michel buijs - clause combining in ancient greek narrative disourse the distribution of subclauses...

7/29/2019 Michel Buijs - Clause Combining in Ancient Greek Narrative Disourse The Distribution of Subclauses and Participial …

http://slidepdf.com/reader/full/michel-buijs-clause-combining-in-ancient-greek-narrative-disourse-the-distribution 1/290



CLAUSE COMBINING IN ANCIENT GREEK NARRATIVE DISCOURSE



MNEMOSYNEBIBLIOTHECA CLASSICA BATAVA

COLLEGERUNT

H. PINKSTER • H.S. VERSNEL

I.J.F. DE JONG • P.H. SCHRIJVERS

BIBLIOTHECAE FASCICULOS EDENDOS CURAVIT

H. PINKSTER, KLASSIEK SEMINARIUM, SPUISTRAAT 134, AMSTERDAM

SUPPLEMENTUM DUCENTESIMUM SEXAGESIMUM

MICHEL BUIJS

CLAUSE COMBINING IN ANCIENT GREEK NARRATIVE DISCOURSE



CLAUSE COMBINING IN ANCIENT GREEK NARRATIVE

DISCOURSETHE DISTRIBUTION OF SUBCLAUSES AND

PARTICIPIAL CLAUSES IN XENOPHON’SHELLENICA AND ANABASIS

BY

MICHEL BUIJS

BRILLLEIDEN • BOSTON

2005



This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

A C.I.P. record for this book is available from the Library of Congress.

ISSN 0169-8958ISBN 90 04 14250 9

© Copyright 2005 by Koninklijke Brill NV, Leiden, The Netherlands Koninklijke Brill NV incorporates the imprints Brill Academic Publishers,

Martinus Nijhoff Publishers and VSP.

All rights reserved. No part of this publication may be reproduced, translated, stored ina retrieval system, or transmitted in any form or by any means, electronic,mechanical, photocopying, recording or otherwise, without prior written

permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Brill provided that

the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers, MA 01923, USA.

Fees are subject to change.

printed in the netherlands



CONTENTS

Preface ............................................................................................ ix

ChapterOne Introduction ....................................................... 11 StatusQuaestionis ................................................................ 1

1.1 ‘Temporal’SubclausesandVerbalAspect .................. 31.2 Conclusion ................................................................... 9

2 AnalysisofAncientGreekNarrativeDiscourse .................. 92.1 TheNatureofNarrativeDiscourse ............................. 9

2.1.1 TemporalRelations ........................................... 112.2 FactorsOperatingontheChoiceamongAlternative

Expressions ................................................................... 132.2.1 PresentationofRealWorldRelations ............... 132.2.2 TextArticulation ............................................... 152.2.3 InformationProcessing ...................................... 17

3 Aims ..................................................................................... 21

ChapterTwo SixNear-minimalPairs...................................... 24Introduction ............................................................................... 241 NMP1: Hellenica4.3.20-21vs. Agesilaus2.15 ......................... 24

2 NMP2: Anabasis6.2.13-15vs. Hellenica2.4.10-11 ................. 383 NMP3: Hellenica3.4.15-16vs. Agesilaus1.19-25 .................... 434 NMP4: Agesilaus1.38-2.2vs. Hellenica4.2.4-8&4.3.1-4 ....... 535 NMP5: Anabasis7.3.23-26vs.Cyropaedia8.4.12-13 ............... 626 NMP6:Cyropaedia5.4.43-50vs. Hellenica3.4.9-10

(&Cyropaedia1.4.19-20, Anabasis2.5.15-24, Anabasis6.3.11-15,and Hellenica3.2.25-26) ....................................... 72

7 Conclusion ........................................................................... 84

ChapterThree PresentationofRealWorldRelations ............... 85Introduction ............................................................................... 851 SubclausesHeadedbyaSemanticallySpecificRelator ...... 85

1.1 ‘Causal’Relations ........................................................ 85



vi CONTENTS

1.2 TemporalRelations ..................................................... 891.3 Conclusion ................................................................... 111

2 SubclausesHeadedbyaSemanticallyNon-specificRelator ................................................................................. 112

3 ParticipialClausesHeadedbyaRelator ............................. 1294 Conclusion ........................................................................... 134

ChapterFour RealWorldConstructionandTextArticulation:PreposedSubordinateClausesand

DiscourseSegmentation .................................... 137Introduction ............................................................................... 1371 RealWorldConstruction .................................................... 1372 TextArticulationandSegmentation ................................... 138

2.1 SegmentationintheStoryofMania( Hellenica3.1.10-16) ...................................................... 140

3 FormandFunctionofSubordinateClauses ........................ 1483.1 Desententialization ....................................................... 1493.2 (Dis-)ContinuityandLinguisticCoding ....................... 153

4 (Dis-)ContinuityatPointsofSegmentation ......................... 1554.1 ThematicCoherence ................................................... 1594.2 TheArticulationofThematicUnits ............................ 163

4.2.1 TemporalDiscontinuity..................................... 1694.2.2 SpatialDiscontinuity.......................................... 1714.2.3 ReferentialDiscontinuity ................................... 177

4.3 ThematicContinuity .................................................... 1784.3.1 TheGenitiveAbsolute ....................................... 179

4.3.2 TheConjunctParticiple .................................... 1874.3.3 SpecialCasesofThematicContinuity............... 1965 Conclusion ........................................................................... 198

ChapterFive ClauseCombiningandInformationProcessing .......................................................... 201

Introduction ............................................................................... 2011 PostposedEmbeddedPredications ...................................... 201

1.1 Postposed‘Causal’§pe¤/§peidÆ-clauses ...................... 2041.2 Postposed‘Causal’ParticipialClauses ......................... 2141.3 SummaryandConclusion ........................................... 220

2 GenitiveAbsoluteConstructionsvs.ConjunctParticiples .. 2212.1 The‘Regular’ConjunctParticiple ............................... 223



CONTENTS vii

2.2 The‘Irregular’GenitiveAbsoluteConstruction .......... 2262.2.1 TextGrammarvs.SentenceGrammar:

Schwyzer(1942) ................................................. 2332.3 SummaryandConclusion ........................................... 252

ChapterSix Conclusion ......................................................... 254

Bibliography ................................................................................... 261

IndexofLinguisticTerms .............................................................. 269GeneralIndex ................................................................................ 273IndexLocorum .............................................................................. 275



PREFACE

Althoughthesubjectofthisbook,thedifferencebetweenthechoiceof a certain subclause and a participial clause, be it a predicative

conjunct participial clause or a genitive absolute construction, isusually addressed in grammars on Ancient Greek, a systematictreatment is surprisingly lacking. While the discourse-analyticdescription of the different clause types—which focuses on howrelations are coded by means of subordinating conjunctions, thedifferences in form and function as discourse boundary markersbetweenpreposed,sentence-initiallyplacedsubclausesandparticiples,andbetweenclausetypeswithrespecttotheinformationflowinon-goingdiscourse—confinesitself tothehistoriographicnarrativeproseof Xenophon, it does not follow that its findings cannot, mutatismutandis,beexpandedtootherauthorsofthegenrewithintheclassicalperiod. By both this feature and the fact that I have providedtranslationsof thepassagesdiscussed,thebookaimstoberelevanttoall those interested in Greek language description and DiscourseAnalysis,aswellasclassicalphilologistsworkingonXenophonandGreekhistoriansingeneral.

QuotationsfromXenophon’sworksarebasedonE.C.Marchant’s

OxfordClassicalText,buttheTeubnereditionofthe HellenicabyC.Hudeandtheofthe AnabasisbyC.HudeandJ.Peters,whichofferamoreextensiveapparatuscriticus,havebeencheckedthroughout.

The work on this book has been made possible by LeidenUniversity,whoofferedmeapositionasresearchassistantinordertowriteaPh.D.thesis,whichwas,finally,successfullydefendedinmay2003.Regrettably,mysupervisorduringtheproject,Sicking(amansooutstandinginhiswidefieldofexpertisethatwheneverIaskmyself questions on anything pertaining to the Ancient Greek language Ithink‘whatwouldSickinghavesaid?’),didnotlivetoseetheprojectfinished.Icanonlyhopethishasbecomeabookofwhichhewouldhaveapproved.



x PREFACE

Fortunately,hissuccessortothechairofAncientGreekatLeidenUniversity, Professor Ineke Sluiter, so stimulatingly and skillfullyhelpedmethroughthefinalstagesofmywork,thattheresulthasbenefittedconsiderablyfromhercontributions.Iamdeeplyindebtedtoherforherableguidanceandmotivation.

IamalsogratefultoProfessorEgbertBakker,Dr.TijnCuypers,ProfessorAlbertRijksbaron,thelateProfessorSiemSlingsandDr.PeterStorkfortheirkindacceptancetoserveasthemembersontheexaminationcommittee.

Many colleagues, at Leiden University, Utrecht University andelsewhere,supportedmeduringtheyearsIworkedonthisproject.Iwouldliketoacknowledgemygratitudetothemasfriends.SpecialmentiondeserveDr.IljaLeonardPfeijfferforhisstimulatingsarcasm,and Dr. Adriaan Rademaker for his warmth, friendliness and hismoralsupportthroughout.

Iwouldalsoliketoexpressmysincerethankstomyparents,whofinancially supported (or, as we would call it euphemistically,‘sponsored’)myeducationfordecades,andfortheirunderstanding.

Mostofall,IwishtothankSusannahHerman,whonotonlynevergottiredofcorrectingmybadEnglish(whichmeansthatifthereisanyoffenceagainstEnglishgrammarorvocabularyinthisbook,itisdefinitely due tomy ownstubbornness),butalsohelpedmethroughvarious final stages of writing this book. Thank you for being abeautifulfriendtomeinthemanywaysthatyouare.

MichelBuijs

Leiden,November2004



CHAPTERONE

INTRODUCTION

This book is about clause types. Its main goal is to enhance ourunderstandingofthedistributionofsubclausesandparticipialclauses

inAncientGreeknarrativediscoursefromafunctionalperspective,tothe interest of classicists and general linguists alike. It aims atclarifyingthefactorsthat,consciouslyorsubconsciously,mayhaveledanativespeakerinhischoicebetweentheseclausetypes.

Forafairlylongtime,thestudyoflinguisticphenomenahasbeenhamperedbythefactthatittookthesentenceasitsobject.Therearelinguisticfeaturesthatcanbeunderstoodbetter,ifnotonlyso,iftheircontributiontotheconstitutionofpiecesofdiscourselongerthanasentenceis ratedat its truevalue. Thedescriptionof AncientGreekhas suffered from the traditional attitude of approaching linguisticphenomenaonthesentence-level.Thisisduetotheauthorityofthestandard grammars that are still widely in use (Kühner& Gerth;Schwyzer & Debrunner; Goodwin; Smyth; Bornemann & Risch;Gildersleeve;Humbert),allofwhichwerewritteninaperiodinwhichtheapproachthatmightbelabelledastextgrammarhadnotyetbeendeveloped.JudgingfromtheincreasingamountofstudiesonAncientGreek published over roughly the last thirty years that describe

linguisticphenomenaincontextandinvestigatetherelationbetweengrammar and discourse, discourse analysis is beginning to find itsplaceinthedescriptionoftheAncientGreeklanguage.

In this book it will be demonstrated that understanding thedistributionofparticipialclauseson theonehand,andsubclauses(inparticular §pe¤-,§peidÆ-,and…w-clauses)onthe other, requires anapproachinwhichthesentence-boundariesarecrossed.

1 StatusQuaestionis

In clause combining, participial clauses and subclauses aresyntacticallydependentclauses:theyarebothembeddedpredications



2 CHAPTERONE

(often called ‘adverbial clauses’).1 The clause on which they aredependent may be (part of) a main clause or another embeddedpredication.Therefore,thesuperordinateclauseismostconvenientlyreferredtoasmatrixclause.

Althoughbothconjunctandabsoluteparticipialclausesmaysharecertain syntactic characterist ics of subclauses, their formalcharacteristics are different. Nonetheless, participial clauses andsubclausesareusuallyregardedas‘alternatives’.2InordertoexplaintheircoexistenceinAncientGreek,thenotionof‘emphasis’hasoften

been used. Typical is Kühner–Gerth (1904, 78): “Statt derPartizipialkonstruktion können auch im Griechischen entweder desgrösserenNachdruckesoderauchderDeutlichkeitwegenNebensätzegebrauchtwerden”.

Unfortunately, those grammars that make use of the notion‘emphasis’failtogiveanaccountofwhatitmeansexactly.Emphasismaybeattributedtothatwordorphrasewhichissupposedtobemore important orsalient from an informational point of view thanother words or phrases in the surrounding context. But since nocriteriaaresuppliedonthebasisofwhichthepresumedemphasizing functionmaybeattributedtooneofthealternativeexpressions,itwillbeclearthatdescriptionsintermsofemphasisaregratuitous.

Althoughis self-evident thatagenitiveabsolute ismoredetachedfrom the syntactical structure of its matrix clause than a conjunctparticiple, l ittle attention has been paid to genitive absoluteconstructionsandsubclausesasalternativeexpressions.3

———1 Theterm‘participialclauses’coversthegrammaticalclassthatisconstitutedby

predicativeconjunctparticipialclausesandgenitiveabsoluteconstructions.Relativeclauses,attributiveparticiplesandparticipleswiththearticleareoutsidethescopeof thisstudy.

2 This kindofapproach isnotconfined toAncientGreek.Cf. e.g.Stumpin astudyonfreeadjunctsandabsolutesinEnglish(1985:4):“afreeadjunctisanonfinitepredicativephrasewiththefunctionofanadverbialsubordinateclause”.

3 Berent (1973), inanarticleon thehistoricaldevelopmentofsubordinationinIndo-Europeanlanguages,statesthatabsoluteconstructionsbearthesamesemanticrelationstotheirmatrixclauseassubclauses,andthatabsoluteconstructionsandsubclauses,despitetheirformaldifference,arefunctionallyidentical.Heassumesfour

chronological layers of Indo-European subordination of clauses, in the course of whichtheabsoluteconstructionlosesitspositioninfavorofthesubclause.ClassicalGreek,whereabsoluteconstructionsandsubclausesoccurnexttooneanother,wouldbe located in a ‘transitory period’. Regrettably, Berent pays no attention to thequestionhowtheyrelateinsuchasituation.Questionsregardingthemuch-disputedoriginofthegenitiveabsoluteconstructionareoutsidethescopeofthepresentstudy.



INTRODUCTION 3

When scholars address the distribution of genitive absoluteconstructionsandsubclauses,theyofteninformusofhowto translatethem;cf. Schwyzer–Debrunner (1950:397):“Wie z.B. imArischen,erscheinenauch imGriechischenvom Beginnder Überlieferung ansog. absolute Partizipialkonstruktionen, d.h. aus Substantiv + Ptz.bestehende Fügungen, die dem übrigen Teil des Satzes selbständig gegenüberstehen(fürdiegeläufigeÜbersetzungsweiseimWerteinesNebensatzes)”.Orthegenitiveabsoluteissaidtobetheequivalentof asubclause:Bornemann–Risch(1973:245):“WeilauchderGen.abs.

eine Nebenhandlung zum übergeordneten Prädikat ausdrückt,entsprichtseineÜbersetzungdereinesPtc.coni.;dahersindbeideindem§246parallelbehandelt,obwohlder(zweigliedrige)Gen.abs.alseine Konstruktion (ein “Syntagma”) mit eigenem Subjekt vomGriechischenausgesehen,inhöheremMaßesatzwertig(ÄquivalenteinesNebensatzes)ist”.

It is, however, the function of embedded predications within alargerstretchofdiscoursethatistobediscussed.Thequestionstoberaised are: “what would be the func ti onal difference between theallegedalternatives?”,and“whatfactorsinfluencetheauthor’schoicebetweenthem?”.Therelevanceofthesequestionsresidesinthefactthat our grammars have raised the issue but have not provided asatisfactoryanswer.Toanswerthesequestions,wehavetolookinadifferent direction: it is the hypothesis of this book that discourseanalysis and a text grammar-oriented approach will yield betterresultsindescribingtheusageofsubclausesandparticipialclauses,inthatitcanmapmorefactorsthatunderliethedistribution,andcan

explainagreaternumberofinstancesmoresatisfactorilythancanbedoneonthebasisofsentence-grammar.

1.1 ‘Temporal’SubclausesandVerbalAspect

Grammars and reference books distinguish a particular class of subclauses, viz. , temporal subclauses. They seem to be easilydistinguishedfromothertypesofsubclausesbyvirtueof themeaning oftheirconjunction.Thus,Kühner-Gerth(1898-1904:II445)dividethe grammatical class of temporal subclauses (Adverbialsätze der

———For arecent discussionofthisissue,and ofancientviewsontheconstruction, seeVasilaros(1993:21-37).



4 CHAPTERONE

Zeitbeziehung)intothreecategories,accordingtotheirsubordinator:a)subclausesindicatingsimultaneitybetweenthesubclauseandthemain clause4 (“bei der Angabe der Gleichzeitigkeit mit demHauptsatze”);b)subclausesindicatinganteriorityofthesubclausewithrespecttothemainclause5(“beiderAngabedesdemimHauptsatzeAusgesagten Vorangehenden”); and c) subclauses indicating posteriorityofthesubclausewithrespecttothemainclause 6(“beiderAngabedesdemimHauptsatzeAusgesagtenNachfolgenden”).

Thisclassificationisapplicabletoagreatnumberofcases;insome

cases,however,itisnotwithoutproblems.Itiswell-knownthatinacertain language some subordinators may be semantically morespecific thanothers. Thus inEnglish,the subordinatorwhenmaybeusedtoindicateallkindsofnon-specific(temporal)relationsbetweenthe content of a subclause and its matrix clause while specificsubordinatorsareavailable(e.g.,before,after ,while,as , till .;seeGivón,1990: 828). It is not a priori excluded that among the manysubordinatorstheAncientGreekspeakerhasathisdisposal,somearesemanticallyspecificandothersarenot.Inthisconnectionitistobenotedthat ˜te ,≤n¤ka, and §n ⁄, to name but a few, are alwaysindicators of a temporal relationship between the subclauseand themainclause,whereas§pe¤,§peidÆ,and…w arereportedtoindicatesometimes a temporal, sometimes a causal relation,7 which wouldmakethemsemanticallynon-specificsubordinators.8

———4 Theylistsubclausesheadedby:˜te;eÔte;ıpÒte;…w,Àsper;˜pvw,˜kvw;˜pou;

≤n¤ka,ıphn¤ka;ısãkiw,ıssãki,ıposãkiw;∑mow;ˆfra,˜fra; m°xri, êxri;¶ste;ßvw,

t°vw,§n ⁄.5 §pe¤,§peiÆ,§pe¤te,§peidÆ;§j o, §j ˜tou, éf' o,§j œn;…w/§pe‹ tãxista,§peidØ tãxista.

6 pr¤n,pr¤n ≥,pr¤n…pr¤n,prÒteron…pr¤n,prÒsyen…pr¤n,prÒteron ≥,pr‹n˜te,pãrow;ßvw,ßvw o,ˆfra,˜fra,¶ste, m°xri, êxri, m°xri o, êxri o,§w ˜,§w o,efiw ˜te kenc.conj .

