylor &Francis.

Lemke, Jay L. 1995b [1993]. 'Discourse, dynamics, and social change'. In Jay L. Lemke,TextualPolitics: Discourse and Social Dynamics, 100-29. London and Bristol, PA: Taylor &Francis.

Lemke, Jay L. 1995c. 'Retrospective Postscript'. In Jay L. Lemke, TextualPolitics: Discourseand Social Dynamics, 154-84. London and Bristol, PA: Taylor & Francis.

Lemke, Jay L. 1997. 'Cognition, context, and learning: a social semiotic perspective'. InDavid Kirshner and James A. Whitson (eds), Situated Cognition: Social, Semiotic, andPsychological Perspectives, 37-55. Mahwah, NJ: Lawrence Erlbaum.

Lemke, Jay L. 1998. 'Multiplying meaning: visual and verbal semiotics in scientific texts'. InJ. R. Martin and R. Veel (eds), Reading Science: Critical and Functional Perspectives onDiscourses of Science, 87-113. London and New York: Routledge.

Lemke,Jay L. 1999. 'Opening up closure: semiotics across scales'. Paper presented at theconference, 'Closure: Emergent Organizations and their Dynamics', University ofGhent, Belgium; May 1999; website: http://academic.brooklyn.cuny.edu/ education/jlemke/index.htm

Lemke, Jay L. 2000a. 'Material sign processes and emergent ecosocial organization'. InPeter Bogh Andersen, Claus Emmeche, Niels Ole Finnemann, and Peder VoetmannChristiansen (eds), Downward Causation: Minds, Bodies and Matter, 181-213. Aarhus:Aarhus University Press.

Lemke, Jay L. 2000b. Across the scales of time: artefacts, activities, and meanings inecosocial systems'. Mind, Culture, and Activity7(4): 273-90.

Liberman, A. and Mattingly, 1. 1985. 'The motor theory of speech perception revised'.Cognition 21: 1-36.

Libet, Benjamin 1996. 'Neural processes in the production of conscious experience'. InMax Velmans (ed.), The Science of Consciousness: Psychological, Neuropsychological andClinicalReviews, 96-117. London and New York: Routledge.

Lieberman, Philip 1984. The Biology and Evolution of Language. Cambridge, MA andLondon: Harvard University Press.

Lindblom, Bjorn 1991. 'The status of phonetic gestures'. In 19natius G. Mattingly andMichael Studdert-Kennedy (eds), Modularity and the Motor Theory of Speech Perception,7-24. Hillsdale, NJ: Lawrence Erlbaum.

REFERENCES 325

Longacre, Robert E. 1983. The Grammar ofDiscourse. New York and London: Plenum Press.Lorenz, Konrad. 1969 [1963]. On Aggression. Trans. Marjorie Latzke. London: Methuen.Lorenzana,Juan Alvarez de 1993. 'The constructive universe and the evolutionary systems

framework'. Appendix in Stanley N. Salthe, Development and Evolution: Complexity andChange in Biology, 291-308. Cambridge, MA and London: MlT Press.

Lukacs, Gyorgy 1980 [1978]. The Ontology ofSocial Being 3: Labour. Trans. David Fernbach.London: Merlin.

Lukacs, Gyorgy 1981. Per l'ontologia dell'essere sociale II: i complessiproblematici piu importanti.Italian trans. Alberto Scarponi. Rome: Editori Riuniti.

Lukacs, Gyorgy 1990. Prolegomeni all'ontologia dell'essere sociali: questioni di principio di unontologia oggi divenuta possibile. Italian trans. Alberto Scarponi. Milan: Guerini eAssociati.

Luria, A. R. 1976. Cognitive Development: Its Cultural and Social Foundations. Trans. MartinLopez-Moril1as and Lynn Solotaroff, Michael Cole (ed.). Cambridge, MA: HarvardUniversity Press.

Manghi, Sergio 1988. '11 soggetto ecosistemico: Identita e complessitii. biosociale'. InFerruccio Andolfi (ed.), Figure d'identita: ricerche sul soggetto moderno, 173-227. Milan:Franco Angeli.

Malinowski, Bronislaw 1923. 'The problem of meaning in primitive languages'.Supplement I in C. K. Ogden and I. A. Richards, The Meaning of Meaning. London:Kegan Paul Trench.

Malinowski, Bronislaw 1935. 'An ethnographic theory of language. Part IV'. In CoralGardens and their Magic, Vo\. 2. London: Alien & Unwin.

Marshal1, J. C. 1980. 'Clues from neurological deficits'. In U. Bel1ugi and M. StuddertKennedy (eds), Signed and Spoken Language: Biological Constraints on Linguistic Form,275-90. Weinheim and Deerfield Beach, Florida: Verlag Chemie.

Martin,James R. 1991. 'Intrinsic functionality: implications for contextual theory'. SocialSemiotics 1(1): 99-162.

Martin, James R. 1992a. English Text: System and structure. Philadelphia/Amsterdam: JohnBenjamins.

Martin, James R. 1992b. 'Macro-proposals: meaning by degree'. In W. C. Mann and S. A.Thompson (eds), Discourse Description: Diverse Analyses of a Fundraising Text, 359-96.Amsterdam and Philadelphia: Benjamins.

Martinec, Radan 2000. 'Rhythm in multimodal texts'. Leonardo 4,3: 289-97.Martinec, Radan 1998. 'Cohesion in action'. Semiotica 120(1/2): 161-80.Matthiessen, Christian 1990. 'Metafunctional complementarity and resonance in syntag

matic organization'. Department of Linguistics, University of Sydney: Mimeo.Matthiessen, Christian 1993. 'The object of study in cognitive science in relation to its

construal and enactment in language'. In M. A. K. Halliday (guest ed.), Language asCultural Dynamic. Cultural Dynamics 6 (1) .

Matthiessen, Christian 1995. Lexicogrammatical Cartography: English systems. Tokyo: International Language Sciences Publishers.

Maze, John R. 1991. 'Representationism, realism and the redundancy of "mentalese'",Theory & Psychology 1(2): 163-85.

McGregor, William B. 1997. Semiotic Grammar. Oxford: Clarendon.McKel1ar, Gordon Bruce 1987. 'The place of socio-semiotics in contemporary thought'. In

Ross Steele and Terry Threadgold (eds), Language Topics: Essays in Honour of MichaelHalliday, Vo\. 11, 523-48. Amsterdam and Philadelphia: John Benjamins.

McKellar, Gordon Bruce 1990. 'The language of neurolinguistics: principles and perspectives in the application of linguistic theory to the neuropsychology of language'.In M. A. K. Halliday, John Gibbons, and Howard Nicholas (eds), Learning, Keepingand Using Language, Vol. 1. Selected Papers from the 8th World Congress of AppliedLinguistics, Sydney, 16-21 August 1987, 331-51. Amsterdam and Philadelphia: JohnBenjamins.

326 REFERENCES

McNeill, David 1992. Hand and Mind: What Gestures Reveal about Thought. Chicago andLondon: University of Chicago Press.

Merleau-Ponty, Maurice 1983 [1942]. The Structure of Behavior. Trans. Alden L Fisher.Pittsburgh: Duquesne University Press.

Natsoulas, Thomas 1998. 'The case for intrinsic theory: Ill. Intrinsic inner awareness and theproblem of straightforward objectivation'. TheJoumal ofMind and Behavior 19(1): 1-19.

Natsoulas, Thomas 2000. 'On the intrinsic nature of states of consciousness'. Consciousness& Emotion I (I): 139-66.

Ninio, A. and Snow, C. E. 1996. Pragmatic Development. Boulder: Westview.Norton, Alec 1995. 'Dynamics: an introduction'. In Robert F. Port and Timothy van Gelder

(eds), Mind as Motion: Explorations in the Dynamics of Cognition, 45-68. Cambridge, MAand London: MIT Press.

Noth, Winfried 1990. Handbook of Semiotics. Bloomington and Indianapolis: IndianaUniversity Press.