7 Rijksbaron(19942:83n.3)observesthat‘temporal’clauseswith §pe¤,…w,etc.+imperfect or aorist indicative may also have a ‘causal nuance’. In instances like Hellenica, 7.2.19 , §kãyeudon m°xri pÒrrv t∞w≤m°rawtherelationbetweentheproposition[theypassedthenightwithoutsleep]andtheproposition[theysleptfarintotheday]doesnotseemtomefirstandforemosttemporal(?aftertheyhadpassedthenightwithoutsleep,theysleptfarintotheday ),withacausal

nuance.Therelator …windicatesarelationbetweenthetwopropositions,andonthebasisofourknowledgeoftheworld,weinterpretthisrelationas‘causal’( inasmuchastheyhadpassedthenightwithoutsleep,theysleptfarintotheday ).TheGreek,however,doesnotspecifythisrelation.

8 However,considerRuijgh(1971:502-503,§412),whoconsiders§pe¤tohaveamorespecificvaluethan te,inasmuchas§pe¤wouldexpresstheideaofanteriority.



INTRODUCTION 5

Itseemsgenerallyacceptedthatthetense/aspectofthefiniteverboftemporalsubclausescanbeusedtodeterminethesemanticrelationbetweenthesubclauseandthemainclause.Thisviewisnotwithoutproblems.Notefor instancethatin English the interpretationof therelationbetweenawhen-clausewithsimpleaspectanditsmainclausemaybeeithersimultaneousorsuccessive(Smith,1983:486-487).InthecaseofAncientGreeksubclausesthatareinterpretedastemporal,thetense/aspectofthefiniteverbisindeedoftenusedtodeterminethenatureofthetemporalrelation.9Ifindeed§pe¤,§peidÆ,and…w

belongto thecategory ofsemanticallynon-specificrelators,we maywonder whether §pe¤, §peidÆ,and …w-clauses should ever beconsideredstrictlytemporalatall.Further,severalinstancesinextantGreektextsrevealthatthenatureofthe(temporal)relationbetweenthesubclauseanditsmatrixclauseisnotunequivocallyplainfromthetense/aspectofthefiniteverbofthesubclause(orthecombinationof tense/aspectofthefiniteverbofthesubclauseandasubordinating conjunction).Considerthefollowingexamplesofsubclauses(example[1]through[3]):

[1] Hellenica3.4.25 d' aÏth ≤ mãxh § , Tissaf°rnhw §n Sãrdesin ¶tuxen n:

thisbattle ,TissapherneshappenedtobeatSardis.

Thesubclause˜te d' aÏth ≤ mãxh §g°netoisunequivocallytemporal,asisindicatedbytherelator˜te.ThereisnootherrelationintheRealWorldbetweenthesubordinatedpropositionthatthisbattletookplaceandthemainpropositionthatTissapherneswasatSardisthan

thattheeventspresentedarecotemporal;nonetheless,thefiniteverbofthesubclauseisanaorist,provingthatthisaspectualformwhenusedfor thefiniteverbofa subclause isnotnecessarilyindicativeof expressingarelationof‘anteriority’.

———9 Anexample ofthis isfoundin Rijksbaron(19942:74,§26Temporalclauses):

“Thefollowingtypesofclausesmaybedistinguished:(a)§pe¤,§peidÆ,˜te,…w+aorist

indicative:thestateofaffairsofthedependentclauseisanteriortothatofthemainclause:‘when’,‘after’.(b)§pe¤,§peidÆ,˜te,…w,aswellasßvw,§n ⁄+imperfect:thestateofaffairsof thedependentclause issimultaneouswiththatofthemainclause:‘when’;ßvw‘solongas’, §n ⁄‘while’.(c)ßvw,pr¤n+aoristindicative:thestateof affairsofthedependentclauseisposteriortothatofthemainclause: ßvw‘until’,pr¤n‘before’”.



6 CHAPTERONE

[2] Agesilaus2.23

tå m¢n dØ m°xri toÊtou koinª aÈtÒw te ka‹ ≤ pÒliw hÈtÊxei: ˜sa ge mØn metå toËto sfãlmata §g°nonto oÈde‹w ín e‡poi …w ÉAghsilãou≤goum°nou §prãxyh.

d' aÔ t∞w §n LeÊktroiw sumforçw gegenhm°nhwtoÁw §n Teg°& f¤louw ka‹ j°nouw aÈtoË ofl ént¤paloi

sÁn MantineËsi, sunesthkÒtvn ≥dh Boivt«n te pãntvn ka‹ÉArkãdvn ka‹ ÉHle¤vn, s sÁn mÒn˙ <tª Lakedaimon¤vn dunãmei>, poll«n nomizÒntvn oÈd' ín §jelye›n Lakedaimon¤ouwpolloË xrÒnou §k t∞w aÍt«n. d˙≈saw d¢ tØn x≈ran t«nkatakanÒntvn toÁw f¤louw oÏtvw aÔ o‡kade épex≈rhsen.

Uptothistimeheandhiscityenjoyedunbrokensuccess;andthoughmany troubles occurred thereafter, it cannot be said that they wereincurredundertheleadershipofAgesilaus.

On the other hand, after the disaster at Leuctra, w hisadversariesinleaguewiththeMantineansw hisfriendsandacquaintancesinTegea,andacoalitionofallBoeotia,ArcadiaandElishadbeenformed,h withtheLacedaemonianforcesonly,while itwasgenerallyexspectedthattheLacedaemonianswouldnotevengooutsidetheirownbordersforalongtimetocome.Helaid

wastethecountryofthosewhohadmurderedhisfriendsandinthiswayreturnedhomeoncemore.

Thefiniteverbofthesubclauseisapresentindicative,aso-calledhistorical present. The semantic relation between the subordinatedandthemainpropositioncannotbedeterminedonthebasisofthetense/aspectofthefiniteverbofthesubclause;itisthecontextwhichisdecisive.Inordertobeabletoqualifykataka¤nousi(and,forthatmatter,strateÊei )asa ‘historical’present,10wehavetoknowwith

whichtexttypewearedealing(narrative,expository).As[2]istakenfromthe Agesilaus, an encomium encompassing different text typeswithin the same text, we have to go one step further and decidewhether this passage as a whole is diegetic. In this case, it is anarrativepassage,11andweareallowedtointerpretthepraesentiaas

———10 The verbs are presented in the historical present to highlight the relative

importance of the actions as related to the specific aims of the speaker in theencomium:theybearonAgesilaus’squalitiesasahumanbeing(kataka¤nousi t ) and as a general (strateÊei sÁn m

<t > ). The same events are recorded at lenghth in Hellenica6.5.10-21.

11 SeeBuijs,unpublished.



INTRODUCTION 7

historical.12 Itistherefore thecontextrather thantheverbformthatguides our decision; we can only decide on the type of relationinvolvedoncewereachthematrixclause.

Whether we are dealing with a ‘temporal’ subclause remainsundecided.Inthefirstplace,thisisamatteroftranslation:inmyopinion,the§pe¤-clausedoesnotprovidea temporalreferenceinthewaythe˜te-clausein[1]does.Thisisbroughttolightbythefactthatseveralinterpretationsmakesense:‘while/whenhisadversariesintheleague with the Mantineans were murdering his friends and

acquaintancesinTegea(…)hetookthefieldwiththeLacedaemonianforces only’ (temporal simultaneityofthetwo propositions; thus theLoebtranslation),or‘after/when his adversaries in league with theMantineanshadmurdered hisfriendsandacquaintancesinTegea…hetook the field with the Lacedaemonian forces only’ (temporalsubsequence of the two propositions; cf. t«n katakanÒntvn toÁw

f¤louw ),oreven‘becausehisadversariesinleaguewiththeMantineanshadmurdered/weremurdering hisfriendsandacquaintancesinTegea(…)he took the field <against them> with the Lacedaemonian forcesonly’.

Rather than focusing on an alleged specific semantic relationbetweenthetwoclauses,weshouldpayattentiontothecontributionthatthe§pe¤-clausemakestothetext.Atapointofsegmentation( d'

aÔ ),thereisabreakinthecontinuityof participants(newadversariesof Agesilausareintroduced),time, place,andaction.Iclaimthatthe§pe¤-clause reflects the speaker’s organization of events in the depictedworldinthatitintroducesanewstageinthedevelopmentofthe

story-linebypresentingafactualstatementwhichthereader/hearerneedstocomprehendthesequel.That certain subclauses regardless of aspectual choice, even in

narrativesequences,donotprovidetemporalreferencesinthemselvesisfurtherillustratedbydifferentkindsofanaphoricexpressionsfoundatthebeginningofthematrixclause,afterapreposedsubclause:

———12 ContrastdeEquitandiratione1.1,whichisfromanon-narrativecontext(§peidÆ:

‘now that’/‘because’): diå tÚ sumb∞nai ≤m›n polÁn xrÒnon flppeÊein¶mpeiroi flppik∞w gegen∞syai, b ka‹ to›w nevt°roiw t«n f¤lvn dhl«sai √ ín nom¤zomen aÈtoÁw ÙryÒtata ·ppoiw prosf°resyai(‘i

w havehadalongexperienceofcavalry,andconsequentlyc familiaritywiththeartofhorsemanship,w toexplaintoouryoungerfriendswhatwebelievetobethecorrectmethodofdealingwithhorses’).



8 CHAPTERONE

[3] Hellenica7.4.23-24

d' §plhs¤asan éllÆloiw, ofl m¢n sÁn t“ ÉArxidãmƒ katå k°raw, ëte kay' ıdÚn poreuÒmenoi, ofl d' ÉArkãdew èyrÒoi sunaspidoËntew,

oÈk°ti §dÊnanto ofl LakedaimÒnioi ént°xein t“ t«nÉArkãdvn plÆyei, éllå taxÁ m¢n ı ÉArx¤damow §t°trvto tÚn mhrÚn diampãj, taxÁ d¢ ofl maxÒmenoi prÚ aÈtoË ép°yn˙skon, Poluain¤dawte ka‹ X¤lvn ı tØn édelfØn toË ÉArxidãmou ¶xvn, ka‹ ofl pãntew d¢ aÈt«n tÒte ép°yanon oÈk ¶latton t«n triãkonta. … d¢ katå tØnıdÚn énaxvroËntew efiw tØn eÈruxvr¤an §j∞lyon, §LakedaimÒnioi éntiparetãjanto.

Nowthatthetwoforceshadcomenear tooneanother,thetroopsof Archidamusincolumn,sincetheyweremarchingalongaroad,andtheArcadians massed together in close order—a theLacedaemonianswerenolongerabletoholdoutagainstthesuperiorweightoftheArcadians,butArchidamusspeedilyreceivedawoundstraightthroughhisthighandspeedilythosewhofoughtinfrontofhimkeptfalling,amongthemPolyaenidasandChilon,whowasmarriedtothesisterofArchidamus;andthewholenumberofthemwhofellatthat time was not less than thirty. But the Lacedaemonians as theyretiredalongtheroadcameoutintoopenground:t they

formedthemselvesinlineofbattleagainsttheenemy.

Thetemporaladverbial §n toÊtƒoccursattheonsetofthematrixclauseofthefirst…w-clause,butherethefiniteverbofthesubclauseisanaorist.Thefactthatinthissituationoftemporaloverlapbetweentwopropositions,thefiniteverbofthesubclauseisanaorist,furtheril lustrates that semantic relations between clauses in clausecombinations should not be established on the basis of thetense/aspectofthefiniteverbofthesubclausealone.Theadditionof

§n toÊtƒbringstolightthatalthoughthesubordinated actionmaybeanteriortothemainaction,thetwo situationsshouldbeunderstoodasbeingsemanticallyrelatedbytemporalsimultaneity.Iwillarguethatthe…w-clause is used tointroduce the frameof reference for aneweventsequence/thematicunitatapointwherethereisabreakinthecontinuityofthediscourse.Inthiscase,thediscourseisdiscontinuouswithrespectto place(thisisinpartduetothemeaningofthefiniteverbofthesubclause‘todrawnearto’,‘approach’)andsituationingeneral,asexpressedbythepostposedparticipialclausesofl m¢n sÁnt“ ÉArxidãmƒ katå k°raw, ëte kay' ıdÚn poreuÒmenoi, ofl d'ÉArkãdew èyrÒoi sunaspidoËntew. The second …w-clause alsointroduces a new situational frame of reference for the upcoming



INTRODUCTION 9

sequence,whichisspatialratherthantemporal,asappearsfromtheadditionof§ntaËya dÆ.

1.2 Conclusion

Indescribingtheusageofparticipialclausesand §pe¤-,§peidÆ-,and…w-clauses,classicalphilologistsandlinguistshavetraditionallybeenconcernedwiththerelationofembeddedpredicationstotheirmatrixclause.Muchattentionhasbeenpaidtothenatureoftherelation

betweentheembeddedpredicationanditsmatrixclauseintermsof semanticnotionssuchascausalityand,especially,temporality.Withinthisapproach,akey-functionfortheinterpretationofthenatureof this relation has been attributed to the tense/aspect of the verbalconstituentoftheembeddedpredication.Thisapproachturnsouttobeproblematicincertaininstances,13andingeneralitseemstoputtoomuchemphasisontemporality,leavingtheorganizationoftextsand the contributionofsubclausesandparticiples totext coherenceunderestimated. As we shall see, they often articulate a discourseboundary at points of segmentation. This discourse function of preposedsubclausesandparticiplescannotbedescribedsatisfactorilyonthebasisofasentence-levelapproach.Thisapproachprovesevenlessusefulwhenweinvestigatethedistributionofthedifferentclausetypes themselves, often considered alternatives. My aim in thefollowing chapters will be to show that context is crucial forunderstanding the function of the different clause types underconsideration.

2 AnalysisofAncientGreekNarrativeDiscourse

2.1 TheNatureofNarrativeDiscourse

Whenwritingahistoricaltext,thehistoriancapturesinwordsRealWorld experience. In order to do so, he makes a cognitivearrangement operate on his Real World impressions in all their

multifariousforms,whichresultsinaconceptionofevents.Theeventis “a hermeneutic construct for converting an undifferentiated

———13 Theissuewillbefurtheraddressedinthediscussionofexamplesinthisbook.



10 CHAPTERONE

continuumoftherawdataofexperience…intotheverbalstructureswe use to talk about experience: narratives, stories” (Fleischman,1990:99).Theevent thus mediatesbetweenRealWorldexperienceandthewayitispresentedintexts.ThetransformationofRealWorldexperience into historical narrative necessarily involves ‘decision’: adecisionoverwhicheventsaretobeselectedforrecording,andadecisionastohowtopresenttheselectedeventsinanarrative,because“narrativesdonotexist,asitwere,insomerealworld,waitingtherepatientlytobeveridicallymirroredinatext”(Bruner,1991:8).

The perception of reality should be distinguished from theexpressionofrealityinanarrative.Itisoftenassumedthatrealitystructurecomespriortonarrativestructure,inthatthestructureof historical narrative copies the structure of reality. However, theconstraints of narrative structure condition our way of looking atreality. The Real World experience takes form only by itsrepresentation in the form of a narrative. The activities of thehistorianorganizing scattered events into meaningfulensembles aretwofold:heisintheprocessof textconstructionontheonehand,andintheprocessofa(re-)constructionofreality,ontheother.

When imposinghis textual frameworkupon reality, thehistorianmaysignalrelationsbetweensituationsintheRealWorld,ormayrefrainfromsignalingthem.Asnarratedeventsoccurintime,thesignaling of temporal succession is usually seen as one of thecharacteristicsofnarrativediscourse.Yettherelationsbetweentheevents of the narrative are always dependent upon the way thehistorianorganizeshisnarrativeasawhole.Thehistorianthuscreates

“anorderoutofmeresuccession”(Hohler,1987:294).With respect to the object of the present study, the following,simplifiedexamplemaywellillustratethepoint.Ifweconsidertworeality-situations whichare tobe recordedastwoseparateevents,aspeakerofAncientGreekmay(1)optfortwoseparate(co-ordinated)finitemainverbs,or (2) present one of the events in grammaticalsubordination.Inthesecondcasehehasthe choicebetween (2a)afiniteverbinasubclauseandafinitemainverband(2b)aparticipialclauseandafinitemainverb.Option (1)resultsinjustpresentingthetwoevents,nomoreandnoless,without—apartfromtheorderinwhichhepresentsthem—indicatingtherelationbetweentheeventsintheRealWorld.Option(2)maybechosenforreasonsofdiscourseorganization,inwhichcasehecanstillchoosebetweengivingaclue



INTRODUCTION 11

about the relation14 in reality between the subordinated and thesuperordinatedeventbyusingasubclause (2a), or refraining fromdoingso,byusingaparticipialclause(2b).Thus,while(1)differsfrom(2)inthatthechoiceofthelatterpertainstotheorganizationofthetextintermsofthepresentationofthetwoevents,(2a)differsfromboth (1)and (2b)inthatonlyin(2a)theorganizationofrealityintermsoftheRealWorldrelationbetweenthetwoeventsisatissue.

In keeping with the argument outlined above, there are threedifferentlevelsinvolvedinthehistorian’sactivity.ThefirstlevelIshall

callthereality-level ;itisthelevelofhisRealWorldexperience,i.e.anyeventnot(yet)capturedintext,andassuchoutsidethedomainof discourse analysis. The second level I shall label the leveloftheReal World construction, which involves everything pertaining to theorganizationoftheRealWorldsituationsofwhichthetextseekstobea m¤mhsiw.Thisisthe levelatwhichrelationsbetweenevents intheRealWorldas(re-)presentedbytheauthorcomein.Thethirdandfinal level I shall call the level of text articulation, which involveseverythingpertainingtotheorganizationofthetext.Questionsaboutthepresentationofeventsinanon-goingnarrativebelongtothislevel.BoththeleveloftheRealWorldconstructionandtheleveloftextarticulationareofrelevancetodiscourseanalysts.

2.1.1 TemporalRelationsThenotionof‘time’isinvolvedinboththe‘happenings’intheRealWorldand the ‘events’aspresented ina narrative.The timefallacythathashauntedAncientGreekphilologistsandlinguistsworkingin

thefieldofAncientGreek(historiographic)narrativediscourseisthatinnarrativetextswedealwithrealtimerelationships,whereasinfactthe notion of ‘time’ involved is that of narrative time —a kind of projectedtemporalitythatusuallydoesnotcorrespondto‘realtime’.

Usuallytheorderinwhichtheeventsinanon-goingdiscoursearenarrated is iconic in the sense that itmirrors the succession inrealtime of Real World ‘happenings’ as perceived by the narrator.Wheneverthenarratorwishestodeviatefromthisstandardprocedurefor reasons of presentation,15 he has to signal this deviation

———14 Eitheraspecificoranon-specificrelation;cf.Chapter1,Section2.2.1.15 Note that therecordingoftwoeventsas occurringsimultaneouslynecessarily

involvesadecisioninpresentation,sincetwoeventscorrespondingtotwoRealWorld‘happenings’that,intheirturn,areexperiencedasoccurringsimultaneously,cannot,



12 CHAPTERONE

linguistically.Moreover,hemaywishtospecifytemporalrelationswhenever he thinks that doing so will contribute to a successfulcommunication with his audience. In order to specify temporalrelations between narrated events, the Ancient Greek speaker hasseveraldevicesathisdisposal.Apartfromusingasubclauseheadedbya specific time subordinator, he may put in temporally specificadverbials,e.g. §n toÊtƒ, metå toËto/taËta,¶peita,orÏsteron.16

AlthoughitoftenseemstobeassumedthattheAncientGreekspeakerwas much concerned with expressing temporal relationships,17 the

actualsituationinAncientGreekdoesnotsupportthisview.Althoughtemporal relationshipsbetweennarratedeventsmay bespecifiedbymeansoftemporallyspecificsubordinators18oradverbials,suchovertmarkingoftenlacks.Ourinterpretationoftemporalrelationshipsisusually determined merely by the order in which the events arepresented,inconjunctionwithourknowledgeoftheworld.