Oldenburg, Jane 1987. 'From Child Tongue to Mother Tongue: a case study of languagedevelopment in the first two and a half years'. Ph.D. thesis. University of Sydney,Department of Linguistics.

O'Toole, Michaell994. The Language ofDisplayed Art. London: Leicester University Press.Painter, Clare 1984. Into the Mother Tongue: A Case Study in Early Language Development. Open

Linguistics Series. London: Pinter.Peng, Fred C. C. 1994. 'Language disorders and brain function'. Acta Neurologica Sinica

3(3): 103-30.Peng, Fred C. C. 2000. 'Review article: "Re-reading Saussure: the dynamics of signs in social

life" by Paul J.Thibault'. Lingua Posnaniensis XLII: 175-98.Perinat, Adolfo and Sadurni, Marta 1999. 'The ontogenesis of meaning: an interactional

approach'. Mind, Culture, and Activity 6(1): 53-76.Petitot-Cocorda,Jean 1985. Les catastrophesde la parole:de RnmanJakobson it Rene Thom. Paris:

Maloine.Piaget,Jean 1946. La formation du symbolechez l'enfant. Neuchatel: Delachaux et Niestlee.Pike, Kenneth L. 1967. Language in Relation to a Unified Theory of the Structure of Human

Behaviour. 2nd, rev. edn. The Hague and Paris: Mouton.Pike, Kenneth L. 1982. Linguistic Concepts: An Introduction to Tagmemics. Lincoln: University

of Nebraska Press.Pinker, Stephen 1994. The Language Instinct: The New Science of Language and Mind. Har-

mondsworth: Penguin.Pinker, Stephen 1997. How the Mind Works. London: Norton.Pitt, V. H. (ed.) 1977. The Penguin Dictionary ofPhysics. Harmondsworth: Penguin.Popper, K. R. and Eccles, J. C. 1977. The Self and its Brain. Heidelberg and New York:

Springer.Premack, D. and Woodruff, G. 1978. 'Does the chimpanzee have a theory of mind?' Behav

ioural and Brain Sciences1: 515-26.Prevignano, Carlo L. and Di Luzio, Aldo 2003. 'A Discussion with John J. Gumperz'.

In Susan L. Eerdmans, Carlo L. Prevignano, and Paul J. Thibault (eds), Language andInteraction: Discussions with John J Gumperz; 7-29. Amsterdam/Philadelphia: JohnBenjamins.

Prigogine, lIya and Stengers, Isabelle 1985 [1984]. Order out of Chaos: Man's New Dialoguewith Nature. London: Fontana.

Prodi, Giorgio 1977. Le basi materiali della significazione. Milan: Bompiani.Prodi, Giorgio 1987. Alia radice del comportamento morale. Genoa: Marietti.Prodi, Giorgio 1989. L'individuo e la sua firma. Bologna: 11 Mulino.Riegel, Klaus F. 1979. Foundations ofDialectical Psychology. New York and London: Academic

Press.Rizzolatti, G. and Arbib, M. A. 1998. 'Language within our grasp'. Trends in Neuroscience

21(5): 188-94.

REFERENCES 327

Robins, RH. 1957. 'Aspects of prosodic analysis'. Proceedings of University of Durham Philosophical Society, series B, Vo!. 1 (Arts): 1-12.

Salthe, Stanley N. 1990. 'Sketch of a logical demonstration that the global informationcapacity of a macroscopic system must behave entropically when viewed internally'.Journal ofIdeas 1:51-6.

Salthe, Stanley N. 1993. Deoelopment and Evolution: Complexity and Change in Biology.Cambridge, MA and London: MIT Press.

Sartre,Jean-Paul 1969 [1943]. Being and Nothingness: An Essay on Phenomenological Ontology.Trans. Hazel E. Barnes. London: Methuen.

Saussure, Ferdinand de 1971 [1915]. Cours de linguistique generate. Paris: Payot.Saussure, Ferdinand de 1993. Eisuke Komatsu (ed.), Cours de linguistique generate:premier et

troisiemecours d 'apresles notes de Reidlinger et Constantin. Collection Recherches UniversiteGaskushuin no. 24. Tokyo: Universite Gakushuin.

Savage-Rumbaugh, Sue, Shanker, Stuart G., and Taylor, TalbotJ. 1998. Apes, Language andthe Human Mind. Oxford: Oxford University Press.

Schank, RC. and Abelson, RP. 1977. Scripts, Plans, Goals, and Understanding. Hillsdale, NJ:Erlbaum.

Seager, William 1999. Theories of Consciousness:An Introduction and Assessment. London andNew York: Routledge.

Shanker, Stuart G. 1996. 'Autism and the theory of mind'. Department of Philosophy andPsychology, York University, Toronto: Mimeo.

Shanker, Stuart G. 1998. 'Philosophical preconceptions'. In Savage-Rumbaugh, Sue,Shanker, Stuart G., and Taylor, Talbot J. 1998. Apes, Language and the Human Mind,77-138. Oxford: Oxford University Press.

Shannon, Claude E. and Weaver, Warren 1949. The Mathematical Theory of Communication.Urbana, IL: University of Illinois Press.

Shepard, R N. 1984. 'Ecological constraints on internal representation: resonantkinematics of perceiving, imagining, thinking, and dreaming'. PsychologicalReview 91:417-47.

Sherrington, C. S. 1940. Man on his Nature. Cambridge: Cambridge University Press.Silverstein, Michael1976. 'Shifters, linguistic categories, and cultural description'. In K. H.

Basso and H. A. Selby (eds), Meaning in Anthropology, 11-55. Albuquerque: Universityof New Mexico Press.

Silverstein, Michael 1985. 'The functional stratification of language and ontogenesis'. InJames V. Wertsch (ed.), Culture, Communication, and Cognition: Vygotskian Perspectives,205-35. Cambridge and Melbourne: Cambridge University Press.

Silverstein, Michael 1987a. 'Cognitive implications of a referential hierarchy'. In MayaHickmann (ed.), Social and functional Approaches to Language and Thought, 125-64. NewYork and London: Academic Press.

Silverstein, Michael 1987b. 'The three faces of "function": preliminaries to a psychology oflanguage'. In Maya Hickmann (ed.), Social and Functional Approaches to Language andThought, 17-38. New York and London: Academic Press.

Silverstein, Michael 1997. 'The improvisational performance of culture in realtimediscursive practice'. In R Keith Sawyer (ed.), Creativity in Performance, 265-312.Greenwich, CT: Ablex.

Sperry, R W. 1952. 'Neurology and the mind-brain problem'. American Scientist 40: 291-312.Sperry, R W. 1980. 'Mind-brain interaction: mentalism yes; dualism no'. Neuroscience 5:

195-206.Stevens, K. N. 1989. 'On the quantal nature of speech'. Journal ofPhonetics 17: 3-45.Studdert-Kennedy, M. and Lane, H. 1980. 'Clues from the differences between signed and

spoken language'. In U. Bellugi and M. Studdert-Kennedy (eds), Signed and SpokenLanguage: Biological Constraints on Linguistic Form, 29-39. Weinheim and DeerfieldBeach, Florida: Verlag Chemie.

Thelen, Esther 1995. 'Timescale dynamics and the development of an embodied

328 REFERENCES

cognition'. In Robert F. Port and Timothy van Gelder (eds), Mind as Motion: Explorationsin theDynamics of Cognition, 69-100. Cambridge, MA and London: MIT Press.

Thelen, Esther and Smith, Linda B. 1994. A DynamicSystems Approach to theDevelopment ofCognition and Action. Cambridge, MA and London: MIT Press.

Thibault, Paul]. 1986a. 'The cognitive hypothesis: a critical comment'. In Text, Discourse,and Context: A Social Semiotic Perspective, 26-45. Victoria University, Toronto:Monographs, Working Papers and Prepublications of the Toronto Semiotic Circle,Vo!' 3.

Thibault, Paul]. 1986b. 'Thematic system analysis and the construction of knowledge andbelief in discourse: the headlines in two Italian newspaper texts'. In Text, Discourse, andContext: A SocialSemiotic Perspective, 44-91. Victoria University, Toronto: Monographs,Working Papers and Prepublications of the Toronto Semiotic Circle, Vo!' 3.