Thisyieldsthe following distinction: whenoftwoactionsA 1andA 2,actionA 2isinterpreted incontext assequentialtoaction A 1onthenarrativetimeplane,narrativetimeispropelledforward;itwillalwaysbedifficulttoassertanythingabout‘realtime’relationships,sinceweare only dealingwith the speaker’s (re-)presentationofreality intheformofnarrative.

———strictlyspeaking, benarratedsimultaneously.On a more global discourse level, anarratormightthereforewishtoorganizehisnarrativethematically:cf.forinstance Hellenica7.3.4-7.4.1 éllå går §pe¤per ±rjãmhn, diatel°sai boÊlomai tå per‹ EÎfronow …ka‹ tå m¢n per‹ EÎfronow e‡rhtai: §gΔ d¢ ¶nyen efiw taËta §j°bhn

§pãneimi(‘however,sinceIhavebegunit,IdesiretofinishthestoryofEuphron…thestoryofEuphronhasbeentold,andIreturntothepointfromwhichIdigressedtothissubject’)—cf.also6.1.19 ka‹ taËta m¢n oÏtvw §pera¤neto: §gΔ d¢ pãlin§pãneimi, ˜yen efiw tåw per‹ ÉIãsonow prãjeiw §j°bhn (‘thus these events wereproceedingtotheirissue;InowreturntothepointatwhichIdigressedwhenItookupthestoryofJason’);in4.8.1ka‹ ı m¢n dØ katå g∞n pÒlemow oÏtvw §poleme›to. §n⁄ d¢ pãnta taËta §prãtteto, tå katå yãlattan aÔ ka‹ tåw prÚw yalãtt˙ pÒleiwgenÒmena dihgÆsomai…(‘asforthewarbyland,itwasbeingwagedinthemannerdescribed. Iwillnowrecountwhat happenedby sea and inthecitieson the coastwhileallthesethingsweregoingon…’),thespecific(temporal)relator §n ⁄marksthesimultaneousoccurrenceofthewarbylandandtheeventsbyseaandinthecitiesonthecoast,whichXenophonchosetonarrateseparately.

16 Perhapsweshouldincludetheexpression§k toÊtou,whichatleastindicatesasubsequentdevelopmentofevents,butlacksthefeatureofbeingsemanticallyspecific;cf.Pelliccia(1989:95-96)onexpressionslike§k toË,§j oand éf' o.

17 Cf.especially thediscussiononthevalueof theaoristandpresentstem,e.g.,Hettrich(1976);foracontrastingopiniononthesubject,seeSicking(1996).

18 SeeChapter1,Section1.1.



INTRODUCTION 13

2.2 FactorsOperatingontheChoiceamongAlternativeExpressions

Now that the levels relevant to the study of narratives have beenestablished,theymayprovideapointofdepartureforamoredetaileddescriptionofthefactorsoperativeontheAncientGreekspeaker’schoice among different clause types in creating (historiographic)narrativediscourse.TheseIconsidertobe:

1. PresentationofRealWorldRelations2. TextArticulation3. InformationProcessing

2.2.1 PresentationofRealWorldRelationsCeterisparibustheorderinwhichtheeventsarenarratedinahistoricaltextmirrorsthespeaker’sperceptionofthesuccessioninrealtimeof the Real World ‘happenings’ they represent. If in his narrative(re-)construction of reality the historian makes use of clausecombining,hemayeitherindicatethatheexperiencessome,initself

unspecified,RealWorldrelationbetweenthesubordinatedandthesuperordinatedevent,orhemayrefrainfromindicatingaRealWorldrelationbetweenthem.Ifhewishestoindicatethat,accordingtohim,aRealWorldrelationbetweenthetwoeventsobtains,hemayuseasubclause.Hemayalsowishtospecifythisrelation.Ifhedoesso,hewillmakeuseofrelatorsthatreferto(aspects)ofstatesofaffairsinreality,whichIcallsemanticallyspecificrelators,suchas §n ⁄,˜te,or≤n¤ka,fortemporalrelations,or d¤otior˜tifor‘causal’19relations.Incase he does not want, or need, to be specific, but still wants toindicatethat,accordingtohim,aRealWorldrelationbetweenthesubordinatedandthesuperordinatedeventobtains,withoutspecifying thenatureofthisrelation,hewillusenon-specificrelators suchas§pe¤,20

———19 Thedisadvantageofthetraditionaltermcausal isthatthistermmaybetakento

exclude relations of reason. Whenever I speak of ‘causal’ relations (in invertedcommas),bothrelationsof causeandrelationsof reasonaremeant;Ishalldistinguishbetweenthetwotermswhennecessary.

20 AviewcomparabletominehasbeenputforwardbyPellicciainanarticleon

Pindar, Nemean7.31-36(1989:74):“Themoresensiblecourse,inmyview,isnottofasten specific meanings on §pe¤andinsteadtodescribeitsfunctionalone:§pe¤introducescircumstantialclausesmodifyingthemainverb;thetime-valueofitsverbisrelative to that of the main verb; the meaning of the clause (temporal, causal,concessive,etc.)inanygivencontextisdeterminedbythedemandsofthatcontext.A good model for comprehending the range (and vagueness) of an §pe¤clauseis



14 CHAPTERONE

§peidÆ,or

…w, which operate not on a semantic level, but on the

discourse level. When, however, the narrator does not want toindicateaRealWorldrelationbetweentheeventsexpressedinthesubordinate clause and the matrix clause, he uses a participialclause.21 The most obvious example of this is provided by thoseinstanceswhere,apartfromthetemporalrelationintheRealWorldthat always obtains between two events, the clause combinationexpressesnoconceivableRealWorldrelationbetweentheevents.22Inthis case the choice of a subclause headed by a semantically non-

specific relator is excluded altogether. But the speaker may simplywishnottoindicateaRealWorldrelationwhichhedidexperience,becauseatacertainpointinthetexthischoiceinfavorofaparticipialclause is influenced by other factors. That sometimes other factorsprevailinthechoiceamongclausetypesisshownbythefactthatweactuallydofind participialclausesheadedby a semanticallyspecificrelatorsuchaska¤perand ëte,or…w.23Insuchcasesthenarratorstillwantsto specify theRealWorldrelationbetweenthecontentoftheclausesintheclausecombination.

PresentationofRealWorldrelationsbetweeneventsissomething thatpertainstotheleveloftheRealWorldconstructionofdiscourseanalysis, as it is concerned with the way the perceived reality isexpressedinthetext.

———furnishedbythecircumstantialparticiple,anditmayonoccasionevenbedesirableto

translatean§pe¤clauseasanunadornedparticipialphrase;thatistosay,todeclinetoclassifyagiveninstanceastemporal,causal,orconcessive,intheinterestsofavoiding theimpositionofsyntacticalpatternsfoundinthespeaker’sownlanguagebutnotintheoriginalGreek”.Althoughtoostrictlyconcernedwithsentence-levelgrammar,Pelliccia’sviewisanimportantstepforward.

21 It should be noted that as far as Real World relations are concerned, it isirrelevantwhether the verbal constituentof the participialclause isan aorist orapresenttensestemform.

22 Cf. Hellenica3.2.21,wheretheonlyconceivableRealWorldrelationbetweentheactionsperformedbyDercylidasinAsiaandtheSpartansinGreeceisindeedatemporalone,butnotethat thisrelation isspecifiedonlybythetemporaladverbialkatå tÚn aÈtÚn xrÒnon.

23 In this connection one may alsothink of the usage of adverbs like metajÊ,eÈyÊw,and ëma.



INTRODUCTION 15

Figure1:PresentationofRealWorldRelations24

Non-linguistic EventA

World EventB

Relating NotRelating

ClauseCombining EventA& EventB EventA& EventB

inText

SpecificRelator Non-SpecificRelator

Temporal Causal Other

§n ⁄ ˜ti ka¤per(adversative) §pe¤ Participial

˜te diÒti efi(relevance) §peidÆ Clause

≤n¤ka …w+ptc. etc. …w without

ëte+ptc. relator

oÂa+ptc.

˜pouetc.

2.2.2 TextArticulation Preposed subordinate clauses may perform the function of textarticulation.Thistakesplaceattheboundaryofdiscourseunits,i.e.,atpointsofdiscontinuityinthediscourse.Boundariesarefoundbothon the levelof text articulation andonthe levelof the Real Worldconstruction.

On the level of text articulation, the boundaries are best calledtextual boundaries; textual boundaries occur at points where the

———24 Only the semantically specific relators most commonly found in narrative

discoursearelistedhere;foracompleteoverviewof(allegedtemporal)relators,mostofwhichforthatmatterdonotqualifyasalternativesfor§pe¤,§peidÆ,and…w,seen.4through6.



16 CHAPTERONE

speakerwantstoindicateatransitionofonepart ofthetext toanother.Atsuchboundaries,subordinateclausesmayperformthefunctionof segmentingtext(withorwithoutlinkageoftextparts)byintroducinganewpointofdeparturefortheupcomingtextsegment.

On the other hand, we also find boundaries that pertain to thecontent ofthetext;atthesecontent-orientedor thematicboundariesthespeaker wants to indicate a transition from one part of the story toanother.ThesethematicboundariesbelongtotheleveloftheRealWorldconstruction.Atthislevel,subordinateclausesmayperform

the function of segmenting text into different thematic units.Itisimportanttonotethatifwehaveathematicboundary,wealsohaveatextual boundary, but not necessarily viceversa: a textual boundarydoesnotnecessarilyimplyathematicboundary—cf.theanalysisof discourse boundaries and of the particles d°and ka¤ as marking discontinuityandcontinuity,respectively,inBakker,1993.

TextArticulation,then,appearstobeafactorinthoseinstanceswherepreposedsubordinateclausesperformafunctioninsegmenting text into discourse units. Generally speaking—as will be furtherdemonstratedinChapter4—,ifachoiceamongsubordinateclausesis influenced by the factor of Text Articulation, narrators tend tochoose a participial clause when they wish to signal a textualboundarybyintroducinganewpointofdeparturefortheupcoming textsegmentwithoutsegmentationofthetextintodifferentthematicunits.Ontheotherhand,ifthenarratorwantstosignalathematicboundary in addition to a textual one, the choice is in favor of asubclause(an§pe¤-,§peidÆ-,or…w-clause). In Section 2.2.1 I have

postulatedthat§pe¤-,§peidÆ-,and…w-clauseareusedforindicatingasemanticallynon-specificrelationbetweenclauseswithinthesentence.Within the opposition participial clause vs. subclause, then, thesubclauseismarkedforcontent-relations,bothonthesentence-leveland,ifpreposed,onthediscourse-level.ThegistofSection2.2.1and2.2.2ispresentedinfigure2:



INTRODUCTION 17

Figure2:PresentationofRealWorldRelationsandTextArticulation

participialclause • norelationindicated • textualboundary

subclause • semanticallynon-

specificrelation’content-relationbetweenclauseswithinthesentence

• textual+thematic

boundary’content-segmentationof discourseunits

2.2.3 InformationProcessing The presentation of information in discourse may be approached

fromvariousangles.In thisSectionIshalldiscusstwooppositionsbywhichinformation-phenomenaareoftenviewed:GivenInformationvs. New Information and Main Information vs. SubsidiaryInformation.

Basically, a taxonomy of given/new information is tripartite.25

Informationconveyedatanypointindiscoursecanbe entirelynew,contextuallyprepared orentirelygiven.Informationstatusisbestviewedasacontinuumwithglidingscales,withontheoneextremeentirelynewinformation (i.e., information not yet processed in the preceding discourseornotpresentinthespeechsituation),andontheotherextreme entirely given information (i.e., information alreadyprocessed in the preceding discourse or present in the speechsituation).Between these two extremes,thereare variousdegrees inthe‘newness’or‘givenness’ofinformation.Fromthepointofviewof textproduction,thestatusofthisinformationdependsonwhatthetext-producerassumesthetext-receivertobeabletoinferfromotherentitiesinthediscoursecontextorspeechsituation.Viceversafromthe

pointofviewoftextcomprehension,thestatusofthisinformationdependsonwhatthetext-receivermayinferfromotherentitiesin

———25 Althoughfurtherdistinctionscanbemade;seePrince,1981;Virtanen,1992.



18 CHAPTERONE

discoursecontextorspeechsituation.Sinceindescriptivelinguisticsitisthelinguist’sworktoconcentrateonthetextitself,I shalllabelthis‘intermediate’categorycontextuallypreparedinformation.

Informationmaybecontextuallypreparedindifferentways.Inthefollowing example,the two subordinateclauses inbold type containcontextuallypreparedinformation:

OscarWilde,ThePictureofDorianGray ,BeginningofChapter8 26

Itwaslongpastnoonwhenheawoke.Hisvalethadcreptseveraltimesontiptoeintotheroomtoseeifhewasstirring,andhadwondered

whatmadehisyoungmastersleepsolate.Finallyhisbellsounded,andVictorcameinsoftlywithacupoftea,andapileofletters,onasmalltray ofold Sèvreschina,and drewbackthe olive-satincurtains,withtheir shimmering blue lining, that hung in front of the three tallwindows. “Monsieur has well slept this morning,” he said, smiling.“Whato’clockisit,Victor?”askedDorianGraydrowsily.“Onehourandaquarter,Monsieur.”

Howlateitwas!Hesatup,andh ,turnedoverhisletters[…].

Afterabout ten minuteshegotup,and,throwingonanelaborate

dressing-gown of silk-embroidered cashmere wool, passed into theonyx-paved bathroom. The cool water refreshed him after his long sleep.Heseemedtohaveforgottenallthathehadgonethrough.Adimsenseofhavingtakenpartinsomestrangetragedycametohimonceortwice,buttherewastheunrealityofadreamaboutit.

,hewentintothelibraryandsatdowntoalightFrenchbreakfast,thathadbeenlaidoutforhimonasmallroundtable close to the openwindow. Itwas an exquisite day. Thewarmairseemedladenwithspices.Abeeflewinandbuzzedroundtheblue-dragonbowlthat,filledwithsulphur-yellowroses,stoodbefore

him.Hefeltperfectlyhappy.Thesecondparagraphstartswithasentenceofreportedthoughtof DorianGray;thenweareinformedthatheperformedthreeactions:hesatup,sippedsometea,andturnedoverhisletters.Thefirstandthethirdactionaredescribedbyafinitemainverb,thesecondoneisexpressedby aparticipialclause. ThatDorian actually sipped someteaissomethingthatwehadnotbeentoldbefore;theinformationconveyedin theparticipialclause,however, ishighlyinferrable,both

fromthesituationingeneral(drinkingteaissomethingpeopledoafterwaking up) and from the preceding discourse (the first paragraph

———26 Takenfrom:CompleteWorksofOscarWilde.WithanIntroductionbyVivianHolland ,

LondonandGlasgow:Collins,19773.



INTRODUCTION 19

describesVictorcominginwithacupoftea,sothatthereaderwouldexpect Dorian todrinksome).The sameapplies,mutatis mutandis,tothe information contained in the subclause “As soon as he wasdressed”: this is contextually prepared information because peoplemaybeexpectedtodressafterwashing,aneventaboutwhichthereader was informed in the preceding paragraph. As a result, thesubclauseeffectsasmoothchange-overfromtheparagraphaboutthebathroom to the paragraph about the breakfast in the library, byexpressing contextually prepared information that naturally fits in

between.Whiletheinformationstatusofentirelynewinformationishigh,

contextuallypreparedinformationhasarelativelylowerinformationstatus. In text comprehension, entities conveying contextuallypreparedinformationare moreeasilyprocessed,and for thisreasonthey can performa text-organizing function.Participialclauses andsubclausesconveyingcontextuallypreparedinformationare typicallyusedforarticulatingthetext:theydescribeactionsthatarelesssalientfor the developmentof the story line than thosedescribed bymainverbs(asinthecaseoftheparticipialclause“havingsippedsometea”in the above example), or articulate thematic units (as with thesubclause“Assoonashewasdressed”above).27Inthecaseofentirelygiveninformation,therecipientofthetextisproperlyspeakingnotinformedatall,sothattheseentitiesmaybesaidtolackinformationstatus;resumptiveclausescontainingsuchinformationareusedfortext-organizationalpurposesonly.So,asaruleofthumb,wemaystatethatthelowertheinformationstatusofanentity,thegreateris

itstext-articulatingfunction:

———27 Note that “having sipped some tea” occurs sentence- and paragraph-medial

whereas“Assoonashewasdressed”occupiesthesentence-andparagraph-initialposition.



20 CHAPTERONE

Figure3:ContinuumofInformationStatus

HIGH

INFORMATION

STATUS

INFORMATION

0

Entirely Contextually EntirelyGiven Prepared New(Restatement/Summary)

ThesecondoppositionpositedatthebeginningofthisSectionistheopposition of Main Information vs. Subsidiary Information. Thisoppositionismainlybasedontheassumptionthatnotallinformationprovidedindiscourseisequallyimportantforthepointthespeakerwants to make. On the sentence level, the notion of subsidiaryinformationisusedtorefertothosepiecesofinformationthatdonotbelongtothemainassertion.Onthediscourselevel,clausesthatdonot belong to the narrative assertion are referred to as containing

subsidiary information; they owe their appearance to the fact thatthey make the text more than an enumeration of events, i.e. theymake a narrative out of successive events. Thus, speakers oftenpresentinformationinordertoorienttheiraddresseetowardsanewpoint to be made or introduce information for reasons of textarticulation(e.g.,‘closure-marking’).Ofteninformationisconveyedinrelationtootherinformationforclarifyingpurposes,e.g.toelucidateor motivate the occurrence of an action or for the speaker’s ownconduct. The opposition discussed here basically amounts to the

followingdichotomy:informationmaybeconveyedasindependentlyinformativeortobeconnectedbytherecipientofthetexttootherinformation in the context. In the second case the information is



INTRODUCTION 21

additional to the main point made, so as to color the audience’smentalpictureofthecurrentsituation.

3 Aims

Themainquestionthatwillbeaddressedinthisbookwillbe:‘Whatarethecontextual factorsunderlyingtheGreekauthor’schoicebetweenaparticipialclauseandan§pe¤/§peidÆ/…w-clause?’Inansweringthis

centralquestion,thisbookneithertriestoconfirmexistingtheoriesnordoesitpresentawatertighttheoryofitsownintermsofstringentrules. What it doesaim at, however, is signaling regularities in theusage of the clause types studied. Signaling regularities, for thatmatter,issomethingwehavetocontentourselveswithindiscourseanalysis—seeBrown& Yule(1983:22):“Thediscourseanalyst,withhis‘ordinarylanguage’data,iscommittedtoquiteadifferentviewof the rule-governed aspects of a language. Indeed, he may wish todiscuss,not‘rules’butregularities,simplybecausehisdataconstantlyexemplifies non-categorial phenomena”. We do not have ‘ordinarylanguage’ dataatourdisposalforAncientGreek,butassumingthatthewriterofanAncientGreektextwasfreetochooseamongthemanywaystocodeamessagelinguistically,andthathischoiceamong allegedalternativeclausetypesisrelatedtothefunctionoftheclauseintheorganizationofcoherenttext,wecanexaminethetransmittedtextasaccuratelyaspossible;itisonlybycloseexaminationincontextthatwewillarriveatabetterunderstandingoftheusageofthese

clausetypes.Inordertoreducetheinfluenceoftherhetoricalaspectsoftextsasmuchaspossible,mydata-baseconsistsoftwocomparable,but not identical narrative textsby Xenophon:the Hellenicaandthe Anabasis.AdditionalexamplesareoccasionallyprovidedfromothertextsbyXenophon,especiallytheencomium Agesilaus,whichcontainsnarrative passages that describe Real World events that are alsorecorded in the Hell en ica,andtheCyropaedia .InChapter2,Iwillpresent a selection of ‘near-minimal pairs’: instances where thealternativeexpressionsareusedincomparablecontexts.InChapter3,IwilldiscussthewayinwhichRealWorldrelationsarecodedbymeansofsubordinatingconjunctions.Semanticallyspecificrelationsbetweenembeddedpredicationsandtheirmatrixclausewillbesetoff fromsemanticallynon-specificones.Thedistributionof §pe¤-,…w-,



22 CHAPTERONE

and§peidÆ

-clauses28 and participles will then be addressed inChapters4and5.Asthediscoursefunctionofembeddedpredicationsturns out to vary according to their position in the sentence, Idistinguish between preposed clauses (embedded predications thatprecede their matrix clause)29 and postposed clauses (embeddedpredications clauses that follow their matrix clause). The differentfunctions of preposed subclauses and participles, and especiallysentence-initialones,asdiscourseboundarymarkersisdescribedinChapter4.Chapter5isdevotedtothedifferencesbetweenclause

types with respect to the information flow in on-going discourse.Section 1 discusses postposed embedded predications which mayreceivea‘causal’interpretation,whileinSection2viewsonthesyntaxof text grammar outlined in the foregoing Chapters are used toaccountfortheoccurrenceofgenitiveabsoluteconstructionswhere,from the point of view of sentence-syntax, the narratorcould haveusedaconjunctparticiple.Finally,theresultsoftheseinvestigationswillbepresentedinChapter6.

As we should be careful about relying on rules derived fromstatistics,Ihavechosennottoprovidethese.Numbersandratios,especially when different factors pertaining to the distribution of participles and subclauses are operative at the same time, merelyproduceanappearanceofobjectivitywhichwouldbeoutofplaceinastudythataimsatpointingouttheimportanceofthecontextinwhichthealternativeexpressionsoccur.

The reader will observe in the discussion of examples that mytreatment of tense stems is not based on any particular theory of

verbal aspect in Ancient Greek. However, I tend to regard verbalaspect, especially where aorist tense stem forms and present tensestem forms in indicatives and participles are at issue, as a textarticulating device, and therewith I shall be found to subscribe

———28 Ihavenotbeenabletofinddecisivecriteriatosufficientlydescribetheinternal

distributionof §pe¤, §peidÆ,and…w . ComparingCyropaedia5.3.57 d¢ ≤m°ra§g°neto,7.2.2 d¢ ≤m°ra §g°neto,and4.1.9 d' ≤m°ra §g°netoonewouldbeinclinedtorefrainfromdistinguishingaspecificcontributionofeachofthethree

subordinators.Furtherresearchisrequiredhere.29 Embeddedpredications occupinga positionearly inthe sentenceperformamore prominent text-organizational function; when necessary, I shall distinguishwithinthecategoryofpreposedclausesbetween sentence-initially placed clauses andclauses that precede their matrix clause, but do not occupy the sentence-initialposition.



INTRODUCTION 23

predominantlytotheviewsonverbalaspectthatSickingpresentedinaseriesofpublications(seeespeciallySicking1991and1996).Thus,averbalactionwillbeseenasindependentlyassertedwhenshowinganaoristtensestemform,whileimperfectsandpresenttensestemformparticiples are considered to be connected to other actions in thecontext(compare,however,forthediscoursefunctionoftheimperfectalsoRijksbaron,1986,1988,and2002).

EnglishtranslationsareinprincipletakenfromtheLoebeditions.In a number of cases, however, I have changed, adapted or

paraphrased these translations in order to stay closer to theinformation-structureoftheGreektext.



CHAPTERTWO

SIXNEAR-MINIMALPAIRS

Introduction

In the case of modern languages, linguists have the possibility of takingtwotextsthatdifferonlyslightly,andofaskinganativespeakerhowheorsheexperiencesthetexts.Thisisnotpossibleinthecaseof adeadlanguagesuchasAncientGreek.Fortunately,thereareseveralparallel passages in the works of Xenophon that provide theopportunitytotakeacloselookatalternativeswhich,atfirstglance,mightgivetheimpressionofbeingdemonstrationsofrandomchoiceonthepartofthespeaker,butatthesametimepresenttousexactlytherequiredmaterial.InthischapterIpresentaselectionofsuchnear-minimalpairs(NMP’s),basedontheirabilitytoclarifytheusageofthe AncientGreek‘alternative’ clause typesthatwereoutlined inChapter 1, in order to investigate their contribution to textarticulation.NeitherinthediscussionofAncientGreekgrammar,norin discourse-centered linguistics have these passages received fullconsideration.

TheclaimsmadeinChapter1willbesubstantiatedandillustratedonthebasisoftheseNM P’sthatillustratethechoicebetween:1)a

conjunct participial clause and a finite main verb, 2) a conjunctparticipialclauseandasubclause,3)asubclauseandamainclause,4)asubclauseandagenitiveabsoluteconstruction,and5)aconjunctparticipialclauseandagenitiveabsoluteconstruction.IntheanalysisoftheNMP’sitwillbespecifiedwhattheeffectoftheindividualclausetypesconsistsin.

1 NMP1:Hellenica4.3.20-21vs.Agesilaus2.15

Inmanyinstancesthe Hellenicaandthe AgesilausdescribethesameReal World situation; sometimes the exact same verbal action isexpressedbydifferentlinguisticmeans.Theexamplebelowillustrates



SIXNEAR-MINIMALPAIRS 25

thisforthetheaction‘takingdinner’,whichisexpressedbyaconjunctparticipial clause in the Hell enica,andbyafinitemainverbinthe Agesilaus.

[1] Hellenica4.3.20-21 [2] Agesilaus2.15tÒte m¢n oÔn

— ka‹ går ∑n ≥dh Ùc° —

. pr— d°…Then—itwasalreadylate—

.Inthemorning…

tÒte m¢n oÔn — ka‹ går ∑n ≥dh Ùc° — sunelkÊsantew toÁw t«n polem¤vnnekroÁw e‡sv fãlaggow

: pr— d°…Then—itwasalreadylate— havingdraggedtheenemy’sdeadwithintheirbattleline,

Inthemorning…

Attheendoftheday,Agesilausandhismentakedinnerandgotosleep ( deipnopoihsãmenoi §koimÆyhsanand§deipnopoiÆsanto ka‹

§koimÆyhsan,respectively);in[2],Xenophonaddstheinformation

thattheydraggedtheenemy’sdeadwithintheirbattleline.Bythisfact,thetwopassagesdifferinthatthe Hellenicapresentsasequenceof two verbal actions, while the Agesilaus presents us with three. Ineithercase,afinitemainverbisusedfortheaction‘goingtosleep’.

Thefinitemainverb§koimÆyhsan oftenoccursintheworksof Xenophontocloseoffadiscourseepisode,unsurprisinglysogiventhemeaning of the verb. Example [1] and [2] thus present ‘typical’scenes;wemightevengoonestepfurtherandpositthatXenophon,

inproducing(historical)narrative,hada‘pre-fab’-schemeinmind,tobeusedforarticulationatsuitablepointsinhistext.

Inordertoprovideananswertothequestionwhatwouldbethedifferencebetweenthechoiceofan(aorist)conjunctparticipialclauseandofafiniteverbinNMP1,Ihavemappedthedistributionofclausetypesbywhichothersequentialactionsarepresentedwhenevertheverbform§koimÆyhsanisusedinthesamecontextasinNMP1.

Figure1:‘goingtosleep’-sequencesintheworksofXenophon.[FV=finiteverb;CP=conjunctparticiple:a=aoriststem;p=presentstem;pf=perfectstem]



26 CHAPTERTWO

FV:3Cyropaedia5.4.19

ka‹ tÒte m¢n oÏtvw § . ëma d¢ tª≤m°r&…Andthusthent .Atdaybreak…

Cyropaedia7.3.1

ka‹ tÒte m¢n oÏtvw §koimÆyhsan. tª d' Ístera¤&…Andthusthent .Onthefollowing day…

Cyropaedia7.5.40

ka‹ tÒte m¢n oÏtvw §koimÆyhsan. tª d' Ístera¤&…Andthusthent .Onthefollowing

day…pfCP+FV:1

Anabasis4.3.2

tÒte m¢n oÔn hÈl¤syhsan mãla ≤d°vw ka‹tépitÆdeia ¶xontew ka‹ pollå t«n parelhluyÒtvnpÒnvn mnhmoneÊontew. •ptå går ≤m°raw ˜sasper§poreÊyhsan diå t«n KardoÊxvn pãsaw maxÒmenoi

diet°lesan, ka‹ ¶payon kakå ˜sa oÈd¢ tåsÊmpanta ÍpÚ basil°vw ka‹ Tissaf°rnouw. …w

oÔn éphllagm°noi toÊtvn ≤d°vw §koimÆyhsan. ëma d¢ tª ≤m°r&…Atthetime,then,theywentintotheirquartersveryhappily,havingprovisionsandlikewisemanyrecollectionsofthehardshipsthatwerenowpast.ForduringallthesevendaysoftheirmarchthroughthelandoftheCarduchianstheywerecontinuallyfighting,andtheysufferedmoreevilsthanallwhichtheyhadsufferedtakentogetheratthehandsofthe

KingandTissaphernes.I ,therefore,thesetroubles

happily.Atdaybreak…aCP+FV:5

Hellenica4.3.20-21

tÒte m¢n oÔn—ka‹ går ∑n ≥dhÙc°—deipnopoihsãmenoi §koimÆyhsan. pr— d°…Then—itwasalreadylate—h

.Inthemorning…




Anabasis4.5.29 taÊthn m¢n tØn nÊkta diaskhnÆsantew oÏtvw§koimÆyhsan §n pçsin éfyÒnoiw pãntew ofl

strati«tai, §n fulakª ¶xontew tÚn k≈marxon ka‹tå t°kna aÈtoË ımoË §n Ùfyalmo›w. tª d' §pioÊs˙≤m°r&…Thatnight,then,h thusg allthesoldiersw amidanabundanceof everything,keepingthevillagechiefunderguardandhischildrenalltogetherwithinsight.Onthenextday

… Anabasis6.4.10

ka‹ tÒte m¢n deipnÆsantew §koimÆyhsan. tª d¢Ístera¤&…Andthusthenh .Onthefollowingday…

Cyropaedia3.1.43

tÒte m¢n dØ j ofl strati«tai. tª d' Ístera¤&…

Andthenh thesoldiers.Onthefollowingday…

Cyropaedia3.3.28

ka‹ §ke¤nhn m¢n tØn nÊkta Àsper ¶prepeprofulakåw poihsãmenoi •kãteroi §koimÆyhsan.tª d' Ístera¤&…Andthatnight advanceguards,aswasproper,eachsidew .Onthefollowing day…

aCPka‹aCP+FV:2Cyropaedia

4.1.7

ka‹ ofl m¢n dØ émf‹ KËron deipnopoihsãmenoi ka‹

fulakåw katasthsãmenoi …w ¶dei §koimÆyhsan. ofl d¢ ÉAssÊrioi…AndCyrusandhisfollowers,h

sentinelsduly,w .TheAssyrians…

Cyropaedia7.2.1

ka‹ ofl m¢n émf‹ tÚn KËron deipnopoihsãmenoi ka‹ fulakåw katasthsãmenoi, Àsper ¶dei,§koimÆyhsan. Kro›sow m°ntoi…AndCyrusandhisfollowers,h

sentinels,aswasnecessary,w.Croesus,however,…




Other Anabasis2.1.1

[…w m¢n oÔn ≤yro¤syh KÊrƒ tÚ ÑEllhnikÚn ˜te §p‹tÚn édelfÚn ÉArtaj°rjhn §strateÊeto, ka‹ ˜sa §ntª énÒdƒ §prãxyh ka‹ …w ≤ mãxh §g°neto ka‹ …wKËrow §teleÊthse ka‹ …w §p‹ tÚ stratÒpedon

ofl ÜEllhnew § ofiÒmenoi tåpãnta nikçn ka‹ KËron z∞n, §n t“ prÒsyen lÒgƒ

dedÆlvtai.] ëma d¢ tª ≤m°r& …[TheprecedingnarrativehasdescribedhowaGreek

forcewascollectedforCyrusatthetimewhenhewasplanninganexpeditionagainsthisbrotherArtaxerxes,whateventstookplaceduringtheupwardmarch,howthebattlewasfought,howCyrusmethisdeath,andhowtheGreeksh

totheircampw ,supposingthattheywerevictoriousatallpointsandthatCyruswasalive.]Atdaybreak…

Cyropaedia2.4.30

Xrusãntaw m¢n dØ taËta ékoÊsaw ka‹ §pigaurvye‹wtª §ntolª tª KÊrou, labΔn toÁw ≤gemÒnaw,

épelyΔn ka‹ paragge¤law ì ¶dei to›w ëma aÈt“ m°llousi poreÊesyai, énepaÊeto. § d¢

˜son §dÒkei m°trion e‰nai,§poreÊeto §p‹ tå ˆrh. KËrow d°, §peidØ ≤m°ra§g°neto,…Onhearingthis,ChrysantasbeingelatedwithhiscommissionfromCyrustookhisguidesandwent

away,andhavinggivenwhatordershethoughtnecessarytothosewhoweretogowithhimwenttorest.W aslongashethoughtreasonable,hestartedforthemountains.Cyrus,whendaylightcame,…

Ishalldiscussexamples[a]through[p]aslistedandclassifiedinthefigure.1

———1 I will not discuss examples [q], where §koimÆyhsanisusedinadependent…w-clause,andexample[r],wherethecompoundmainverb épekoimÆyhsanisusedinan§pe¤-clause.IhaveincludedtheNMPdiscussedatthebeginningofthisSectioninthecorpus([1]=[e];[2]=[n])inordertobeabletodiscusstheindividualmembersofthepairinconnectionwiththerelevantothermembersofthecorpus.



30 CHAPTERTWO

Ineveryinstancethereisaboundaryinthetextafterthesentenceofwhich §koimÆyhsan is the main finite verb. This boundary,separating(and,atthesametime,linking)the twoadjacentdiscourseunits, is thematic in nature: a Real World boundary (viz., thetransition from the one day to the other) is reflected by a textual boundary in that one event sequence ends, and a new one starts.Further,thetransitionfromoneeventsequencetoanotherissignaledbyahighdensityoflinguisticmarking.Conspicuousistheusageof transition-marking m°n… d°.Inallinstancesthe m°n-membercloses

off adiscourse unit, while the next event sequence starts in the d°-member(cf.Bakker,1993:302-305).Onlyin[k]dowehave m°ntoiopeningthenewsentence,butthisiseasilyaccountedfor(seebelow).

Indicatorsofthematicbreaksbetweenadjacentdiscourseunitsaretemporal,spatial,referentialoractiondiscontinuity(seeChapter4,Section 1). In all instances, there is temporal discontinuity in thetransitionfromtheonedaytotheother.Thisismarkedbytheusageof temporal adverbials: in [a]-[i] and [l]-[p] (total: 14) we have atemporaladverbialinthe m°n-memberineveryinstance(12x tÒte;twiceareferenceto‘thatnight’: taÊthn/§ke¤nhn…tØn nÊkta ).In13ofthese14instances,thenextsentenceopenswithareferencetothefollowingday,invariousforms:6xtª d' Ístera¤&;2xtª d' Ístera¤&pr–;2xpr— d°;2x ëma d¢ tª ≤m°r&;1xtª d' §pioÊs˙ ≤m°r&.Thefourteenthtimewehave ≤n¤ka d' ∑n §n m°sƒ nukt«n.Thisis[m],whichisdifferentfromtherestinthatforseveralparagraphswelearnabouteventsduringthenight.Thenewdayisnotmentioneduntil5.3.57, after a resumptive ‘closing line’: tØn m¢n dØ nÊkta oÏtvw

§poreÊonto: §pe‹ d¢ ≤m°ra §g°neto…2Whereasthetransitionsinthese14instancesoccurinpassagesthat

arealltemporallyorganized,passages[j]and[k]areorganizedaround

———2 Note that [m]is indeeddifferentlyorganized:herethe initial d°connectsthe

sentenceinwhichwefindthe m°n-membertothepreceding(asamatteroffact,itisthe d°-memberofa m°n… d°sequence),whereastÒte m°n…≤n¤ka d°operatesunderthescopeoftheinitial d°.AsregardstheusageofsubclausesinCyropaedia5.3.51-57,thelargercontextof[m],itistellingthatatthepointwherethereisneedofa

temporalreference,i.e.after tÒte m°n,asemanticallyspecific(temporal)subordinatorsuchas ≤n¤kaisused(comparetheother13instanceswherewefindatemporalreference)atapointwherethenarrativeremainstemporallyorganized,whereasatapoint where a subclause is used to indicate a thematic boundary in an on-going narrativesequence,asinCyropaedia5.3.57,wefindthatan§pe¤-clauseisused;cf.thediscussionof[j]-[k]below.




their participants(ofl m¢n dØ émf‹ KËron

…ofl d¢ ÉAssÊrioi

andofl m¢n émf‹ tÚn KËron…Kro›sow m°ntoi,respectively).Still,thereisa

thematic break: there is referential discontinuity. However, theintroductionoftheaction‘goingtosleep’evokesthenotionof‘night,’and a reference to the break of day is to be expected. In a waycomparableto[m]§koimÆyhsan…≤n¤ka d' ∑n §n m°sƒ nukt«n…§pe‹ d¢ ≤m°ra §g°neto, this reference to the thematic break of temporaldiscontinuityisexpressedbymeansofan§pe¤-clause.Inthecaseof[j]thishappensin4.1.9afteraresumptive‘closingline’( oÏtv

dØ §kle¤pousi tÚ stratÒpedon ka‹ ép°rxontai t∞w nuktÒw. …w d'≤m°ra §g°neto ka‹ ¶rhmon éndr«n §fãnh tÚ t«n polem¤vnstratÒpedon …); in the case of [k] in 7.2.2 (§peidØ d¢ ≤m°ra§g°neto ), after we have been told what happened during the night.Thus, in [j]-[k] the temporal break is postponed; first the focus of attention shifts to the actions performed by a different participantduringthenight.ThefullNP ofl d¢ ÉAssÊrioi[j]isacommonmeanstoarticulateatopicswitch.InthecaseofKro›sow m°ntoi([k])thereisanextranuance:with m°ntoi ,theactionundertakenissignaledasdifferentfromwhatonewouldexpectonthebasisofthecontextorsituation.3TheactionundertakenbyCroesusisthatheandhisarmyfled straight towards Sardis (eÈyÁw §p‹ Sãrdevn ¶feuge sÁn t“strateÊmati ). The unexpectedness of this undertaking is clear fromCyrus’sreaction(7.2.2):§peidØ d¢ ≤m°ra §g°neto, eÈyÁw §p‹ Sãrdeiw∑ge KËrow (‘when daylight came, Cyrus led his army straight onagainstSardis’).

Inthecorpus,the m°n-memberisopenedbyka¤9outof16times.