Thibault, Paul]. 1987. 'Interview with Michael Halliday'. In Ross Steele and TerryThreadgold (eds), Language Topics: Essays in Honour of MichaelHalliday, Vol. II,601-27.Amsterdam and Philadelphia: Benjamins.

Thibault, Paul]. 1989. 'Semantic variation, social heteroglossia and intertextuality:thematic and axiological meanings in spoken discourse'. Critical Studies1(2): 181-209.

Thibault, Paul J. 1990. 'Questions of genre and intertextuality in some Australian televisionadvertisements'. In R. Rossini Favretti (ed.), The Televised Text, 89-131. Bologna:Patron.

Thibault, Paul]. 1991a. SocialSemiotics as Praxis: Text, Social Meaning Making and Nabokov's'Ada: Theory and History of Literature series, Vo!' 74, Wlad Godzich and JochenSchulte-Sasse (eds), Minneapolis and Oxford: University of Minnesota Press.

Thibault, Paul]. 1991b. 'Grammar, technocracy and the noun: technocratic values andcognitive linguistics'. In Eija Ventola (ed.), Recent Systemic and OtherFunctionalViews onLanguage,281-305. Trends in Linguistics Studies and Monographs. Berlin: Mouton deGruyter.

Thibault, Paul]. 1992a. 'Minor clauses and interpersonal meaning'. Dipartimento diScienze del Linguaggio, Universita degli Studi di Venezia: Mimeo.

Thibault, Paul J. 1992b. 'Grammar. ethics, and understanding: functionalist reason andclause as exchange'. SocialSemiotics 2(1): 135-75.

Thibault, Paul]. 1993. 'Using language to think interpersonally: experiential meaning andthe cryptogrammar of subjectivity and agency in English'. In M. A. K. Halliday (guested.), Language as CulturalDynamic. CulturalDynamics 6(1-2): 131-86.

Thibault, Paul]. 1994. 'Text and/or context?: an open question'. State-of-the-Art article. InThe Semiotic Reviewof Books (Toronto) 5(2), May 1994.

Thibault, Paul]. 1995. 'The interpersonal grammar of Mood and the ecosocial dynamicsof the semiotic exchange process'. In Ruqaiya Hasan and Peter Fries (eds), On Subjectand Theme: From thePerspective of Functions in Discourse, 51-89. Amsterdam and Philadelphia: Benjamins.

Thibault, Paul]. 1997a. Re-reading Saussure: TheDynamics of Signs in Social Life. London andNew York: Routledge.

Thibault, Paul]. 1997b. 'The sounds of language'. Saussure and Beyond, Lecture No. 4.Cyber Semiotics Institute (University of Toronto), website: http://www.chass.utoronto.ca/epc/srb.

Thibault, Paul]. 1998a. 'Embodiment, perception, consciousness, personhood: Thecascading/collecting dialectic of langue and parole in the individual'. Cyber SemioticsInstitute, University of Toronto (Canada); website: http://www.chass.utoronto.ca/epc/srb.

Thibault, Paul]. 1998b. 'Metasemiosis'. SRB Insights. The Semiotic Reviewof Books 9(2), May1998: 8-12.

Thibault, Paul]. 1998c. 'Inner speech'. In Paul Bouissac (ed.), TheEncyclopedia of Semiotics,312-14. Oxford and New York: Oxford University Press.

Thibault, Paul]. 1999a. 'Communicating and interpreting relevance through discourse

REFERENCES 329

negotiation: an alternative to relevance theory'. Journal ofPragmatics 31: 557-94.Thibault, Paul J. 1999b. 'Putting Humpty Dumpty's theory of meaning back together again:

can Saussure help?' Belgian Essays on Language and Literature 9: 7-34. Liege: BelgianAssociation of Anglicists in Higher Education.

Thibault, Paul]. 2000a. 'The dialogical integration of the brain in social semiosis: Edelmanand the case for downward causation'. Mind, Culture, and Activity 7(4): 291-311.

Thibault, Paul]. 2000b. 'The multimodal transcription of a television advertisement:theory and practice'. In Anthony Baldry (ed.), Multimodality and Multimediality in theDistance Learning Age, 311-85. Campo Basso: Lampo.

Thibault, Paul]. 2001. 'Multimodality and the school science textbook'. In Carol TaylorTorseIlo, Giuseppe Brunetti, and Nicoletta Penello (eds), Corpora testuali per ricerca,traduzione e apprendimento linguistico, 293--335. Padua: Unipress.

Thibault, Paul]. 2002. 'Interpersonal meaning and the discursive construction of action,attitudes and values: the Global Modal Program of one text'. In Peter Fries, MichaelCummings, David Lockwood, and William Spruiell (eds), Relations and Functions withinand around Language, 56-116. London and New York: Continuum.

Thibault, Paul]. 2003a. 'Body dynamics, social meaning-making, and scale heterogeneity:reconsidering contextualization cues and language as mixed-mode serniosis'. In SusanEerdmans, Carlo Prevignano, and Paul]. Thibault (eds), Discussing John J Gumperz,127-47. Amsterdam/Philadelphia: Benjamins.

Thibault, Paul]. 2003b. 'Contextualization and social meaning-making practices'. In SusanEerdmans, Carlo Prevignano, and Paul]. Thibault (eds), Discussing John J Gumperz;41-61. Amsterdam/Philadelphia: Benjamins.

Thibault, Paul]. 2004. 'Agency, individuation, and meaning-making: reflections on anepisode of bonobo-human interaction'. In Geoffrey Williams and AnnabeIle Lukin(eds), Language Development: Functional Perspectives on Species and Individuals 112-36.London and New York: Continuum.

Thibault, Paul.J. In press a [2004]. 'Semiology'. In Jeff Stepnisky and Andrew Wernick(eds), The Encyclopaedia ofSocial Theory, London and Newbury Park, CA: Sage.

Thibault, Paul.J. In press b. Agency and Consciousness in Discourse: Self-other Dynamics as aComplex System, London and New York: Continuum.

Thibault, Paul J. and Van Leeuwen, Theo 1996. 'Grammar, society, and the speech act:renewing the connections'. Journal ofPragmatics 25: 561-85.

Threadgold, Terry 1997. Feminist Poetics. London and New York: Routledge.Togeby, Ole 2000. 'Anticipated downward causation and the arch structure of texts'. In

Peter Begh Andersen, Claus Emmeche, Niels Ole Finnemann, and Peder VoetmannChristiansen (eds) , Downward Causation: Minds, Bodies and Matter, 261-77. Aarhus:Aarhus University Press.

Trager, G. 1.. and Smith, H. 1.. 1951. Outline ofEnglish Structure. Norman, OK: Battenburg.Trevarthen, Colwyn 1978. 'Modes of perceiving and modes of acting'. In Herbert 1.. Pick

and E!liot Saltzman (eds), Modes of Perceiving and Processing Information, 99-136.Hillsdale, ~: Erlbaum.

Trevarthen, Colwyn 1980. 'The foundations of intersubjectivity: Development of interpersonal and cooperative understanding in infants'. In D. Olsen (ed.), The Social Foundations ofLanguage and Thought, 316-42. New York: Norton.

Trevarthen, Colwyn 1987. 'Sharing makes sense: intersubjectivity and the making of aninfant's meaning'. In Ross Steele and Terry Threadgold (eds), Language Topics: Essaysin Honour ofMichael Halliday, Vo!. I, 177-99. Amsterdam/Philadelphia: Benjamins.

Trevarthen, Colwyn 1992. 'An infant's motives for speaking and thinking in the culture'. InAstri Heen Wold (ed.), The Dialogical Alternative: Towards a Theory ofLanguage and Mind,99-137. Oslo: Scandinavian University Press.