Whereas d°isusedtopresentanextindependentstepinasequenceof events, ka¤ indicates that what follows is an addition to whatprecedes.4Otherdevicesthatmarktheconnectiontothepreceding include: m¢n oÔn(3x); m¢n dÆ (3x), and the use of anaphoric deicticelementssuchasoÏtvw(3x),taÊthn(1x),and§ke¤nhn(1x).Finally,§koimÆyhsanisanaoristindicative,andtheaoriststemisapttocloseoffasequence.5

Several linguistic features of Ancient Greek narrative have thusbeenillustratedonthebasisofthiscorpus.Ishallnowturntothe

———3 For m°ntoiusedfordenialofexpectation,seeSlings(1997).4 Cf.Bakker(1993)onka¤asacontinuity-marker.5 Theimperfect§koim«ntoisfoundonlyonceintheentireworksofXenophon

( Anabasis4.5.14),whereitisusedinarelativeclause.



32 CHAPTERTWO

grammarofexpressingsequentialactionswithspecialattentiontothehierarchythatobtainsamongthem.

On the level of text articulation, the use of the finite verb§koimÆyhsaniscentraltotheclosureofthethematicdiscourseunit.The simplest form of closing off a discourse segment is using anindependentclauseconsistingofonefiniteverb,asillustratedin[a]-[c](ka‹ tÒte m¢n oÏtvw §koimÆyhsan ). In these exampleswehave a‘oneeventsentence’.Example[d]isslightlymorecomplicated,butinfactstillcontainsoneevent(§koimÆyhsan ), for the perfect participal

clause…w oÔn éphllagm°noi toÊtvnreferstoastateofmindratherthan an event, and resumes previously expressed information(propositionaloverlap).

AlthoughinprincipleGreeksyntaxoffersdifferentalternativestoexpress two events in a sentence, in fact we only find an aoristconjunctparticiplefollowedbyafiniteverb([e]-[i]).Participialclausessuch as these are generally called clause chaining participles. Of course,theaoriststemisused,astheeventencodedbytheparticipialclauseisexpressedasaself-containedpieceofinformation(theeventsareapartofthenarrativeassertion).6Thattheyarenottimeadjunctsisindicatedbythepresenceofotherconstituentsprovidingatemporalreference.Thechronologyofthetwoeventsisdeterminedbythewaytheyarepresented(clausechainingparticiplefirst,thenthefiniteverb)and the hearer’s/reader’s knowledge of the world. In terms of textarticulation though, theclausechainingparticiplesare noton aparwiththefiniteverboftheirsentence.‘Takingdinner’([e],[g]),‘being separatelyquartered’[f],‘beingpresentedwithgifts’[h]or‘stationing

advanceguards’[i]arelesscentraltothethematicstructureofthediscourse.Theaction‘goingtosleep’isthereforepresentedasthemaininformation,whereastheotheractionsarepresentedbyclausechainingparticiples.Theusageofanaoristparticiplefollowedbyafiniteverb,therefore, isa meansto express two independent events,andatthesametimetoconstituteamongthemahierarchythatisfunctional with respect to the articulation of the larger passage of whichtheyarepart.

———6 Contrastfor instance §n fulakª ¶xontew tÚn k≈marxon ka‹ tå t°kna aÈtoË

ımoË §n Ùfyalmo›w in [f], where the (postposed) present participle providesinformationthatistobeconnectedtothepreceding§koimÆyhsan.




Nowwhatifthenarratorwishestoexpressmorethantwoeventsinasentence?Thefirstoptionseemstobe:addingaclausechaining participle; in principle there is no limit to extending a chain of participles.7Inthecorpus,however,thisoptionisnotfound.Instead,we find two otheroptions, both involving the use of ka¤ :eithertheclausechainingparticiplesarecombined,resultingincombinationsof two([j]-[k])orthree([l]-[m])participialclauses,orthefiniteverbsarecombined ([n]). The use of ka¤ involves linking of what belongstogetherassetofffromtherest.Thislinkingoccursbothwiththe

clause chaining participles and with the finite verbs, i.e. on theappropriatehierarchicallevel.Thus,ifwecompare[e]to[j]-[k],thelowerhierarchicalleveloccupiedby deipnopoihsãmenoialonein[e]is extended by one event ( fulakåw katasthsãmenoi ) in [j]-[k];comparing [j]-[k] to [l]-[m], again the lower hierarchical level isextendedbyoneevent(purã… kaÊsantew and suskeuasãmenoipãnta, respectively), whereas if we compare [e] to [n], in [n] thelower hierarchical level is now takenbysunelkÊsantew toÁw t«npolem¤vn nekroÁw e‡sv fãlaggow,whereastheaction‘takingdinner’is transferred to the higher hierarchical level and is linked to theaction‘goingtosleep’.

Thechoiceofthehierarchicallevelatwhichthelinkingtakesplaceisnotrandom.Herewegobacktothe NMPthatwasthepointof departureforthisdiscussion.

Xenophon could have written *sunelkÊsantew (…) k a ‹ deipnopoihsãmenoi §koimÆyhsanin Ages i laus 2.15 ([n] = [2]),expanding the sequence he wrote in the parallel passage in the

Hell enica 4.3.20([e]=[1])withoneparticipialclause.Letusfirstobserve that if we find linking of participial clauses before§koimÆyhsan,we always find the series to beopened bytheaction‘takingdinner’( deipnopoihsãmenoiin[j]-[l]; deipnÆsantewin[m]),andnaturallysobecauseofthechronologyintheRealWorldthatismirroredbythewayeventsarepresentedinthetext.Theactionsthatare combined with the action ‘taking dinner’ by ka¤ areagainof relativelylowimportanceforthediscoursesegmentasawholeatthepointwhere it ends: theseactions belong to the routineofmilitary

———7 For example: Anabasis 1.1.7 ı d¢ KËrow Í toÁw feÊgontaw

strãteuma § M¤lhton…(‘Cyrust theexilesu,c anarmy,andl Miletus’).



34 CHAPTERTWO

warfare—‘stationingsentinels’([j]-[m],noteÀw/Àsper ¶dei

), ‘lighting firesinfrontoftheoutposts’[l],or‘packingupeverything’[m]—andareeasilyaddedtotheschemeoncetheaction‘takingdinner’hasbeenmentioned.In Agesilaus2.15thesituationisdifferent.Itisclearthattheclause sunelkÊsantew toÁw t«n polem¤vn nekroÁw e‡sv

fãlaggow,lackinginthe Hellenica,isrelatedtothedescriptionofthebattlefield after a battle in the Ages ilaus (2.14), also lacking in the Hellenica.Althoughtheprecisereasonfortheappearanceoftheclauseremainsobscure,8 it clearly expresses an independent step in the

sequenceofevents,onewhichisnotpartofthetypicalscheme.Thus,*sunelkÊsantew (…)ka‹ deipnopoihsãmenoi §koimÆyhsanwouldhaveresultedinlinkingtwoactionsthatarenoteasilycombinedfroman informational point of view, which a for tio ri applies to* deipnopoihsãmenoi ka‹ sunelkÊsantew (…)§koimÆyhsan ,asthisorderwouldnotmirrortheorderofRealWorldevents.9Therefore,thingscouldhardlyhavebeenarticulatedinadifferentmanner.

Agesilaus2.15showshowlinkingandhierarchyareinvolvedinthechoice between alternative ways of expressing actions (note that

§deipnopoiÆsanto ka‹ §koimÆyhsanheretogetherbuildupthesetof actions typically found at the closure of a discourse segment). Yetotheroptionsareavailable.Ifweconsiderthepossiblecombinationsof (preposed) participles and finite verbs, three actions may beexpressedinonesentenceinthefollowingways:

———8 Itdependsonwhetherweacceptthewordst«n polem¤vn,transmittedbythe

manuscriptsbutregardedbysomescholarsasproblematic(t«n pol.del.Schneider,§k t«n pol.Weiske, t«n épolom°nvnJacobs).Ifwedonot,thereasonforthepresence of the clause might be that once we have been told that bodies lay all

around,itmightbetakenasanoffenceagainstGreekreligiontoleavethemthereduringthenight.Butifwedoaccept themanuscriptreading,whichweshould,theclause is there to highlight Agesilaus’s behavior in accordance with Xenophon’sencomiasticgoal.

9 The action ‘taking dinner’ occupies the front position only when actionsbelongingtotheeveningroutinesofmilitarywarfareareinvolved;cf.[j]-[m].




Figure 2: Combinations of three actions expressed by (preposed) participles and finiteverbs

1.FINITEVERBka‹FINITEVERBka‹FINITEVERB

2.PARTICIPIALCLAUSE+FINITEVERBka‹FINITEVERB

3.FINITEVERBka‹PARTICIPIALCLAUSE+FINITEVERB

4.PARTICIPIALCLAUSEka‹PARTICIPIALCLAUSE+FINITEVERB

5.PARTICIPIALCLAUSE+PARTICIPIALCLAUSE+FINITEVERB

Option1attributesequalstatustothethreeactions,whichis,forthereasonsmentionedabove,uncalledforin Agesilaus2.15,asislinking thefirsttwoactionsandsettingthemofffromthethird(option4),orrefraining from linking and expressing the three actions as threeindividual and informationally autonomous successive steps in thenarrative sequence, giving the first two actions the same, yet lowerhierarchicalstatuswithrespecttothelast(option5).Onlyoption3seems to be a real alternative for the one Xenophon has actuallychosen.Yetthisoptionisnotviableheresincesuchasequencewouldhavepresentedtheaction‘dragging’andtheaction‘goingtosleep’asequallyimportantstepsintheorganizationofthedevelopmentofthenarrative. Now that a discourse segment is closed off, the action‘draggingthebodies’isnotonaparwiththeaction‘goingtosleep’,around which the action sequence within a larger whole isthematicallyorganized.

Option 3 is in fact found in [o]. Here, ép∞lyon §p‹ tÚstratÒpedonand§koimÆyhsanarelinkedonthesamehierarchicallevelbyka¤,withthe‘takingdinner’hierarchicallydowngradedtoaparticipialclause in the second halfof the set. Although the action‘goingbacktothecamp’couldhavebeenexpressedbyaparticiple(* épelyÒntew §p‹ tÚ stratÒpedon ),Xenophon’sdecisionnottodososeemstobemotivatedbythefactthatthisactioninvolvesarelocationimportantenoughforthedevelopmentofthenarrativetowarrantafiniteverb( ép∞lyon ).

Example[p],finally,isthemostcomplicatedone,anddifferentfrom the rest. It consists of five actions, expressed by a presentparticiple and an aorist participle linked by ka¤ preceding animperfectfinitemainverb;thismainverbislinkedbyasecond ka¤to




thetext,inthatthisverbprovidesthemostbasicbuildingblockofthenarrative. This is most conspicuous in examples [a] through [c],where the action sequence closes off with §koimÆyhsan withoutexpressionoffurtheractions.If,however,everynarrativesequencewerenarratedinthisway,thetextwouldloseits texture:therewouldbe no ‘relief’, so to speak, because the actions constituting thestructuralbackbonewouldnotbedistinguishedfromthesurrounding material.Atthispoint,thepreposedparticipialclausescomein.Inallinstances in the corpus where a sequence consists of one or more

preposedparticiplesandisclosedoffbythesinglefinitemainverb§koimÆyhsanwhichexpressesthecentralaction([d]through[m]),theinformation status of the preceding actions is low; the participlesexpressactionsthatbelongtotheroutineofmilitarywarfare,ortheparticiple is resumptive (either by summary, as in [d], or byrestatement,asin[h]).Sometimesthenarratorusesmorethanonefinitemainverbin the sequence([n]through[p]).In [o]and[p]hebreaksupthesequenceinpartsinordertopreventthereadersfromlosinggriponthenarrativesequence.Inthecorpus,example[n](=[2])standsout.In Agesilaus2.15theaoristconjunctparticipialclausesunelkÊsantew toÁw t«n polem¤vn nekroÁw e‡sv fãlaggowexpressesanactionthat,initscontext,hashighinformationstatus;theaction isratherunexpectedandofsignificancefor thepictureof AgesilausthatXenophoniscreatinginhisencomium.Thethematicstructureofthesetend-of-day scheme,however, ispreservedbythefinitemainverbs§deipnopoiÆsanto ka‹ §koimÆyhsan;asaresultof this,theaction‘theytookdinner’isexpressedbyafinitemainverb

here,linkedby ka¤ on the appropriate hierarchical level, so as tocreate a set of structural elements (§deipnopoiÆsanto ka‹§koimÆyhsan )that islinguisticallydistinguishedfromthesurrounding information (sunelkÊsantew…). Ag e s i l aus 2.15 ([n] = [2]) iscomparabletoyetdifferentfrom Hellenica4.3.20([e]=[1]),wheretheparticiple deipnopoihsãmenoi isusedmerelytoexpressanactionof ratherlowinformationstatusbeforethefinitemainverb§koimÆyhsanclosesoffthesequenceasastructuralbuildingblockinthethematicorganizationoftwoadjacentdiscourseunits.Weconcludethatanactionmaybehierarchicallyupgraded(ordowngraded,ifwetakethe Agesilaus-passageasour pointof departure)forreasonspertainingtotheneedsofaspecificcontext.



38 CHAPTERTWO

Subclauses are absent among the comparable means of clausecombininginthiscorpus.Thisissignificantandpointsinthedirectionofapositionwhere,contrarytowhatoftenisthoughtandtaught,weshouldregardsubclausesasasubstantiallydifferentcategoryinclausecombining as far as the organization of narrative discourse isconcerned. In the next NMPwewillconsiderthealternationofa(present stem) conjunct participial clause and a (present stem)subclause(§pe¤-clause).

2 NMP2:Anabasis6.2.13-15vs.Hellenica2.4.10-11

[3] Anabasis6.2.13-15 [4] Hellenica2.4.10-11Jenof«n m°ntoi §boÊleto koinª~met' aÈt«n~ tØn pore¤anpoie›syai, nom¤zvn oÏtvw ésfalest°ran e‰nai μ fid¤& ßkastonst°llesyai: éllå N°vn ¶peiyen aÈtÚn kay' aÍtÚn poreÊesyai, ékoÊsaw toË XeirisÒfou ˜tiKl°androw ı §n Buzant¤ƒ èrmostØw fa¤h triÆreiw ¶xvn ¥jein efiwKãlphw lim°na: ˜pvw oÔn mhde‹w metãsxoi, éll' aÈto‹ ka‹ ofl aÈt«nstrati«tai §kpleÊseian §p‹ t«ntriÆrvn, diå taËta suneboÊleue.ka‹ Xeir¤sofow, ëma m¢n éyum«nto›w gegenhm°noiw, ëma d¢ mis«n §ktoÊtou tÚ strãteuma, §pitr°pei aÈt“ poie›n ˜ ti boÊletai.Jenof«n d¢ ¶ épallage‹w t∞w stratiçw§kpleËsai: y

, pÒtera l“on ka‹ êmeinon e‡h strateÊesyai ¶xontitoÁw parame¤nantaw t«nstrativt«n μ épallãttesyai,§sÆmhnen ı yeÚw to›w flero›wsustrateÊesyai.Xenophon,however,wasdesirousof makingthejourneyincompanywithCheirisophus,believingthatthiswas

§k d¢ toÊtou labΔn ı YrasÊboulowtoÁw épÚ Ful∞w per‹ xil¤ouw ≥dhsuneilegm°nouw, éfikne›tai t∞wnuktÚw efiw tÚn Peiraiç. ofl d¢triãkonta §pe‹ æsyonto taËta,eÈyÁw §boÆyoun sÊn te to›wLakvniko›w ka‹ sÁn to›w flppeËsi ka‹to›w ıpl¤taiw: ¶peita §x≈roun katåtØn efiw tÚn Peiraiç èmajitÚn énaf°rousan.ofl d¢ épÚ Ful∞w ¶

mØ éni°nai aÈtoÊw,

,sunespeirãyhsan §p‹ tØnMounix¤an.




asaferplanthanforeachofthemtoproceedindependently;butNeonurgedhimtogobyhimself—hehadheardfromCheirisophusthatCleander,theLacedaemoniangovernoratByzantium,hadsaidhewascomingtoCalpeHarbourwithtriremes;itwasNeon’spurpose,then,thatnooneelseshouldgetashareinthisopportunity,butthathehimself andCheirisophusandtheirsoldiers

shouldsailawayuponthetriremes,andthiswasthereasonforhisadvicetoXenophon.AndCheirisophus,hewassodespondentoverwhathadhappenedand,besides,feltsuchhatredtowardthearmyforitsaction,thatheallowedNeontodowhateverhechose.

getclearofthearmyandsailawayhome;

astowhetheritwasbetterandmoreproperforhimtocontinuethe

journeywithsuchofthesoldiersashadremainedwithhim,ortoberidofthem,thegodindicatedtohimbythesacrificesthatheshouldstaywiththem.

HereafterThrasybulustookthemenofPhyle,whohadnowgatheredto

thenumberofaboutonethousand,andcamebynighttoPiraeus.TheThirtylearnedofthis,andsoatoncesetoutagainsthim,withtheLaconianguardsmenandtheirowncavalryandhoplites;thentheyadvancedalongthecarriageroadwhichleadsuptoPiraeus.ThemenfromPhyle,f

preventtheircomingup,y

,theygatheredinacompactbodyonthehillofMunichia.

Wehavetwodifferentstoriesthistime,yettheexamplesshowtheuseofacomparableconstruction:wehaveanindependentmainclausewithanaoriststemverbalcontituentinbothcases.Inthismainclauseaparticipanttriestodosomethingforsometime(¶ti m¢n §pexe¤rhs-);thecorresponding d°-memberopenswithapreposed(presentstem)conjunctparticipialclausein[3]andapreposed(presentstem)§pe¤-clausein[4];withinthesequenceasawholetheconjunctparticipialclause and the subclausemay beseenas alternative clause types inclausecombining.

In [3], the question is addressed how Xenophon, Neon, andCheirisophus thought the journey should be continued. Neon hadtriedtopersuadeXenophontogobyhimself( N°vn ¶peiyen aÈtÚnkay' aÍtÚn poreÊesyai ). This is indeedwhat Xenophon considered



40 CHAPTERTWO

doingforsometime(Jenof«n d¢ ¶ti m¢n §pexe¤rhsen épallage‹wt∞w stratiçw §kpleËsai ).10 The suggestion that he would not get

clearofthearmyandsailawayhomeiselaboratedoninthenextsentence:HeraclestheLeader,whomheconsultedastowhetheritwasbetterandmoreproperforhimtocontinuethejourneywithsuchof the soldiers as had remained with him, or to be rid of them,indicatedtohimbythesacrificesthatheshouldstaywiththem.

In[4],ThrasybulustakesthemenofPhyleandarrivesatPiraeus.Atopicswitchoccurs:theThirty(ofl d¢ triãkonta )are the newtopic

andwearetoldwhattheydidwhentheylearned11ofThrasybulus’sarrival:theysetouttotherescueofthoseinPiraeus(§boÆyoun );next,they advanced along the carriage road which leads up to Piraeus(§x≈roun ).Astheimperfectsindicate,theseactionsaresupposedtobeconnected with the sequel, and the action sequence ends with thereaction of the men from Phyle (again a topic switch: ofl d¢ épÚFul∞w ),reported byaorists: forsometimethey triedtoprevent theircomingup,buteventuallytheygatheredinacompactbodyonthehillofMunichia.

Ineitherexample,thereader/hearerisinformedonthebasisof thelexicalmeaningof ‘§pexe¤rhs-’,thatwhatiscommunicatedhereis an attempt to undertake an action ( épallage‹w t∞w stratiçw§kpleËsaiand mØ éni°nai aÈtoÊw,respectively).Thepresenceof m°n,on the other hand, creates the expectation that the attemptundertakenwillbeabandoned,tobefollowedbyadifferentaction.Whetherthesentencecontainingthecorresponding d°openswitha(conjunct)participialclause,as in[3],orwithasubclause,asin[4],

thearticulationbymeansof m°n… d°servesthepurposeofmarking discoursecomplementation. 12 Therefore, it is implausible that thesubordinate clauses, although having been preposed, perform thefunctionofdiscoursesegmentationhere.Rather,itwillbeargued,thedifferenceinsyntaxreflectsthedifferenceintheRealWorldrelationthatobtainsbetweentheeventsexpressedinthetwoclausesofthe

———10 See Lendle (1995, 376): “Die Formulierung ¶ti m¢n §pexe¤rhsen…(vgl.

Hell.2.4.11)isthierindemSinnzuverstehen,daßerbereitsseiteinergewissenZeitundnachwievorerwog,dasHeerzuverlassen”.11 A subclause of the type §pe‹ æsyonto taËtaoftenindicatesfocalizationat

pointswhereadiscourseparticipantundertakesanactioninresponsetoanactionof someotherdiscourseparticipant.

12 Asopposedtotransition-marking m°n… d°;seetheIndexofLinguisticTerms.




clausecombination.Thetwosubordinateclausesunderconsiderationhavepresentstemverbalconstituents.Fromapresentationalpointof view,thismeansthattheverbalactionisnotindependentlyasserted,buthastobeconnected withtheverbalactionofthemainproposition.

In[3],an§pe¤-clause,ifatallapttoindicatetheobtainingRealWorld relation, would have needlessly complicated participanttracking.Thenarratorpresentsthereaderwithinformationnecessaryforacorrectunderstandingofthemainproposition(§sÆmhnen ı yeÒw )by using the conjunct participle yuom°nƒ d¢ aÈt“ t“ ≤gemÒni

ÑHrakle› ka‹ koinoum°nƒ, pÒtera…:withouttheinformationthatXenophonofferedsacrifice,thetextwouldhavebeenincoherent.Asthingsstand,thedativecase(yuom°nƒ … aÈt“ )isusedtoexpresstherelationobtainingintheRealWorldbetweentheeventsexpressedinthetwoclausesoftheclausecombination,viz.,thatsomeoneofferedsacrificeand that the god responded, byfilling the slotof the thirdcomplementto §sÆmhnen.Atthesametime,itpreparesforatopicswitchtoı yeÒw.

In[4]thereisthespecificcontributionofthesubordinator§pe¤.Inhis (re-)construction of reality in the formofnarrative, the speakerdoes several things at the same time: he records two Real World‘events’13:1)thatthelineofthetownwallseemedtorequirealargeforceforitsdefence,and2)thatthemenfromPhylegatheredinacompactbodyonthehillofMunichia,andfurther ,byusing§pe¤,hesignalsthataccordingtohimarelationexistsbetweenthetwoeventsthatare combined, without specifying the natureof this relation.In[4], we are presented with the speaker’s evaluation of the reality

situation that brings the men from Phyle—the current discoursetopic—roundtoperformtheactionwithwhichtheparagraphcomestoaclose:‘inviewofthefactthat ,<tothem>,thelineofthetownwall,extensiveasitwas,seemedtorequirealargeforceforitsdefence,whereas they were not yet numerous, they gathered in a compactbodyonthehillofMunichia’.

ThepresentationoftheRealWorldsituationin[4]bymeansof thesubclausehasthe effect thattheeventexpressedinthesubordinateclauseisrelayedfromthepointofviewofthecurrentdiscoursetopic,the men from Phyle. This effect is established bya combinationof threefactors:1. theorderinwhichthetopicandthepresentstem

———13 Forthenotionof‘event’,seeChapter1,Section2.1.



42 CHAPTERTWO

subclausearepresented;142.thelexicalmeaningof§dÒkei

itself;3.thedeletion of the dative phrase to §dÒkei indicating to whom thesituationseemedthecase.Thesituationfoundinexample[4]shouldthereforebeinterpretedas:“forsometimethemenfromPhyletriedtopreventtheircomingup,butthenthey realized thatthelineofthetownwallrequiredalargeforceforitsdefence,and therefore<theygaveupthisattemptand>theygatheredinacompactbodyonthehillofMunichia”.Thusparaphrased,thereader/hearerstillmissessomeessentialinformation,viz.,whythefactthatthelineofthetown

wallrequireda largeforceforitsdefencewouldbeaproblemforthemenfromPhyle.Theanswertothisquestionis:<theyrealizedthat>the line of the town wall was extensive and they were not yetnumerous.15Thenarratorprovideshisreaderwiththisinformationbymeans of two conjunct participles inside the §pe¤-clause( m°gaw…n/oÎpv pollo›w oÔsi ). The information provided by theseparticipial clauses that are part of the construction may be called‘subsidiary’ to the extent that it does not belong to the mainproposition,butistheretohelpthereaderinhisattempttodecodethemeaningcontainedinthemainproposition.

The major upshot of the analysis of this N M P consists in theobservationthatthechoicebetweenaconjunctparticipialclauseandasubclause(§pe¤-clause)maybedeterminedbyfactorsconcerningthepresentationofdifferentRealWorldrelationsbetweenthecontentof thesubordinateclauseandthecontentofitsmatrixclause.In[3]sucharelationobtains,butitisdifferentfromtheonein[4],inthatin[3]—asituationinwhichonemayexpectaresponsefromthegod,so

thatatopicswitchoccurs( Jenof«n ’ ı yeÒw )—therelationbetweenthe act of sacrificing and the answer to it is self-evident, and isprimarily expressed by the dative case of the conjunct participle,

———14 Cf.Sicking,1996:71: “Iftheseobservationsarecorrect,preposedPS[present

stem]subclauses,apartfrompreparingthereader’s understanding ofthemainclause,atthesametimecanbeusedfordirectinghis perspective,orpointofview,bymakingthedramatispersonathe‘focalizer’oftheeventsmentionedinthesubclause.ItwillbenotedthatthisisnotacharacteristicofpreposedPSsubclausesassuch,butaresult,orside

effect,of1.thetopicofthe<mainclause>havingbeenestablishedbeforeasubclauseaboutadifferenttopiccomesin,and2.PSmakingthesubclauseandthemainclauseintooneintegratedunitofinformation,thusbringingthecombinationofthetwounderthe‘scope’ofthetopicinvolved.”

15 Thatnumbersareto increaseontheirsideis apparentlyto beexpected,andthisexpectationisindeedmetinthesequel;cf. Hellenica2.4.11-12.




functioningasthethirdcomplementto§sÆmhnen

.In[4],wheretheadjacentclausesofthe§pe¤-clausehavethesamesubject(ofl d¢ épÚFul∞w ), the speaker presents anevaluationofthereality situation bythe participants that brings them round to perform the actionpresentedinthematrixclause.Thenon-specificrelator §pe¤indicatesthat the subordinated proposition is to be related to the mainproposition;thewaythespeakerpresentsaRealWorldsituationintheclausecombination,incombinationwiththelinguisticmeansusedto present the subordinated situation, has the effect that the

subordinated situation is relayed from the point of view of thediscourseparticipants.Inthecaseoftheconjunctparticipialclausein[3],thesituationisexternallyfocalized.

InthenextNMPwewillelaborateonthediscussionof NMP1& 2,asintwocomparablecontextsofsomelengthwewillagainencounterthe alternationofan (aorist stem)conjunctparticipialclause and an(aoriststem)independentfiniteverb,aswellasthealternationofa(present stem) conjunct participial clause and a (present stem)§pe¤-clause.

3 NMP3:Hellenica3.4.15-16vs.Agesilaus1.19-25

[5] Hellenica3.4.15-16 [6] Agesilaus1.19-25… diå m¢n dØ taËta eÈyÁw polloÁw§raståw t∞w aÍtoË fil¤aw§poiÆsato. gign≈skvn d' ˜ti ≤ m¢nporyoum°nh ka‹ §rhmoum°nh x≈ra

oÈk ín dÊnaito polÁn xrÒnonstrãteuma f°rein, ≤ d' ofikoum°nh m¢n speirom°nh d¢ é°naon ín tØntrofØn par°xoi, §pem°leto oÈ mÒnon toË b¤& xeiroËsyai toÁw§nant¤ouw, éllå ka‹ toË pr&Òthtiprosãgesyai. ka‹ pollãkiw m¢nprohgÒreue to›w strati≈taiw toÁw èliskom°nouw mØ …w éd¤kouwtimvre›syai, éll' …w ényr≈pouw

ˆntaw fulãttein, pollãkiw d¢ ıpÒte metastratopedeÊoito, efi a‡syoitokataleleimm°na paidãria mikrå§mpÒrvn, ì pollo‹ §p≈loun diå tÚnom¤zein mØ dÊnasyai ín f°rein



44 CHAPTERTWO

genom°nhw d¢ taÊthw t∞wflppomax¤aw, yuom°nƒ t“ÉAghsilãƒ tª Ístera¤& §p‹ proÒdƒ êloba g¤gnetai tå flerã. toÊtou m°ntoi fan°ntow str°caw §poreÊeto§p‹ yãlattan.

efi mØ flppikÚnflkanÚn ktÆsaito, oÈ dunÆsoito katåtå ped¤a strateÊesyai, ¶gnv toËtokataskeuast°on e‰nai, …w mØ drapeteÊonta poleme›n d°oi.ka‹ toÁw m¢n plousivtãtouw §kpas«n t«n §ke› pÒlevnflppotrofe›n kat°leje: p d°, ˜stiw par°xoito ·ppon ka‹ ˜plaka‹ êndra dÒkimon, ˜ti §j°stai aÈt“ mØ strateÊesyai, § oÏtvtaËta suntÒmvw prãttesyaiÀsper ên tiw tÚn Íp¢r aÍtoË époyanoÊmenon proyÊmvw zhto¤h.

§k d¢ toÊtou §peidØ ¶ar Íp°faine,sunÆgage m¢n ëpan tÚ strãteuma

efiw ÖEfeson:

aÈtå ka‹ tr°fein, §pem°leto ka‹toÊtvn ˜pvw sugkom¤zointÒ poi.to›w d' aÔ diå g∞rawkataleipom°noiw afixmal≈toiwpros°tatten §pimele›syai aÈt«n,…w mÆte ÍpÚ kun«n mÆy' ÍpÚ lÊkvn diafye¤rointo. Àste oÈ mÒnon oflpunyanÒmenoi taËta, éllå ka‹ aÈto‹ ofl èliskÒmenoi eÈmene›w aÈt“§g¤gnonto. ıpÒsaw d¢ pÒleiwprosagãgoito, éfair«n aÈt«n ˜sa doËloi despÒtaiw ÍphretoËsipros°tatten ˜sa §leÊyeroi êrxousipe¤yontai: ka‹ t«n katå krãtow énal≈tvn teix°vn tª filanyrvp¤&ÍpÚ xe›ra §poie›to.

m°ntoi énå tå ped¤a oÈd¢ §n tªFrug¤& § strateÊesyai diåtØn Farnabãzou flppe¤an, ¶dojen aÈt“ flppikÚn kataskeuast°one‰nai, …w mØ drapeteÊonta poleme›n d°oi aÈtÒn. toÁw m¢n oÔnplousivtãtouw §k pas«n t«n §ke›pÒlevn flppotrofe›n kat°leje.

d°, ˜stiw par°xoito ·pponka‹ ˜pla ka‹ êndra dÒkimon, …w§j°soito aÈt“ mØ strateÊesyai:

oÏtvw ßkastonproyÊmvw taËta prãttein Àsper êntiw tÚn Íp¢r aÍtoË époyanoÊmenonproyÊmvw masteÊoi. ¶taje d¢ ka‹

pÒleiw §j œn d°oi toÁw flpp°awparaskeuãzein, nom¤zvn §k t«nflppotrÒfvn pÒlevn eÈyÁw ka‹ fronhmat¤aw mãlista ín §p‹ tªflppikª gen°syai. ka‹ toËt' oÔn égast«w ¶doje prçjai, ˜tikateskeÊasto tÚ flppikÚn aÈt“ ka‹eÈyÁw §rrvm°non ∑n ka‹ §nergÒn.§peidØ d¢ ¶ar Íp°faine, sunÆgagepçn tÚ strãteuma efiw ÖEfeson:

Theimmediateresultwasthathehadmanyardentsuitorsforhisfriendship.Recognizingthatacountryplunderedanddepopulatedcouldnotlong




AfterthiscavalrybattleAgesilausonthenextdaywasofferingsacrificeswithaviewtoanadvance,whentheliversofthevictimswerefoundtobelackingalobe.Aftertheunexpectedappearanceofthissign,heturnedandmarchedtothesea.

,unlessheobtainedanadequatecavalryforce,hewouldnotbeabletocampaignintheplains,heresolvedthatthismustbeprovided,sothathemightnothavetocarryonaskulkingwarfare.Andaccordinglyheassignedtherichestmenofallthecitiesinthatregiontothedutyofraisinghorses;b

thatwhoeversupplieda

horseandarmsandacompetentmanwouldnothavetoservehimself,hthesearrangementstobe

carriedoutwithalltheexpeditionthatwastobeexpectedwhenmen

supportanarmy,whereasaninhabitedandcultivatedlandwould

yieldinexhaustiblesupplies,hetookpainsnotonlytocrushhisenemiesbyforce,butalsotowinthemoverbygentleness.Andhewouldoftenwarnhismennottopunishtheirprisonersascriminals,buttoguardthemashumanbeings;oftenwhenshifting camp,ifhenoticedlittlechildren,thepropertyofmerchants,left

behind—manymerchantsofferedchildrenforsalebecausetheythoughttheywouldnotbeabletocarryandfeedthem—helookedafterthemtoo,andhadthemconveyedtosomeplaceofrefuge.Again,hearrangedthatprisonersofwarwhoweretoooldtoaccompanythearmyweretobelookedafter,thattheymightnotfallapreytodogsorwolves.Itthus

cameaboutthathewonthegoodwillnotonlyofthosewhoheardofthesefacts,butevenoftheprisonersthemselves.Inhissettlementwiththecitiesthathewonover,heinvariablyexcusedthemfromallserviledutiesandrequiredonlysuchobedienceasfreemenowetotheirrulers;andbyhisclemencyhemadehimselfmasteroffortressesimpregnabletoassault.

hewasnotabletocampaignintheplainseveninPhrygia,owingtoPharnabazus’cavalry,hedecidedthatacavalryforcemustbeprovided,sothathemightnothavetocarryonaskulkingwarfare.Hethereforeassignedtherichestmenofallthecitiesinthatregiontothedutyof raisinghorses;further,h

thatwhoeversuppliedahorseandarmsandacompetentmanwouldnothavetoservehimself;a inthiswayh thateveryoneofthemcarriedouttheserequirementswith



46 CHAPTERTWO

wereeagerlylookingforsubstitutestodieintheirstead.

Afterthis,atthefirstsignofspring,hegatheredhiswholearmyatEphesus.

thezealofamaninquestofasubstitutetodieinhisstead.Healsospecifiedcitiesthatweretofurnishcontingentsofcavalry,feeling surethatfromthehorsebreeding citiesridersproudoftheirhorsemanshipwouldbeforthcoming.Thisagainwasconsideredanadmirablestrokeonhispart,thatnosoonerhadheraisedhiscavalrythanitbecameapowerfulbodyreadyfor

action.AtthefirstsignofspringhegatheredhiswholearmyatEphesus.

Inthesetwoparallelpassagesfromthe Hellenicaandthe Agesilausthesame Real World situation is described. The passage that will beconsideredindetaildealswiththesubjectoftheraisingofamountedforce by Agesilaus. The same Real World experience has beentransformedintotextinadifferentmannerinthetwopassages,both

withrespecttotheamountofinformationthatisconveyedandthelinguisticmeansthatareusedtoarticulatethetext.Itisimportanttorememberthatthe Hellenicaandthe Agesilausare

of a different text type: they are a historical narrative and anencomium, respectively. In the Hellenica,theraisingofamountedforce is described as part of an ongoing narrative sequence. Thecorrespondingpassageinthe Agesilausispartofadifferentlyplannedpieceofdiscourse.Sections1.6-2.31oftheencomiumarededicatedtotheactions(¶rga )ofAgesilaus’sreign16coveringmorethanhalfofthe

text. Inhis account ofAgesilaus’sactions,Xenophon uses narrativeepisodestoillustrateAgesilaus’squalitiesasageneral,for,hesays, I believethathisdeedswillthrowtheclearestlightonhisqualities .17Hehas

———16 Theopeningandclosureofthispartofthetextaremarked:1.6:˜sa ge mØn §n

tª basile¤& dieprãjato nËn ≥dh dihgÆsomai (‘I will now give anarrative of theachievementsofhisreign’)…3.1:ka‹ taËta m¢n dØ e‡rhtai ˜sa t«n §ke¤nou ¶rgvn

metå ple¤stvn martÊrvn §prãxyh(‘such,then,istherecordofmyhero’sdeeds,sofarastheyweredonebeforeacrowdofwitnesses’).Similarly,theopeningandclosure

oftheaccountofAgesilaus’sactivitiesinAsiaaremarked:1.10: §n to¤nun tª ÉAs¤&¥de pr≈th prçjiw §g°neto(‘well,hisfirstactinAsiawasthefollowing’)…1.38: t«n m¢n dØ §n tª ÉAs¤& prãjevn toËto t°low §g°neto(‘thisthenwastheendofhisactivitiesinAsia’).

17 Agesilaus 1.6:˜sa ge mØn §n tª basile¤& dieprãjato nËn ≥dh d : épÚ går t«n ¶rgvn ka‹ toÁw trÒpouw aÈtoË kãllista nom¤zv katadÆlouw ¶sesyai.




chosentheformofnarrative,forhowcouldonegiveaclearerimpressionof whatkindofgeneralhewasthanbynarratingthethingshedid? 18The Agesilausonthewholemightberegardedas‘behavioraldiscourse’,albeitwithnarrativechunksinterwovenintheaccountofthehero’sactions:19thetext is agent oriented, and Agesilaus’s actions are narrated inchronologicalorder,yetnarrativeepisodesarepresentedin‘blocks’,that are selected for the purpose of eulogy. It is clearlythe speakerwhopresentsthenarrativeepisodesasexamples:heis,explicitlyorimplicitly,presentthroughoutthetext,20oftencommentinguponthe

actions by giving an evaluative statement. It is therefore to beexpected that the linguistic and informational differences in thecorresponding passage are linked up with the difference in theconstitutionofthesetexts.