Van Leeuwen, Theo 1999. Speech, Music, Sound. London: Macmillan.Varela, F. and Maturana, R. 1980. Autopoiesis and Cognition. Dordrecht: ReideJ.Vihman, Marilyn May 1991. 'Ontogeny of phonetic gestures: speech production'. In

330 REFERENCES

Ignatius C. Mattingly and Michael Studdert-Kennedy (eds), Modularity and the MotorTheory of Speech Perception, 69--84. Hillsdale, !'U: Erlbaum.

Vihman, Marilyn and de Boysson-Bardies, Benedicte 1994. 'The nature and origins ofambient language influence on infant vocal production and early words'. Phonetica 51:159-69.

Vihman, Marilyn May, Mackek, Marlys A., Miller, Ruth, Simmons, Hazel, and Miller,jim1985. 'From babbling to speech: a re-assessment of the continuity issue'. Language61(2): 397-445.

Volosinov, V.N. 1973 [1930]. Marxism and thePhilosophy ofLanguage.Trans. Ladislav Matejkaand I. R. Titunik. New York and London: Seminar Press.

Volosinov, V. N. 1976 [1927]. Freudianism: A Marxist critique. Trans. and ed. I. R. Titunik andNeal H. Bruss. New York and London: Academic Press.

Volosinov, V. N. 1983. 'The construction of the utterance'. Trans. Noel Owen. In AnnShukman (ed.), Bakhtin School Papers: Russian Poetics in Translation, No. 10, 114-37.Oxford: RPT Publications (in association with Department of Literature, University ofEssex, Colchester).

Vygotsky,Lev 1986 [1934]. Thought and Language.Trans. Alex Kozulin. Cambridge, MA andLondon: MIT Press.

Vygotsky, L. S. 1987 [1934]. 'Thinking and speech'. In R. W. Rieber and A. S. Carton (eds) ,The Collected Works of L. S. Vygotsky, Vot. 1: Problems of General Psychology, 37-285. NewYork: Plenum Press.

Walkerdine, Valerie 1997. 'Redefining the subject in situated cognition theory'. In DavidKirshner and j ames A. Whitson (eds) , Situated Cognition: Social, Semiotic, and PsychologicalPerspectives, 57-70. Mahwah, NJ: Lawrence Erlbaum.

Wall, Patrick D. 1996. The placebo effect'. In Max Velmans (ed.), The Science of Consciousness: Psychological, Neuropsychological and ClinicalReviews, 162-80. London and NewYork:Rout/edge.

Wells, Cordon 1999. Dialogic Inquiry: Towards a Sociocultural Practice and Theory of Education.Cambridge and New York: Cambridge University Press.

Wertsch,James V.1991. Voices of theMind: A Sociocultural Approach to Mediated Action.Londonand Singapore: Harvester Wheatsheaf.

Wheeler, John Archibald 1988. 'World as system self-synthesized by quantum networking'.IBMJoumal of Research Development 32(1): 1-12.

Whitson, James A. 1997. 'Cognition as a semiosic process: from situated mediation tocritical reflective transcendence'. In David Kirshner and James A. Whitson (eds),Situated Cognition: Social, Semiotic, and Psychological Perspectives, 97-149. Mahwah, !'U andLondon: Lawrence Erlbaum.

Whorf, Benjamin Lee 1956a [1939]. 'The relation of habitual thought and behaviour tolanguage'. Injohn B. Carroll (ed.), Language, Thought, & Reality,134-59. Cambridge,MA: MIT Press.

Whorf, Benjamin Lee 1956b [1941]. 'Languages and logic'. In John B. Carroll (ed.),Language, Thought, & Reality, 233-45. Cambridge, MA: MIT Press.

Wilden, Anthony 1980 [1972]. System and Structure: Essays in Communication and Exchange.2nd edn. London: Tavistock.

Wilden, Tony 1981. 'Semiotics as praxis: strategy and tactics'. Recherches Semiotiques/SemioticInquiry 1(1): 1-34.

Wilson, Elizabeth A. 1998. Neural Geographies: Feminism and the Microstructure of Cognition.London and New York: Rout/edge.

Zelazo, Philip D. 1999. 'Language, levels of consciousness, and the development of intentional action'. In Philip D. Zelazo, Janet W. Astington, and David R. Olson (eds) ,Developing Theories of Intention: SocialUnderstandingand Self-control, 95-117. Mahwah, NJ:Lawrence Erlbaum.

Name Index

Abelson, R P. 7Abercrombie, David 121, 123, 124, 125Andersen, Peter B0gh 5Arbib, M. A. 267,268,269,270Aristotle 183Armstrong, David F. 72, 80,90, 91Astington,j. W. 174

Bakhtin, Mikhail M. 97, 183,211, 215,270Baron-Cohen, S. 174Bates, Elizabeth 37Bateson, Gregory 8, 13, 14, 23, 24, 26, 27,

34,87,148,172,173,192,282,300,301Battacchi, Marco W. 174,213,246,247Battistelli, Piergiorgio 174Beaugrande, Robert de 48, 308Bellugi, Ursula 190Benson, James 38Bernstein, Basil 279Bernstein, N. 79,82Berthoz, Alain 74,75,77,79,81,83,105Bhaskar, Roy 31,32Bickhard, Mark H. 41,121,136Bogdan, Radul. 146,205,206,207,209,247Bohr, Nielsen 33, 35Bolinger, Dwight 92, 98, 99Bouissac, Paul 34, 40Bourdieu, Pierre 279Boysson-Bardies, Benedicte de 85, IIIBraten, Stein 52Browman, Catherine 90,91,92,145,162Biihler, Kad 194, 197, 200, 221

Campbell, Donald T. 41, 121, 136Carruthers, Peter 262,263,264,271,273,

274, 275, 277, 278Catford, j. C. 122Celani, Giorgio 174Chabris, C. F. 105

Chomsky, Noam 7, 80, 226, 282, 283Clark, E. 205Clark, H. 205Cleirigh, Chris uz, ll3, ll5, ll9, 120,

121, 122, 123, 124, 145, 154, 160,295Cole, Michael 207,216Coleridge, Samuel Taylor 81Corina, David P. 190Coulthard, Malcolm 84, 85Cowley, Stephenj. 40,44,45,124,190,283Cremonini, Giorgio 97

Damasio, Antonio 55,147,148,177,210,248, 249, 250, 251, 252, 253, 254, 255,256, 257, 258, 259, 260, 261, 264, 275

Davidse, Kristin 70,132,167,198,199Davidson, Rj. 189Deacon, Terrence 16,55, 126, 173, 174,

175,190,191,219,276,282Descartes, Rene 224, 225, 226, 302, 303,

305, 307Di Luzio, Aldo 115Dory, R 172Dre~ke,E 205,224

Eccles, j. C. 172Edelman, Gerald M. 19,20,51,52,55,62,

71,84,88,92,101,139,140,141,144,146, 160, 163, 164, 176, 179, 184, 185,209,210,214,215,220,221,228,231,238,242,246,255,256,258,262,264,273,285,302,310,317

Fadiga, L. 268Fillmore, Charles 269Finnemann, Niels Ole 20, 24, 29Firth,j. R 108, ll5, ll7, ll9Fitch, W. Tecumseh 7Flohr, Hans 56, 177, 283, 284, 285, 286,

332 NAME INDEX

287,288,289,290,296,297,301,310,317

Fowle~ Carol 86,89,93,101,102,103Frasnedi, Fabrizio 97Fries, Peter 38

Gee,james Paul 3Gendlin, Eugene 23Gibson,jamesJ. 8,11,12,13,14,60,86,

87,88,102,121,152,185,186,202,218,219,220,222,228,229,230,234,246,247,276,291,296,303,304,306

Gillett, Grant 177Goldfield, Eugene C. 141Goldstein, Louis 90,91,92,145,162Goodwin, Charles 3Creaves, William 38Grene, Marjorie 11, 12Griisser,O.J. 106Gumperz,johnJ. 115

Haken, H. 245Halliday, M. A. K 5,23,31, 37,40,42,46,

47,48,50,55,63,84,94,95,99,107,117, 127, 130, 134, 154, 156, 162, 163,164,166,167,176,194,195,196,197,198,199,200,202,203,205,206,207,212,214,226,239,240,241,242,243,244,245,257,266,269,270,278,282,291, 292, 295, 298, 300, 301, 315