Thetextprecedingthecorrespondingpassageisnotcomparable.Inthe Hellenica,whenXenophonwrites§poreÊeto §p‹ yãlattan,itshouldbeunderstoodthatthisisinthedirectionofEphesus,towardshiswinterheadquarters.Ephesusis,indeed,thelocationin3.4.16.Ontheleveloftextarticulation,anewtextsegmentisstartedwiththeconjunct participial clause gign≈skvn d°…;thistextsegmentiscloselyboundupwiththeprecedingonewithinalarger,thematicallycoherentwhole:theraisingofthemountedforcefitsintotheon-going sequence of retreat and marching to Ephesus. In the Agesilaus,thepassage under consideration is preceded by a large episode onAgesilaus’sconducttowardsothersinwarfare,meanttoillustratehis

filanyrvp¤a.21Withtheclause§pe‹ m°ntoi énå tå ped¤a oÈd¢ §n tª

———18 Agesilaus1.9:…p«w ên tiw saf°steron §pide¤jeien …w §stratÆghsen μ efi

aÈtå d ì ¶prajen;19 Foracharacterizationofdiscoursetypes,seeLongacre,1983.20 Cf. in the present passage: (1): d°oi aÈtÒn( Agesilaus )vs. d°oi ( Hellenica ): the

rathersuperfluous aÈtÒnindicatesthatthespeakeristalkingaboutAgesilaus;(2):…w§j°soito( Agesilaus )vs.˜ti §j°stai ( Hellenica ): by using the indicative in indirectdiscoursetheauthorrendersthequotationverbatim,wherastheoptativeisusedwhenthespeakerintegratesthequotationintohisnarration(Cf.Goodwin,1897,261(670):SIMPLESENTENCESININDIRECTDISCOURSE :IndicativeandOptativeafter˜tiand…wandinIndirectQuestions:“afterpasttensestheindicativeandoptativeareinequallygooduse;theoptativebeingusedwhenthewriterincorporatesthequotationentirely

intohisownsentence,andtheindicativewhenhequotesitintheoriginalwordsasfarashisownconstructionallows”).21 IntheclauseıpÒsaw d¢ pÒleiw prosagãgoito,theaoristoptativeillustrates

thatthepresentstemisnotcompulsoryforso-called‘distributive’actions.Theaoristisusedbecauseeveryconstituentintheclause,includingtheverbalconstituentwithhighinformationstatus,hasfocusfunction.Whileboththeunderlyingdiscoursetopic



48 CHAPTERTWO

Frug¤& §dÊnato strateÊesyai diå tØn Farnabãzou flppe¤an,

Xenophon returns to the narrative lineheabandonedin1.16: ı d¢ÉAghs¤laow ént‹ toË §p‹ Kar¤an fi°nai eÈyÁw éntistr°caw §p‹Frug¤aw §poreÊeto: ka‹ tãw te §n tª pore¤& épant≈saw dunãmeiw

énalambãnvn ∑ge ka‹ tåw pÒleiw katestr°feto ka‹ §mbalΔn éprosdokÆtvw pamplhy∞ xrÆmata ¶labe22 (‘but instead of marchingonCaria, Agesilaus forthwith turnedroundand made forPhrygia.Pickingupthevariousforcesthatmethimontheroute,heproceededtoreducethecitiesandcapturedavastquantityofbooty

bysuddenattacks’).Itis followedbyanevaluativesection,startingin1.17withstrathgikÚn oÔn ka‹ toËto §dÒkei diaprãjasyai, ˜ti §pe‹pÒlemow proerrÆyh ka‹ tÚ §japatçn ˜siÒn te ka‹ d¤kaion §j §ke¤nou§g°neto, pa›da ép°deije tÚn Tissaf°rnhn tª épãt˙, fron¤mvw d¢ka‹ toÁw f¤louw §ntaËya ¶doje plout¤sai(‘thisachievementalsowasthoughttobeaproofofsoundgeneralship,thatwhenwarwasdeclared and cozening in consequence became righteous and fairdealing, he showedTissaphernes to bea childat deception. Itwasthought,too, thathemadeshrewduseofthisoccasiontoenrichhisfriends’),whichinitsturnisfollowedbythebehavioralunits1.18(§pe‹ gãr )—22onhis filanyrvp¤a.

The§pe‹ m°ntoi-clausepresentsuswithaproblematicsituationforAgesilaus:acampaignin the plains was impossibleeven inPhrygia,owing to Pharnabazus’ cavalry. This contrasts sharply with theprosperoussituationresultingfromAgesilaus’spraiseworthybehaviorwearepresentedwithin§§18-22.Theparticle m°ntoisignalsthatthisproblematicsituationisunexpected,forbothAgesilaushimself(hence

his decision at that point to raise a mounted force) and thereader/hearer, with respect to the preceding episode. Theunexpected, problematic situation, then, is directly related toAgesilaus’sreactiontothissituationbymeansof§pe¤,whichindicates

———‘Agesilaus’spraiseworthyconducttowardsothersinwarfare’andthediscoursesub-topic‘notharmingenemies’arecontinued,atransitionismadefromhisbehaviortowardshumanadversariestohisbehaviortowardscities.Thefollowingsentence,connectedtothefirstbyka¤,elaboratesonthisbyreferringtofortresses.Before

abandoningthecurrentdiscoursetopic,inthefinalsentenceofthisepisode(ka‹ t«nkatå krãtow énal≈tvn teix°vn tª filanyrvp¤& ÍpÚ xe›ra §poie›to ), Xenophonexplicitly refers to what he had been demonstrating for several units: Agesilaus’s

filanyrvp¤a.22 Onthe functionalchoiceof theaoristherewhere theparallelpassagein the

Hellenicareads§lãmbane,seeBuijs,unpublished.




arelationbetweenthetwopropositionsintheRealWorld.Agesilaus’sdecision toraise a mounted force, however, is easily understood inviewofthissituation.

The first set of clause combinations to be compared aregign≈skvn d¢ ˜ti efi mØ flppikÚn flkanÚn ktÆsaito, oÈ dunÆsoitokatå tå ped¤a strateÊesyai, ¶gnv toËto kataskeuast°on e‰nai, …w

mØ drapeteÊonta poleme›n d°oiinthe Hellenicaand§pe‹ m°ntoi énåtå ped¤a oÈd¢ §n tª Frug¤& §dÊnato strateÊesyai diå tØnFarnabãzou flppe¤an, ¶dojen aÈt“ flppikÚn kataskeuast°on e‰nai,

…w mØ drapeteÊonta poleme›n d°oi aÈtÒninthe Agesilaus.Wehavetostartwiththeobservationthatthe§pe¤-clauseinthe Agesilausseemstopresent a more extensive version of the situation expressed by thegign≈skvn-clause,inthatitcontainsonepieceofinformationlacking inthe Hellenica,viz., diå tØn Farnabãzou flppe¤an.Itscounterpartinthe He lleni ca would be the adjective ‘flkanÒn’, which can be fullyappreciatedonlyinthelightofthe Agesilaus-passage,asPharnabazusisnotmentionedinthisepisodeinthe Hellenica.23Theheavyprocessing load in the Agesilaus may have lead the narrator to use a finite(§pe¤-)clause, not a participle. Further, the Real World situationspresented in the two passages are relayed from different points of view:ifthe two locutionsareanalysedon thesentencelevel,wecanobservethatinthe HellenicaXenophonpresentsuswithan internal considerationonthepartofAgesilaus(gign≈skvn ), which leads to aresolution(¶gnv ); that in the Agesilaus on the other hand Xenophonpresentsuswithanassertion( énå tå ped¤a oÈd¢ §n tª Frug¤& §dÊnatostrateÊesyai diå tØn Farnabãzou flppe¤an ),which ispresentedas

themotivationforthedecisionthatfollows(¶dojen aÈt“ ).While inbothpassagesAgesilausfunctionsas thefocalizer,theturngign≈skvn …¶gnv is an instance of internal focalization, whereas the turn §pe‹

m°ntoi … oÈd° … §dÊnato strateÊesyai … ¶dojen aÈt“ isindicativeofexternalfocalization,i.e.onthepartofthe laudator .Byvirtue of the imperfect §dÊnato and the non-specific relator §pe¤indicatingaRealWorldrelationobtainingbetweenthecontentofthesubordinateclauseandthecontentofitsmatrixclause,thespeakerpresentsarealitysituationthatbringsaparticipantroundtoperform

———23 The question which passage was written first is much disputed, and goes

beyondthescopeofthisstudy.



50 CHAPTERTWO

theactionpresentedinthematrixclause.24Whatismoreisthatthisisinlinewiththeaimsofthewriteroftheencomium:inthe Agesilaus,weare presented with a picture of Agesilaus, who in the light of hisevaluationoftheRealWorldsituationmakesacunningdecision.

Thenanarrativesequencedescribeshowthemountedforcewasraised.ThispartofthetextismarkedbyoÔninthe Agesilausinordertosetitofffromtheintroductorysentence§pe‹ m°ntoi… d°oi aÈtÒn,with which a return to the narrative sequence was achieved. Theparticlesignalsthatatthispointwehavebeenprovidedwithenough

introductoryinformationtobepreparedtopassovertothespeaker’smainpoint,i.e.presentingthethingsAgesilausdidtoachievehisgoalas an illustration of his qualities—which gives the writer of theencomium the opportunity to portray his hero as an outstanding general.

Comparingthetwopassagesthatdescribetheraisingofamountedforce, we may claim that the difference in linguistic devices usedwithin the N M P is to be explained by the speaker’s need toaccommodatethetwopassagesontheraisingofamountedforcetotheirrespectivecontexts.Thecomparisonofthesetwopassagesfromthe Hellenicaandthe Agesilausbringstolightthatthechoiceamong apparentlyalternativewaysofexpressingoneselfinAncientGreekisdependentuponthe communicativeaimsthespeakerhasinmind,whichare,intheirturn,dependentonthecontext ,andthe texttypetheyareusedin.Inthisrespect,achoiceoftheotheralternativewouldnothavebeenungrammatical,butwouldhaveresultedinadifferentkindofcommunication. ProeipΔn d°… §po¤hsen oÏtvandproe›pe d°

…ka‹ §po¤hsen oÏtvw in the above texts are cases in point: if aGreekspeakerwantstoconvey:‘heissuedaproclamation…’and‘inthiswayhebroughtitaboutthat…’hehasseveraloptionsathisdisposal(wecaneasilythinkofotheralternatives).Weareinclinedtotranslatetheparticipialclause-optionby‘byproclaiming’andthetwofiniteverbsbytwofiniteverbs,butthisisnotthepointhere.25

———24 Cf.§pe‹ d¢ m°gaw ı kÊklow Ãn poll∞w fulak∞w §dÒkei de›syai oÎpv pollo›w

oÔsi, sunespeirãyhsan §p‹ tØn Mounix¤aninNMP2.

25 Intheclausecombinationproeip≈n…§po¤hsen oÏtvtheparticipialclausewouldbeanexampleofwhathasbeencalled‘coincident’use(cf.Rijksbaron,20023:125:“theparticiple,whileexpressingacompletedstateofaffairs,isnotanteriortothemainverb,butcoincideswithit”).Theintroductionofthisextracategoryisnecessaryonlyifonebelievesthat“theaoriststemsignifiesthatastateofaffairsiscompleted”(Rijksbaron,20023:1)andthat“thevalue[completed]oftheaoriststemmayserveto




4 NMP4:Agesilaus1.38-2.2vs.Hellenica4.2.4-8&4.3.1-4

[7] Agesilaus1.38-2.2 [8] Hellenica4.2.4-8& 4.3.1-4

toigaroËn ofl §n tª ÉAs¤& ÜEllhnewoÈx …w êrxontow mÒnon éllå ka‹ …wpatrÚw ka‹ •ta¤rou épiÒntow aÈtoË§lupoËnto. ka‹ t°low §dÆlvsan ˜tioÈ plastØn tØn fil¤an pare¤xonto.§yeloÊsioi goËn aÈt“suneboÆyhsan tª Lakeda¤moni, ka‹taËta efidÒtew ˜ti oÈ xe¤rosin•aut«n deÆsoi mãxesyai. t«n m¢n dØ §n tª ÉAs¤& prãjevn toËto t°low§g°neto.

ékoÊsantew taËta pollo‹ m¢n§dãkrusan, pãntew d' §chf¤santo bohye›n met' ÉAghsilãou tªLakeda¤moni: efi d¢ kal«w téke›g°noito, labÒntew aÈtÚn pãlin¥kein efiw tØn ÉAs¤an. ka‹ ofl m¢n dØsuneskeuãzonto …w

ékolouyÆsontew. ı d' ÉAghs¤laow §n m¢n tª ÉAs¤& kat°lipen EÎjenon èrmostØn ka‹ frouroÁw par' aÈt“oÈk ¶latton tetrakisxil¤vn, ·na dÊnaito dias–zein tåw pÒleiw: aÈtÚw d¢ ır«n ˜ti ofl pollo‹ t«nstrativt«n m°nein §peyÊmoun mçllon μ §f' ÜEllhnawstrateÊesyai, boulÒmenow …w belt¤stouw ka‹ ple¤stouw êgein

mey' •autoË, îyla proÊyhke ta›wpÒlesin, ¥tiw êriston strãteumap°mpoi, ka‹ t«n misyofÒrvn to›wloxago›w, ˜stiw eÈoplÒtaton lÒxon¶xvn sustrateÊoito ka‹ ıplit«nka‹ tojot«n ka‹ peltast«n.proe›pe d¢ ka‹ to›w flppãrxoiw, ˜stiweÈippotãthn ka‹ eÈoplotãthntãjin par°xoito, …w ka‹ toÊtoiwnikhtÆrion d≈svn. tØn d¢ kr¤sin

¶fh poiÆsein, §pe‹ diaba¤hsan §kt∞w ÉAs¤aw efiw tØn EÈr≈phn, §nXerronÆsƒ, ˜pvw eÔ efide¤hsan ˜titoÁw strateuom°nouw de› eÈkrine›n.∑n d¢ tå îyla tå m¢n ple›sta ˜pla§kpeponhm°na efiw kÒsmon ka‹ıplitikå ka‹ flppikã: ∑san d¢ ka‹st°fanoi xruso›: tå d¢ pãnta îylaoÈk ¶latton §g°nonto μ épÚtettãrvn talãntvn. tosoÊtvn m°ntoi énalvy°ntvn, pampÒllvn xrhmãtvn ˜pla efiw tØn stratiånkateskeuãsyh.

———26 Íf¤ketoA:corr.Dm.2.



54 CHAPTERTWO

§poreÊeto diå t«n aÈt«n §yn«nœnper ı P°rshw t“ pamplhye›stÒlƒ: ka‹ ∂n §niaus¤an ıdÚn ı bãrbarow §poiÆsato, taÊthn me›onμ §n mhn‹ kayÆnusen ı ÉAghs¤laow.oÈ går …w ÍsterÆseie t∞w patr¤dowproeyume›to.

,26

Larisa›oi m¢n ka‹ Krann≈nioi ka‹Skotoussa›oi ka‹ FarsãlioisÊmmaxoi ˆntew Boivto›w, ka‹

pãntew d¢ Yettalo‹ plØn ˜soi aÈt«n fugãdew tÒte ˆntew

, krita‹kat°sthsan Lakedaimon¤vn m¢nM°naskow ka‹ ÑHripp¤daw ka‹ÖOrsippow, t«n d¢ summãxvn eÂw épÚ pÒlevw. ka‹ ÉAghs¤laow m°n,§pe‹ tØn kr¤sin §po¤hsen, ¶xvn tÚstrãteuma §poreÊeto tØn aÈtØnıdÚn ¥nper basileÁw ˜te §p‹ tØn Ñ Ellãda §strãteuen.(…)28

ı d' ÉAghs¤laow speÊdvn m¢n §k t∞wÉAs¤aw §boÆyei:ˆnti d' aÈt“ §n ÉAmfipÒlei égg°llei Derkul¤daw ˜ti nik“°n te aÔ LakedaimÒnioi, ka‹ aÈt«n m¢nteynãnai Ùkt≈, t«n d¢ polem¤vnpamplÆyeiw: §dÆlou d¢ ˜ti ka‹ t«nsummãxvn oÈk Ùl¤goi peptvkÒtewe‰en. §rom°nou d¢ toË ÉAghsilãou:âAr' ên, Œ Derkul¤da, §n kair“g°noito, efi afl sump°mpousai pÒleiw≤m›n toÁw strati≈taw tØn n¤khn …wtãxista pÊyointo; épekr¤nato dØ ıDerkul¤daw: EÈyumot°rouw goËnefikÚw taËt' ékoÊsantaw e‰nai. OÈkoËn sÊ, §pe‹ pareg°nou,kãllista ín épagge¤laiw; ı d¢ êsmenow ékoÊsaw, ka‹ går ée‹ filapÒdhmow ∑n, e‰pen: Efi sÁtãttoiw. ÉAllå tãttv, ¶fh, ka‹

prosapagg°llein ge keleÊv ˜ti §ånka‹ tãde eÔ g°nhtai, pãlinparesÒmeya, Àsper ka‹ ¶famen.ı m¢n dØ Derkul¤daw §f' Ñ EllhspÒntou pr«ton §poreÊeto:

.Larisa›oi m¢n oÔn ka‹ Krann≈nioika‹ Skotoussa›oi ka‹ Farsãlioi

sÊmmaxoi ˆntew Boivto›w, ka‹pãntew d¢ Yettalo¤, plØn ˜soi———

27 otoiA:exHellcorr.Richards, aÈt“H.Sauppe.28 4.2.9§n d¢ toÊtƒ…4.2.23ka‹ aÏth m¢n dØ ≤ mãxh oÏtvw §g°neto.




§tÊgxanon, §kakoÊrgoun aÈtÚn27

§fepÒmenoi. ı d¢ t°vw m¢n ∑gen §nplais¤ƒ tÚ strãteuma, …

ThereforetheGreeksinAsiamournedhisdepartureasthoughtheywerebiddingfarewellnotmerelytoaruler,buttoafatheroracomrade.Andattheendtheyshowedthattheiraffectionwasunfeigned.At

anyratetheywentwithhimvoluntarilytoaidSparta,knowingastheydidthattheymustmeetanenemynotinferiortothemselves.Thisthenwastheendofhisactivities

aÈt«n fugãdew tÒt' §tÊgxanon,§kakoÊrgoun aÈtÚn§pakolouyoËntew. ı d¢ t°vw m¢n∑gen §n plais¤ƒ tÚ strãteuma,…Uponhearingthismanyburstintotears,buttheyallvotedtogowithAgesilaustotheaidofLacedaemon,resolved,ifmattersthereshouldturnoutsuccessfully,tobringhimbackagainwiththemtoAsia.Andtheymadepreparationstofollowwith

him.AsforAgesilaus,heleftbehindhiminAsiaEuxenusasgovernor,andwithhimagarrisonofnotlessthanfourthousandmen,sothathecouldkeepthecitiessafe;andseeingthatmostofhisownsoldiersweremoredesirousofremainingthanof undertakingacampaignagainstGreeks,inthedesiretoleadwithhimthebestmenandasmanyaspossible

heofferedprizestothecities,fortheonewhichshouldsendthebestforce,andtothecaptainsofthemercenaries,fortheonewhoshould

jointheexpeditionwiththebestequippedcompanyofhoplites,of bowmen,andofpeltasts.Helikewiseannouncedtothecommandersof cavalrythathewouldalsogiveaprizeofvictorytotheonewhoshould

furnishthebestmountedandbestequippedbattalion.AndhesaidthathewouldmakethedecisionintheChersonese,upontheircrossingfromAsiatoEurope,hispurposebeingtoletthemunderstandthoroughlythattheymustselecttheirtroopswithcare.Asfortheprizes,mostofthemwerebeautifullywroughtarms,bothforhoplitesandforhorsemen;there

werealsowreathsofgold,andtheprizesalltoldcostnotlessthanfourtalents.Asaresult,however,oftheexpendingofthissum,armsworthavastsumofmoneywereprovidedfor




factalltheThessaliansexceptthoseof themwhochancedatthattimetobeexiles,followedathisheelsandmolestedhim.Foratimeheledthearmyinahollowsquare,…

factalltheThessaliansexceptthoseof themwhochancedatthattimetobeexiles,followedafterhimandmolestedhim.Foratimeheledthearmyinahollowsquare,…

Several pieces of information found in the diegetic passage fromXenophon’sencomium Agesilaus1.38-2.2arealsoencounteredinhis Hellenica,albeitinadifferentlyorganized(narrative)setting.InthisSection,specialattentionwillbepaidtotwosubordinateclausesinthe

Agesilauswhichcontaininformationthatisconveyedinthe Hellenicaaswell, though expressed by a different clause type. The pieces of information under consideration are: (1) “he crossed the Hellespont ”,expressedbyasentence-initialaoriststemconjunctparticipialclausein Agesilaus2.1( diabåw d¢ tÚn ÑEllÆsponton ) and by a sentence-initial§pe¤-clausewithanaorist stemverbalconstituent in Hellenica4.2.8(§pe‹ d¢ di°bh tÚn ÑEllÆsponton ),and(2)“havingpassedthrough Macedonia he arrived inThessaly”,expressedbyan aorist stemconjunctparticipial clause and a finite main verb in a sentence-initial§pe¤-clausein Agesilaus2.2(§pe‹ d¢ §jame¤caw Makedon¤an efiw Yettal¤an

éf¤keto )andbyanaoriststemconjunctparticipialclauseandafinitemain verb in an independent clause in Hel lenica 4.3.3 (ı d'ÉAghs¤laow diallãjaw Makedon¤an efiw Yettal¤an éf¤keto ).