Handel, Stephen 120, 124, 266Harre, Rom 6,12,38,176,177,178,213Harris, P. 174Harris, Roy 44, 89Hasan, Ruqaiya 6,156,200,214,279Hauser, Marc D. 7Hebb, D. O. 285, 286, 296Heisenberg, W. 35Hjelmslev, Louis 40,54,59,60,61,65,

66,67,69,70,104,116,117,126,236,292

Iedema, Roderiek A. M. 77Iwamoto, Kazuyoshi 38

jacobs, R. A. 105jakobson, Roman 190jespersen,Otto 66johnson, Mark 165, 181johnson, M. H. 52johnston, Trevor 47,186,192juarrero, Alicia 4, 13, 16, 17, 19,20,63,

138, 254, 269, 302

Kanzi 38Kauffman, Stuart A. 103, 104Kaye, Kenneth 37Kelso,J. A. S. 84,90,141Kinsbourne, Mareel 4, 17, 18,21,22,52,

271,310,311,312,317Koenig, O. 105Kosslyn, S. M. 105Kress, Gunther 47, 192

Lakoff, George 165, 181Lamb, Sydney M. 40, 292Lane, H. 91, 102, 105Langaeker, Ronald W. 75,76,156,198,

271, 272Latour, Bruno 77, 186Lave,jean 3Laver,john 84,86Lernke, jay L. 3,8,9, 16,20,22,23,25,26,

27,30,34,39,40,41,43,44,47,48,54,65,70,72,82,109,118,137,141,146,151,178,181,182,186,187,192,193,207,216,237,251,252,255,256,264,265,267,282,293,294,295,300,304,306,308,309,312,313

Liberman, A. 100,268Libet, Benjamin 171, 172, 173, 174,

180Lieberman, Philip 84, 85, 86Lindblom, Bjorn 93, 94Longaere, Robert E. 95, 97Lorenz, Konrad 213Lorenzana, juan Alvarez 142Lukaes, Gyorgy 99,216,236,237,317Luria, A. R. 188,216

Malinowski, Bronislaw 188, 194Manghi, Sergio 3Marshall,J. C. 73,74Marsolek, C.J. 105Martin,jamesR. 40,117,142,215Martinec, Radan 78,79,93,151,192Martinet, Andre 292Matthiessen, Christian 6, 46,94, 95, 97,

196, 197, 239, 298Mattingly, I. 100Maze.john R. 11MeGregor, William B. 127,269MeKellar, Gordon Bruee 33, 38, 48MeNeill, David 186Merleau-Ponty, Mauriee 78, 100Morton, J. 52

NAME INDEX 333

Natsoulas, Thomas 233Ninio, A. 205Norton, AJec 145Noth, Winfried 34

Oldenburg,jane 240Olson, D. R. 174O'Toole, Michael 192

Painter, Clare 239Panbanisha 38Peirce, Charles Sanders 34,35, 116, 126,

249, 305Peng, Fred C. C. 190, 217, 218, 254, 268,

271,272,289,290,294Perinat, Adolfo 213Petitot-Cocorda,jean 87Piaget,jean 213Pike, Kenneth L. 94, 125, 134Pinker, Stephen 283Pitt, V. H. 129Popper, K. R. 172Premack, D. 174Prevignano, Carlo 115Prigogine, Ilya 10, 178, 182Prodi, Giorgio 16,35,38,87, 186

Riegel, Klaus F. 176,183Rizzolatti, G. 267, 268, 269, 270Robins, R. H. 94Rosenblum, Lawrence D. 101, 102, 103

Sadumi, Marta 213Salthe, S. N. 3,4,8, 10, 14, 16,20,25, 39,

40,43,54,55,61,62,63,70,74,75,77,79,80,82,83,86,88,106,112,114,118,126,145,147,148,159,161,176,177,181, 182, 183, 187, 190, 206, 216, 224,226, 227, 230, 232, 237, 239, 250, 255,257,261,265,287,288,311

Saltzman, E. L. 84,90, 141Santilli, Kathy 190Sartre, jean-Paul 176Saussure, Ferdinand de 45,47,49,52,59,

60,61,65,66,69, 133, 184, 189, 236,271

Savage-Rumbaugh, Sue 38, 282Schank, R. C. 7Seager, William 225, 302, 303, 305, 306,

307, 309Shanker, Stuart G. 55, 74, 224, 282Shannon, Claude E. 34

Shepard, R. N. 81Silverstein, Michael 42,194,201,215,255Smith, H. L. 84Smith,LindaB. 10,19,51,52,55,172,

185,210,211,241,242,243,245Snow, C. E. 205Sperry, R. W. 172Stengers,Isabelle 10, 178, 182Stevens, K. N. 141Stokoe, William C. 72, 80, 90, 91Studdert-Kennedy, M. 91, 102, 105

Tagliatela,jared 38Talbot, Taylor]. 282Thelen, Esther 10, 19,51,52,55, 162, 163,

166,172,185,210,211,241,242,243,245

Thibault, Paul]. 3,6,8,19,29,34,38,39,40,47,48,49,52,60,62,72,74,87,90,118126,141,151,154,156,173,181,184,189,192,200,213,214,215,217,237,241,256,275,282,293,301,310

Threadgold, Terry 83Togeby, Ole 17, 144Tononi, Giulio 176, 262Trager, G. L. 84Trevarthen, Colwyn 36,37, 179, 202, 212,

214,240,282Tuller, B. 84,90,141

Vaid,jyotsna 190Van Leeuwen, Theo 47, 192,215Vihman, Marilyn May 85, IIIVolosinov, V. N. 74,241,317Vygotsky, Lev S. 49,216,273,275

Walkerdine, Valerie 3Wall, Patrick D. 203Weaver,Warren 34Wells, Gordon 48Wertsch, lames V. 282Wheeler, John Archibald 282, 306Whitson,james A. 70Whorf, Benjamin Lee 280Wilcox, Sherman E. 72, 80, 90, 91Wilden, Anthony 23, 34, 148, 178, 179,

215, 222Wilson, Elizabeth A. 48Woodruff, G. 174

Zelazo, Philip D. 6

Subject Index

ability, symbolic referential 174abstraction, symbolic 214abstractness, conceptual 70account, hypostatic local 31acoustic cue 266act

articulatory 88dialogic 37, 270indexical 162intentional 270

action 250, 251artefact mediated 216intentional 267

action schema 75, 76actional semiotic 79, 83activity 8, 14, 18

biomechanical 45contextualizing 164distributed 316exploratory 12, 200, 220gestural 105linguistic

motor dimension of 75meaning-making 3, 4, 6, 10, 13, 14, 17,

47, 48, 49, 50, 72, 108motor 185, 217, 218, 268, 271multimodal social 208neural (neuronal) 19,76,171,174,246,

260, 315neuromuscular 14,41,42,43,45,150perceptual 160,303performatory 185, 186reading 81ritualistic 213self-organizing 52semiotically mediated 138sensori-motor 13, 15, 18,81,82, 100,

144,181,203,294,302,309internalized 82

symbolic 213vocal tract (articulatory) 53,65,71,83,

86,88,91,94,100,101,117,125,136, 137, 146, 157, 159, 187

activity-structure 98, 140act-token 270affect 189affordance 4,11,12,13,14,15,16,19,20,

24,36,41,43,60,86,162,164,310agency 15,16,32,33,63,95,97,182agent 4,5,15,45,63,64,65,77,147,183,

217,233,279alterity 316alternative, paradigmatic 98amplification, irreversible act of 306analysis, metalinguistic 44ape-language research 282arrow of time 317artefact 15,16,44,45,51,53

textual 310array

ambient optic 264acoustic 100, 101visual 234

articulation 40,43,83,89,93,120,121,123, 159, 163, 218, 271, 289, 291,292

phoneticgestural theory of 90

articulator 87,89,92,94,101,103,105Artificial Intelligence (AI) 72association 11attention 204, 283

poles of 246, 276attraction, basin of 104attractor 4,19,43,65,103,104,187,209,

239, 245, 266, 269, 270, 279, 280,305, 312

attractor basin 148, 167

SUBJECT INDEX 335

attractor landscape 167,245audition 271awareness 20, 288

integrative field of 311, 312intrinsic 223, 233meta-semiotic 214non-perceptual 178, 229perceptual 276, 300, 308phenomenal 56, 289

babbling 162, 163behaviour, motor 4bifurcation 255binding 172body, social 77bonding, sentimental 205,207bonobo 38boundary condition 14,29,121,167,186,

187,231,232,239,278,289,313,315

Braille 78, 297brain, hemispherical organization of 17,

55brain lesion 22breath group 84, 85, 86, 92, 93, 104Broca's area 140, 267, 268

cascading 261case structure 269category 5,7

digital 222, 225experiential 167,197,212,299,300lexicogrammatical 21observer's 227, 230perceptual-motor 243, 306phonological 221,222, 291, 294, 295schematic 22, 165, 197semantic 298semiotic 82symbolic 225, 226, 228value-producing 189

categorization (category formation) 220,225, 228, 240, 242

conceptual 105, 167, 238, 258, 302linguistic 168perceptual-motor 178,216,274,275,

302phonological 266sensori-motor 21symbolic 231systems of 184, 309

causalityAristotelian 137efficient 24,31, 166

final 145, 183transcategorial 32, 33

causeefficient 49,61,72, 138, 183final 138formal 72, 138material 61, 138

clause 196, 198, 298minor 156

closuredialogic 52functional 236informational 224semiotic 309

eo-adaptation 15eo-articulation 89,91, 104codification 44

meta-linguistic 45coding

analogic 23digital 23

coding orientation 279, 280eo-evolution 15, 67, 177cogent moment 41,112,115,145,182,287cognition 3,7,8,13,14,49,150

computationalist model of 150information-processing model of 238

cognitive science paradigm (cognitivescience) 225, 240

coherence, interactional 215cohesion

lexico-semantic 47phonological 119

communication, analogic 148competence, linguistic 282,283,315complementarity 12,31,33,36,68,82

dialogic 277wave/particle/field 136,148

complexity, organizational 12,244concatenation (phonological) 119condensation, semantic 275consciousness 55

core 147,250, 253, 262dialogic basis of 183extended 147,253,255object of 55primary 184stream of 311trajectory of 182, 183

conditionboundary 78, 79, 84, 100enabling 4,5, 16, 177, 183initial 246initiating 5,19,83,120,220,313

336 SUBJECT INDEX

neural architecture as 106constituency, hierarchical 39constraint 32

context-sensitive 84contextual 4,14,30,63,312ecosocial 24formal 84informational-semiotic 145integrative 105higher-order 16metafunctional 278semiotic 6time-bound 83

construal, symbolic 62content 60, 61content form 64, 67, 69content purport 64content stratum (plane) 46, 61, 63,80, 107content substance 64, 66, 67context 14,165context of situation 155contextualization 27,33,81, 251,252contextualization cue 115contrast, discrete/typological 25control, locus of 315control parameter 159control structure 100, 101cortex

cerebral 283,284,289,294,297frontal 240pre-frontal 185, 260

crypotype 95cycle

articulatory 84, 295respiratory (respiration) 84,93rhythmic 295vocal-tract 160

Darwinist view 24deixis 134

evaluative/modal proximity 132,198,199, 276,277

person 132,199,276,277temporal proximity 132,198,199,276,

277demarcation (phonological) 119depolarization, postsynaptic 287development, language 241developmental dynamic 163developmentallifespan psychology 176dialogue 183,231,232difference 9,20,34,45,51,52,80,132,

185,186,193,195,228system of 49,193

typological-categorial 135, 222value-laden 223, 310

differentiation 12,25,46,52,61,142,187,189, 206, 228

analogue 23categorial 256functional 50of self and other 210sensori-motor 140symbolic 211topological-continuous 168,262typological-categorial 142

discourse 48, 78, 117, 140, 157, 181, 200meta-semiotic 45multimodal 192

discourse semantics 274discourse voice 50discrimination

conceptual 164conceptual-semantic 163

distinctioncategorial 237, 243phonological 53semiotically salient 184typological-categorial 51, 52, 178, 193,

225, 292diversity, metafunctional 86,136,145, 164dualism 313

between physical vs. phenomenal brainstates 283

Cartesian 74, 302mind-body 89mind-brain 302nco-Cartesian 89

dyad 12,19,37,38,49,102,176,179,231,244, 270, 278, 282

mother-infant 35, 161dynamics

ecosocial 9material-physical 9neural 16

egoreception 219emergence 313

developmental 80,206historical 65, 80, 182synchronic 32

emotion 98, 99, 259, 260, 264empathy, phonetic 124, 125enactment, interpersonal 273entextualization 147entrainment 15,53entropy 10, 159epigenetic process 283

SUBJECT INDEX 337

eventdistal 100, 101, 102ecological 87,88, 102environmental 28, 86, 204, 207, 220, 230mental 222neural 251perceptual 93

environment 8, 14ecosocial 3physical-material 6

evolutioncortical 185semiotic 146

exchange 35,37,38,213,228dialogic 178, 207dyadic 211,245energy 227information 231interpersonal 214,215,270of matter, energy and information

meaning 8, 71socio-affective 243

experiencecontextual nature of 32epigenetic 240perceptual 248value-laden 262

explanationcausal 102causal-reductive model of 34

explorationabstract symbolic 185haptic 165interpersonal 217movement based 242sensori-motor 7,212,241

expression 60,61,63, 71expression form 64, 69, 87, 88expression purport 64, 66, 104expression stratum (plane) 54,63, 70, 80,

87, 89, 234expression-substance 62, 64, 66, 68, 87, 88,

91exteroception 219,220,221,246,262

face 185,187,189,210,211feature

distinctive 122paradigmatic 120

feedback, positive 244, 266, 287feedback loop 266,284feeling-state 259, 260, 264field

articulatory 122

deictic 200interpersonal 129, 130, 135phonogenetic 120symbolic 197

figure 128, 196, 197, 198,277,278,297,298

finalization 183, 270Finite 131,132,198,199,216,269,276,

277finiteness 156firstness 34, 35, 93, 249foot 122, 123, 124, 125, 154, 157, 295form 71

graphological/phonological 59frequency, fundamental 103freedom

degrees of 10,74,75,79,83,84, 139,141, 165,228,244, 312

semiotic 215topological 309

friction 74, 75, 77, 85, 94, 217function

brainstem 218"etc. 'higher cognitive 74higher mental 21indexical 115, 119, 120mental 209motor 61,217,218,268,271neural 209

gaze 37,45,185,192,195,201,202,247,248

genre 5,44,45,140,142,188,194,213,239, 279

gesture 26,27, 29, 30, 51, 62, 65, 68, 86,89,91,92,100,101,102,186,192,193,246,266

perceptual (phonological) salience of 94vocal 164

Given 128, 129,299goal-seeking 215goal-state 215grammar 46, 47

formal 46interpersonal 269universal 48

grapheme 118, 234graphology 70ground 201, 202

non-symbolic 226grounding 135,156,158,177,199,211,

247, 260, 262, 277group, neuronal 220

338 SUBJECT INDEX

habitus 5, 74, 279hand 184, 185, 187, 189hemisphere

cortical 189left and right 18,189,190,191,212

heterogeneity, scalar (or scale) 43,54,65,70,97,116,145,157,181,279,293

heteroglossia, social 50, 214hierarchy

combinatorial 142,143constituent 125integration/implication (of iconic,

indexical, and symbolic modes ofsemiosis) 221,223

integrative 177,237presupposition-cum-implication 211,

219,222,223,225,230,240referen tial 126scalar 181, 184, 216,287specification (implication) 126, 195three-level 19, 34, 35, 36, 40, 109, 110,

143,158,181,250,289,311,313holism

interactional-affective 260relational 259

homunculus 20,316homogeneity, scalar (or scale) 43,70

icon 126,135,195,211,222,230,240iconicity 255Ictus 113, 121, 122, 124, 125, 154idea (Descartes) 303, 305image

acoustic 271, 272auditory 76,271,272body 181mental 249,250,251,257,258,265,266,

275, 308unified field of 249

neural 260motor 75,76

imagerymental 76proto-indexical mental 255visual 81

implication, scale of 46impulse

afferent 219efferent 219neural 289, 290

index 126,135,195,202,211,222,230,240

individuation 16,38,140,146infodynamics 55

information 6, 11, 12, 13, 22, 36, 42, 87,89,93,114,115,141,147,152,202,220,221,226,227,245,246,249,264,286,289,291

Dretske's definition of 224environmental 303haptic 261indexical 88macroscopic 226perceptual 60,61,162,262,278stimulus 25, 27, 60, 121, 164, 203, 204,

218,219,220,222,225,228,230,231,233,243,274,306,315

multimodal association of 243information potential 20information-processing 8information theory 34InnenweU 25,61,69,78,175,176,203,217,

227,230,236,251,255,256,279instantiation 133, 156integration

contextual 49, 158, 232phonological 119space-time 81

intent, teleological 100intention 4,17,31,33,36,48,74,100,101,

102communicative 115proximate 138, 139semantically organized 268

intentional content 296intentionality 68interact (semantic/dialogic) 38,215,216,

217interaction

dialogic 36, 68forceful (between body and world) 165,

166, 168non-linear 17

Interaction System 9, 264, 265interdependence, semiotic-material 23interface, bio-kinematic 83internal milieu 261,290,293interpersonal moral order 6, 214interpretance, system of 15, 16,20,25,28,

35,63,71,83,175,178,180,181,223, 224, 226, 231, 232, 233, 242,250,251,255,257,262,287,300,303,304,305,306,307,308,309,315

self as 174interpretant sign 305, 306in terpreter 35intersubjectivity, primary 19, 36, 179, 282

SUBJECT INDEX 339

intonation 85,92,94,98,99,101,113,119,161,295

introception 262invariant

perceptual 230phonological 272, 274visual 234

isolation, ontological 43

judgement 305faculty of 224

kinaesthesis 81,82,186,187,261kinematics 82kinesics 24knowledge

declarative 212procedural 212,213representational 213sensori-motor 213

labour, social division of 99,142,237language

Cartesian view of 74formalist models of 7formal theories (models) of 48, 137metafunctional organization of 164-165metafunctional theory of 195mixed-mode character of 153multifunctional character of 194physical manifestation of 78stratal organization of!stratified nature

of 136, 271stress-timed 123syllable-timed 123

language faculty 7, 73, 74language gene 24langue 49, 60langue interieure 61,69,76lateralization (of brain functions) 190, 191level

focal 14,15, 16integrative 39,127,146,147,148,161,

173,177,201,206,221,248,254,259,262,273,293

lexicogrammar 42, 46, 55, 62, 69, 70, 77,78,83,99,107,126,127,140,205,211,225, 274

as state space 245metafunctional organization of 152, 153symbolic possibilities of 77

limit cycle 103linguist, integrational 44linguistics, systemic-functional 70, 79

logic, propositional 194logogenesis 145,149,150,294,297loop (looping)

Batesonian 250reentrant 207,256,258,288,289

magnetic resonsance imaging (MRI) 171map

first-order neural 252, 253sensori-motor 252

mappingneural 265reentrant 71,209,210,220,240somato-sensory 262

materialism 313eliminative 223

meaningexperiential 47,86,128,196iconic 266interpersonal 47,56,86,212linguistic 47logical 47, 86referential theory of 229symbolic 256textual 47,86, 128, 129

meaning-makingmultimodal 19multiplicative character of 192

meaning-potential 13,239,244Meaning System 264, 265mechanism

causal 10, 73central 73, 74, 78internal 72lower-scalar 72

mediation, semiotic 44, 138, 139, 194,221

memory 214,234,235,253,258,263, 283ecosocial 310long-term 111, 290short-term 262, 263, 264, 277, 278value-category 310value-laden 308

message unit 156, 157meta-discourse 46metafunction 46, 103, 107, 126, 127, 149,

150,191,192,193experiential 124, 125, 196interpersonal 94,124,199,200logical 93, 130, 131,277textual 123, 200, 201

meta-redundancy 25,26,27, 29, 30, 31, 32,56,89,164,176,193,227,230,244,272,293,295,300,309

340 SUBJECT INDEX

meta-redundancy hierarchy 28meta-representation 287, 288, 294micro-function 42, 147, 166, 168, 245mind 6, 7, 8, 175, 188, 225, 229, 282, 302,

305, 313theory of 55, 174, 176, 184

mirror neurons 56,267,268,269,270modality 129, 130, 198,217

semiotic 49modalization 216model

Cartesian-Euclidean-Newtonian 73cybernetic 72

modulation 216mood 129, 130, 132, 201, 205, 215, 269moraicity 114,121,295morpheme 70, 155, 158morphosyntax 43movement, theory of 78multimodality 46mutuality, of animal and environment 12

nativist hypothesis 282naturalism 223necessity, indexical 61,77,85, 134, 230,

300, 316negotiation, discursive 215neo-cortex 185, 187, 218

pre-frontal 215net, neural/neuronal 294, 296

self-organization of 284, 285, 301network

neural 22, 305, 312neuronal 41

neural cell assembly 284, 285, 294neuron 171neuronal group selection 241New 128,129,299,300nominal group 133, 134, 155, 156, 157,

276notation, written 89

observer 77,78,105,222,223,232,233,248, 290

categorial reach of 127observer system 249Onset 112ontogenesis 48, 139, 245, 246, 294ontogeny 230organization

cerebral 21emergent 313heterarchical 21hierarchical 21

metafunctional 18, 19,91, 103, 104phonological 161,295,296

oscillator, nonlinear 84

paradigmatic contrast set 305paralanguage 24pathway

cortieo-bulbar 218, 294developmental 19,316goal-seeking 270

patternafferent stimulus 287mental 249paradigmatic 30syntagmatic 30thematic 116

percept, analogue 225perception 7,11,12,29

allocentric 107ecological theory of 11, 296event 291

periodicity, rhythmical (rhythmic) 84, 85,92,93,104,295,296

periodicity wave 128perspective (view)

field, particle, and wave 121first-person 173externalist 224, 229inner 287,289,297internalist 224internalized 255intra-organism and inter-organism (on

language) 48observational 175observer 41, 137, 287particle 128self-referential 177, 250,255,261,276,

317third-person 173, 257

phase space 42, 121, 151semantic 241

phonation 86,112,160,218,289phoneme 70,91,119,121,222,266phonetics 72phonology 43, 70, 72, 87, 117

metafunctional organization of 152, 153quantal theory of 141

phylogenesis 48, 139pick-up

of information 12, 13, 28perceptual 27, 54, 118

pitch 94,95,97,111,113,119,295pitch fluctuation 124polarization, of self and world 210, 259

SUBJECT INDEX 341

positing, teleological 237positron emission-tomography (PET) 171,

267possibility, symbolic 77potential, bio-kinematic 84power, causal 41practices, meaning-making 9precursor world 213primate 80, 247principle, contextualizing 37Principle of Alternation 20,22,54,56, 109,

137, 294, 312program, central 82prominence, tonic 113, 295proprioception 220,221,246,261,262process

bio-physicaltopological-categorial properties of 82

neurological 19,20context-sensitive nature of 252

neuronal 43self-organizing 49

project 77reductionist 39

proposal 269proposition 130,208,269prosody 95,97, 107, 115, 116, 119

articulatory 86, 92field-like nature of 107gestural 101interpersonal 94,97, 107semantic 95

proto-imperative 211proto-indicative 211proto-language 38,42,55,139,147,156,

161, 163, 167, 168,239,240,243,244

proto-meaning 212,254,265,266,267,289, 290

proto-modalization 38proto-proposition 205, 206, 208proto-self 147,252,254,255proto-semiosis 220,245,251,267

multiple realizability of by neurologicalprocesses 267

proto-Subject 205, 208

quantification 156quantum mechanics 33,35, 121

rank scale 154, 157, 158lexicogrammatical 151phonological 94,114,151,161,295

reactance 95

reading 81realization 70,116,144,182,254,266,312reasoning, syllogistic 188recategorization 141,144,146,163receptor 105

kinaesthetic 74peripheral 203

receptor cell 11, 289, 291, 303, 304receptor organ 25, 61recursion 7, 93reduction, downwards 31, 73reductionism, physical 302, 313redundancy 26,27,28,29,251,301

contextual 15, 36, 165, 187,308reentry 275reference 126

hierarchy of 219symbolic 175

reference scale 62reflection

experiential 273meta-semiotic 233

register 142regulation, sensori-motor 21relation

contextualizing 23,30,34, 176,229,247,265,293,300,310,317

heteroglossic 50relationship

indexical 61,116realizatory/realizational 40,78,274,301

Remiss 121,122,124,154repertoire

conceptual 139, 140, 146phonetic 139sensori-motor 146

representamen 305representation 12, 173, 175,240,307,312,

314mental 7,8

resourcesomatic 77tactic 40value-creating 47

retina 235Rheme 128, 129,299rhetoric 97Rhyme 112,295rhythm 93,94,112,116,119,123,124,

154, 160, 296robustness 64, 65, 233, 257routine

articulatory 271auditory 296

342 SUBJECT INDEX

cogmtive 271,297conceptual 55,138,143,144,160,163,

295lexicogrammatical 295motor 75, 272, 274

imagined 187phonological 274, 295sensori-motor 55,140,141,143,144,

163, 293, 295rule

deterministic 72innate 239linguistic 283

salience, motivational 184, 189scale, space-time 10, 17, 65schemata (schema)

action 165Body-Forceful-Interaction-Environment

165, 166corporeal 82interpersonal action 167movement 83sensori-motor 167

score, gestural 91,92secondness 34,35,75,77,93,94,249R~I~~l~l~l~lnl~l~

183,198,206,220,223,229,247,249,252,254,258,261,265,274,276,288,312,313,317

autobiographical 253experiential unity of 262grounded perspective of 248neurally realized 261qua self-organizing system 34

self-awareness 250, 275, 317selfhood 182self-organization 38, 55, 56, 141, 176, 177,

242,261,262,317self-reference 246sememe 70semiosis 6, 29, 31, 38, 43

iconic, indexical, and symbolicmodalities of

indefinite 306infant 42, 213internalized (inner) 177, 238linguistic

mixed-mode character of 151stratified nature of 233symbolic 262topological 23typological 23, 43

semiotic action formation 44

sensation 223,225,229,291,302,305acoustic 291bodily 258empiricist theories of 222haptic 106

sharing, mental 205, 207sign 34,39,40,52,67,202,303,314

iconic 119indexical 115, 116,243,287indexically creative or entailing 255indexically presupposing 255symbolic 287system of 80, 203

signified 107signifier 59, 60, 62simulation, internal 76, 107solipsism, Cartesian 236,257sound (Saussure) 60,61space 105, 106

allocentric 106attractor 50, 108distant 106egocentric 106extra-corporeal 106extra-personal 106interpersonal neural 269personal 106n-dimensional semantic/semiotic 50,

245multidimensional physiological 142multidimensional semantic 55, 194, 239,

245, 278, 279self-organizing 271semantic neural 17symbolic neural 6, 144, 159, 160, 187,

193,272,278,297,298,317space-time, Newtonian 38specialization, hemispheric 191specification hierarchy 39, 54, 55,99, 254,

255,260,262,266,294,317specification, levels of 38speech, inner 10,81, 187, 189,263,271,

272,273,274,275,297speech function 215state

emotional 98mental 5, 19,22,49,247metastable 182neurophysiological 32, 33

stimulusperceptual 152physical 303sensory 76

stimulus-response dichotomy 230

SUBJECT INDEX 343

stratification 217,274linguistic 39,40,63, 70, 118, 142ontological 33, 34semiotic 43, 233

structuralism 49structure

constituent 122coordinative 141dissipative 10,25,112,146,147,159,

161, 182, 244phonological 85, 90

Subject 132,199,269,299substance, phonetic 59substrate

neural 18neuromuscular 115topological 82, 103, 104

supersystem 160,227,229,238,255,264supervenience 80syllable 84,90,94,112,113,119,120,154,

157,159,222,266,295,296syllable boundary 120syllable weight 160symbol 51,126,135,195,211,222,230,

240internal 240

symbol string 51, 74symmetry-breaking 78symesthesia 81synapse 285,286syntagm 98, 124system

abiotic 24central nervous 13, 15,28,51,227,234,

260,261,290,293,294,297closed 10dynamic open 8,10,11,40,48,49,138,

178,179,207,244,279ecosocial 43evolutionary 142far-from-equilibrium 244goal-seeking 24,216,317higher-scalar 35infant 45language 49,50, 51, 52, 60multimodal 45organism-plus-environment 36perceptual 186peripheral nervous 227, 290, 293, 294phonological 63,111,291,292scalar hierarchical 20, 43selective recognition (brain as) 55, 139,

165

self-organizing 11, 182, 236, 244, 245,310, 313

semiotic resource 237skelotomuscular 75somatic recognition (brain as) 164, 187,

222,234,272,274somato-sensory 261symbolic 300transitivity 79three-level 14three-level hierarchical 15vestibular 261

text 3,16,72,117,154,263,276,277,278,279, 309

texture, phonological 120Theme 128, 129,299theory

breath-group 84homeokinetic 84motor 100, 101, 102movement 90systemic-functional 47,48, 117, 154,238,

273Thermodynamics, First and Second Law of

226Thing 133, 134, 155, 157thinking 188

conceptual 7higher-order 216linguistically realized 271reflexive 277, 278symbolic 189,263verbal 188

thirdness 71,80,93thought (Saussure) 60,61

semiotically mediated 215time

absolute (clock) 295,296evolutionary 240historical-biographical 75real 151relative 295

timescale(s), non-adjacent 43tone group (see also breath group) 84,85,

295tonicity 113, 121, 161,295,296tool

body as 106brain as 293

topic-comment parsing 205topic predication 206, 207, 209trajectory 3,4,5,6,10, 13, 14, 16, 18,20,

344 SUBJECT INDEX

21,50,83,138,141,147,150,159,160, 161, 167, 177, 181, 217, 228,230,242,253,261,270,275,278,313

action 260, 297agentive determinability of 114developmental 275, 283historical-biographical 4, 44, 50, 55, 111,

140, 227, 265individuating 102,147,313lifespan 180logogenetic 146, 159, 310ontogenetic 3,54,182,226rhythmic 112semiogenetic 4,5,25,51social (moral) 176syllabic 115temporal 151vocal-tract 111,112,139

transcranial magnetic stimulation 267transduction 26, 63, 203, 219, 233, 292, 297,

315symbolic 88, 234

transformation 93transitivity role 97, 156, 197, 198type-category 197type-specification 156

Umwelt 11, 12,37, 60, 69, 71, 79, 83, 85,176, 178, 189, 211, 217, 236, 282,287, 293

vagueness 45,46,50,176,180,184,190,207,230,234

valence,affective 204,212value 47,49,51,52,71,88,89, 101, 104,

132,133,184,189,207,209,210,211,212,214,217,220,242,244,245,282,309,316

conceptual 223interpersonal 195social 200symbolic 203

value bias 51,52,53,54value stance 299value system 160variability 11, 29, 50variation (variety), topological-eontinuous

23,25,26,42,52,109,118,125,127,135,161,167,178,190,193, 200,216,218,222,230,234,243,259,292

vocalization 6vocal tract 41,67, 103, 152, 189vowel 40

wave, rhythmic 151Wernicke's area 140

brain mind and the signifying body open linguistics

Documents

language policy

development of language

language pathology

study of language

functional view of language

languagebased approach

language instinct debate

communicative use of