Inthe Agesilaus, Xenophon’s narrative treatment of Agesilaus’s¶rga(seeSection3)canbedividedroughlyintotwoparts:anaccountofhisactivitiesinAsia(1.9 -1.38)andhisactivities inGreece(2.1-2.27)andEgypt(2.28-31).Example[7]presentsthefinalepisodeofa

largediscourseunitconcerningAgesilaus’sactivitiesinAsia.Boththebeginning and end ofthe episodeare marked explicitlyand overtly;the closing-line t«n m¢n dØ §n tª ÉAs¤& prãjevn toËto t°low§g°netomarkstheendofatextblockthatiscentraltotheconstitutionof the text as Xenophon planned it, as is often the case with suchclosing-lines;cf.also Hellenica4.2.23ka‹ aÏth m¢n dØ ≤ mãxh oÏtvw§g°neto(‘andinthismannerthatbattletookplace’).Notethehighdensity of linguistic marking in connection with the relatively lowinformation status of these sentences: anaphoric deictic elements

(toËto, aÏth,oÏtvw ), a semantically empty and therefore anaphoricverbalconstituent(§g°neto ); transition-marking m°n … d ° (for thisnotion,seeBakker,1993:302-305),andthelowinformationstatusof



58 CHAPTERTWO

the sentence as a whole because of the recapitulation of previouslyprocessedinformation,asmarkedby dÆ.29

ThetransitionfromAgesilaus’sactivitiesinAsiatohisactivities inEurope therefore constitutes a major break in the structure of theworkasawhole.Itistobenotedthatthistransitionconstitutesa Real Worldboundary;RealWorldboundariesincludetransitionfromoneday/season/yeartotheother,orthecrossingofanaturalbordersuchas a river or a strait. Real World boundaries may, but need notcoincidewithboundariesattheleveloftextarticulationorthelevelof

theRealWorldconstruction.30

Afteralarge-scalebreakhasbeenarticulated,thetransitionfromAsiatoEuropeitself( diabåw d¢ tÚn ÑEllÆsponton )canbeexpressedaspartofthenewepisodethatisintheprocessofbeingorganized.The topic is continuous, and the nature of the information that hecrossedtheHellespontisscript-predictableaftertheclosinglinet«n

m¢n dØ §n tª ÉAs¤& prãjevn toËto t°low §g°neto.The conjunct participialclauseseemstobethepropergrammaticaldevicetolaunchoff the new action sequence (‘having crossed the Hellespont hemarched on by the same route as the Persian king Xerxes hadfollowed when he made his expedition against Greece’). Thusexpressinghimself,Xenophoncreatesanopportunityforcomparing AgesilausandthePersianKing.ThedistancethathadbeentraversedbyXerxesinayearwascoveredbyAgesilausinlessthanamonth.Thecomparisonisinplaceintheencomiasticcontext;31XenophoncompareshisherotothePersianKingthroughouttheencomium(seeespecially Agesilaus9.1-5).

Example [8] presents the corresponding episode in the Hellenica.AgesilauspersuadeshisalliestogowithhimtotheaidofSparta,withtheintentionofreturningtoAsiatogetherafterwards.Hisalliesmakepreparationstofollowhim.AgesilaushimselftakescareofthesafetyofthecitiesinAsia.Heoffersprizesinordertoleadwithhimthebest

———29 Analmostprototypicalexampleofthiskindofmarkedclosureisprovidedby

Cyropaedia2.2.17:taËta m¢n dØ §ntaËya ¶lhjen. §k d¢ toÊtou…30 Bakker(1993:286)notesonthesubclause…w d¢ ≤m°rh §g°netoinHerodotus

2.121g:“theframeopenedbyatemporalsubclauseis“thematic”(content-oriented),in that the discourse boundary coincides with a natural caesura between twosequencesofeventsnarrated.”TheissueoflinguisticmarkingofthematicboundarieswillbediscussedinChapter4.

31 Asthegenreoftheencomiumdevelops,comparisonwillbecomeoneofitstopoi inlatertimes.




menandasmanyaspossible.Thedecisioninthecontestwillbetakenin the Chersonese, i.e. after having crossed from Asia to Europe.Xenophonspendsashortparagraphontheprizestobewon.Hethencontinues his narrative by mentioning Agesilaus’s crossing of theHellespont,andtheappointmentofthejudges.Thedecisionismade,and we learn that Agesilaus marched on by the same route asXerxes32. With that, we leave Agesilaus and are informed aboutcotemporalevents( Hellenica4.2.9§n d¢ toÊtƒ…4.2.23ka‹ aÏth m¢n

dØ ≤ mãxh oÏtvw §g°neto ).

Inthe Hellenica,thenaturalboundaryofthetransitionfromAsiatoEuropedoesnoteffectthatdeepanincisionwithintheorganizationof thetext ,asthecrossingtheHellespontispartofanon-goingstory-linethatwillbeabandonedsoonthereafter.Thesubclause§pe‹ d¢ di°bhtÚn ÑEllÆspontonisusedatapointinthetextwhere,comparedtotheratherstraightforwardwayofnarratinginthe Agesilaus,thereisarelatively highdegree ofturbulenceasfaras theorganizationof thetextisconcerned.Atthesametime,wearetakenbacktothemaineventlineafterashortdigressionontheprizestobewon(∑n d¢ tå

îyla … pampÒllvn xrhmãtvn ˜pla efiw tØn stratiånkateskeuãsyh ) and to express arelocation from the one side of theHellesponttotheother;inthisconnection,thesubclausemakesfortextual coherence by expressing the relocation as a link in anexpectancychain,i.e.,theoccurrenceofasituationanticipatedintheprecedingdiscourseistakenasthenewpointofdepartureforwhatfollows.In Hellenica4.2.6-8,wehavesuchanexpectancychaininthesequence tØn d¢ kr¤sin ¶fh poiÆsein, §pe‹ diaba¤hsan §k t∞w

ÉAs¤aw efiw tØn EÈr≈phn, §n XerronÆsƒ … §pe‹ d¢ di°bh tÚn Ñ EllÆsponton … krita‹ kat°sthsan … §pe‹ tØn kr¤sin §po¤hsen.WhilethisstrategicmoveonAgesilaus’spart(ır«n ˜ti ofl pollo‹ t«nstrativt«n m°nein §peyÊmoun mçllon μ §f' ÜEllhnawstrateÊesyai, boulÒmenow …w belt¤stouw ka‹ ple¤stouw êgein

mey' •autoË ) isabsent inthe Agesilaus,33Xenophonenlargesonitinthe Hellenica;onlytheredoesthecrossingofanaturalbordercoincidewithathematicbreakinthestory.Further,theidentityofthejudgesisrevealed,whichinvolvesatopicswich,whileAgesilausremainsan

———32 Notethattheextendedcomparisonisleftouthere.33 PerhapsXenophondeemedthereluctanceofAgesilaus’ssoldierstoundertake

thecampaignunfittingfortheencomium;noteincontrast §yeloÊsioi goËn aÈt“suneboÆyhsan tª Lakeda¤moniinthe Agesilaus.



60 CHAPTERTWO

activediscourseparticipant(notethethirdpersonsingular: di°bh

).Asopposedtotheconjunctparticipialclauseinthe Agesilaus,the§pe¤-clauseinthe HellenicahelpstokeeptrackofAgesilaus,whostillhasaroletoperformbeforeheisbroughtoffthestage(ka‹ ÉAges¤laow

m°n ),atapointwherenotonlytheprecedingunithasadifferenttopic(the îyla ),but also the followingmainclause featuresnewdiscourseparticipants (the krita¤ , who are introduced by name). Thecomplexityofthenarrativeinthe Hellenicaseemstohavetriggeredtheuseofasubclause.

Ishallnowdiscussthesecondinstanceofdifferentexpressionsusedtoconveythesamepieceofinformation,viz.,thatAgesilaus“having passedthroughMacedoniaarrivedinThessaly.”34

Inthe Agesilaus,theactionsequencethatisstartedwith diabåw d¢tÚn ÑEllÆsponton §poreÊetoisimmediatelythereaftertemporarilyabandonedinordertocompareAgesilausinafavourablemannerwiththePersianking(seeabove).InthecourseofthecomparisonbetweenAgesilausandXerxes,theprecisespatialsettinggoesastray,andsodoestheactionsequenceitselfwhenaparenthesis-likesentenceofreportedthoughtintervenes(oÈ går …w ÍsterÆseie t∞w patr¤dowproey u me ›t o ). Hereafter there seems to be need for some(re-)organizationofthediscourse:atthepointwheretheabandonedstory line is picked up, Agesilaus is continued as the dominantdiscourseparticipantandarelocationiseffected,asthenextepisodetakes place in Thessaly. Further, the enemies that Agesilausencounteredthereareintroduced.Allthisoccursatacaesurainthetext, for up to the point where the §pe¤-clauseoccurs,Agesilaus’s

retreat from Asia has been unproblematic, but once Thessaly isreached,hegetsintoserioustrouble.Thepreposed§pe¤-clauseisusedto move Agesilaus to Thessaly, so that the enemies can beconvenientlyintroducedinthemainclause.Onceintroduced,theyceasetobedominantdiscourseparticipants,asmarkedby m°n,forinthe d°-memberthecameraimmediatelyturnsback,sotospeak,totheherooftheencomium.Again,thesubclauseisusedatapointinthetextwherethereisarelativelyhighdegreeofturbulenceasfarasthe

———34 Idonotconsider the alternation of§jame¤cawand diallãjawofanymajorimportanceinthisconnection;throughouttheparallelpassagesofthe Agesilausandthe Hellenicasuchalternationsarefound.Forinstance,whereverXenophonusestheverb ékolouy°vor§pakolouy°vinthe Hellenica,heusesßpomaior§f°pomaiinthe Agesilaus,asinthepassagescited.




organization of participants, location, and action sequence isconcerned.

Inthe Hellenica, Agesilaus is re-introduced as a topic into thediscourseafteraconsiderableperiodofabsencein4.3.1(notethefullNP ı d' ÉAghs¤laowandtheparticipialclausespeÊdvn … §k t∞wÉAs¤aw, which helps to anchor him in the addressee’s previouslystoredknowledge).Dercylidas(re-)appears,reportingtohimthattheSpartans were victorious. A conversation between Agesilaus andDercylidas takes place, in which Agesilaus dismisses Dercylidas to

reportthevictorytothecitiesthataresendingtheirtroops.Attheendof this conversation, Dercylidas sets out for the Hellespont, andAgesilaus passes through Macedonia and arrives in Thessaly. The Hellenicafeaturestwoindependent(main)clausesco-ordinatedby m°n… d°toexpressthis.Thesesentencescloseoffthedirectspeechunit;in the m°n-memberscript-predictableinformationispresented(note

dÆ ),andDercylidas isbrought off the stage; the d°-membertakesupAgesilaus,whoistobecomethetopicofthenewsequence.Agesilaus’sarrivalinThessalyisformallysetapartfromthesubsequentepisodebythepresenceoftheparticleoÔn,givingtheLarisa›oi m°n (oÔn) …ı d°-sequenceadifferentstatusthantheprecedingbyindicatingthatthepointatwhichthenarrativeaimedhasbeenreached.TheparticleoÔn is conspicuously absent in the Ag esilaus; its presence in the Hellenica,whereithelpstoarticulatedifferentepisodesofthetext,underscorestheobservationsthatthesetextsaredifferentlystructured.

Thecomparisonofthetwotextsinthis NMPrevealsthatifthesame Real World situation is to be narrated in two texts that are

differently constituted, the choice among the different clause typesavailableinAncientGreekisprincipallydeterminedbythespecificcontextinwhichagivenpieceofinformationispresented.Facedwiththetaskofprocessinginformationpertainingtodifferentaspectsofthecommunication, such as participant-tracking, discontinuity in thespatialsetting,ormainvs.subsidiaryactionsequences,thespeaker,itisassumed,willarticulatehistextaccordingtohowhethinksthesedifferent tasks will be achieved best. Given the fact that thecommunicativeaimsthespeakerhadinmindaredifferentforthe

Agesilausandthe Hellenica,itis,initself,notsurprisingtofinddifferentclausetypesusedtoexpressthesameRealWorldsituationintextsthat,as awhole, are differentlyconstituted.Whatshouldinterest ushereisthewayinwhichtheseseveralclausetypescontributetothe



64 CHAPTERTWO

catchwhomeveryouchoose,andretreatingyoushallnotfeartheenemy”.

herethanyouwerewhenfirstyou joinedus?”

Theseexampleshavebeentakenfromlargerdiscoursesdescribingtheeventsattwosymposia.Inbothtexts,athematicunitisroundedoff bymeansofaclosing-line: §ntaËya m¢n dØ g°lvw §g°netoand§p‹toÊtoiw m¢n dØ g°lvw §g°netointhe AnabasisandtheCyropaedia,respectively.Again,certaincharacteristicelementsarefoundinthese‘closinglines’:anaphoricdeicticelements(§ntaËya/§p‹ toÊtoiw ), the

aorist indicative §g°neto with an inanimate subject, transition-marking m°n … d ° , and an informational content that thereader/hearer will readily accept in the light of what precedes, assignaledby dÆ.Thecontinuationafterthissentenceistosomeextentidentical,inthatonlyaspecificsceneduringthebanquetisroundedoff, while the description of the banquet itself continues; thecontinuation of the banquet, however, is expressed with differentlinguistic means: a subclause is used in the Anabasis (§peidØ d¢

proux≈rei ı pÒtow ),whereastheCyropaediacontinueswithagenitive

absolute (proÛÒntow d¢ toË sumpos¤ou ). If two contexts differ sominimallyquanarratedsequence,itseemspromisingtoinvestigatethenarrativestructureofthelargerdiscourseepisodeinwhichtheyoccur,as well as the relation between the subordinated propositions andtheirmatrixclauses.

Thereportofthebanquetin Anabasis7.3.15-33startswithSeuthesinvitingthegeneralsandthecaptainstodinner(§15).Thesequelisbasicallytripartite:thefirstepisodetakesplacebeforetheguestsenter;

a certain Heracleides solicits rich gifts for the king (§§16-20).Xenophonappearstohavea problem,forhehadcomeacrossfromParium with nothing but a boy and enough money to cover histravelling expenses. Then we have the dinner itself (§§21-25). Thisepisodeconsistsofadescriptionofthediningroom(§21),theoddwayinwhichthefoodisdistributed(§22),andthepeculiarbehaviorofoneArystas,andSeuthes’s reactionto it, closingoff with the outburstof laughter (§§23-25). The third episode (§§26-33) describes the afterdinner happenings: toasting the king and presenting him with gifts(§§26-32)—asequenceinwhichthreeminorparticipantsappearonthe stage, who are followed by Xenophon himself, presenting asolutionto his problem:heoffers himselfand his friendstoSeuthes.Theepisodeendswithanentertainment-scene(§§32-33).




Figure4:TheBanquetinCyropaedia8.4.1-27

invitation KËrow … §kãlese …sunekãlese

participantsintroduced

arrivalatbanquet

…w d' ∑lyon oflklhy°ntew §p‹ tÚ

de›pnon

systemoftheseating plan

dinner §pe‹ d¢ §de¤pnoun Cyrus’sgenerosityendofdinner §pe‹ §dedeipnÆkesan ka¤

… filanyrvp¤a

afterdinnerdrinks

§k toÊtou dØ §pe‹Íp°pinon

howtoshowloyaltybestbetrothalofG.’sdaughter

metå d¢ taËta Cyrusdistributes

giftsleave-taking ka‹ tÒte m¢n dØ oÏtvw¶lhjen ≤ skhnÆ:§janistam°nvn d'

aÈt«n §jan°sth ka‹ ıKËrow ka‹jumproÊpemcen aÈtoÁw§p‹ tåw yÊraw

A comparison of the two figures offers a clue to the answer whyXenophon preferred a subclause in the Anabasis and a genitiveabsoluteintheCyropaediainordertoexpresscomparableinformation:inthe Anabasis,thesubclauseoccursattheonsetofoneofthepartsof theover-allscheme,creatingarelativelydeepincisiononthelevelof theRealWorldconstruction:thereisatransitionfromthedinner-episodetothepresentationofgiftsthatwasanticipatedtoin§§16-20.As it happens, the theme-predictable episode of the ‘after dinnerdrinks’ had been more or less initiated during the Arystas-section:horns of wine had been carried round (§24: k°rata d¢ o‡nouperi°feron ) and everybody apparently had finished their meals foreveryonetookahorn(§24:ka‹ pãntew §d°xonto ),exceptingArystas,



68 CHAPTERTWO

whohadnotfinishedasyet(§24:sxolãzei går ≥dh, §gΔ d¢ oÈd°pv

).The subclause is therefore used for content-oriented discoursesegmentation; atthe same time,it makesfor thematiccoherence byreferringtotheafterdinnerdrinks(ı pÒtow ) that had already beenintroducedin§§24-25(hence,probably,§peidÆ ).

IntheCyropaediawehaveproÛÒntow d¢ toË sumpos¤ou,thenounmostprobablyreferringtothewholeeventratherthantooneofitsindividual parts. In fact, the genitive absolute is used at a minorboundary between two individual conversations; the three sections

consistingof§§9-12,§§13-23,and§§24-27allbelongtothehigh-levelthematic unit of the after dinner drinks, as indicated in §9 by §ktoÊtou dØ §pe‹ Íp°pinon.39 There is articulation of small-scalediscoursesegmentation(note d° ), indicatingtheonsetofanextphaseinthe text.Thegenitiveabsolutemoves narrative timeforwardandalthoughanewthemewillbeaddressedinthisphase,thisphaseisstillto be viewed as part of the text segment concerned with the afterdinner drinks. By its content, the genitive absolute provides thereader/hearer with the information that the underlying discoursetopicofthesymposiumdescribedisstillrelevant;assuch,thegenitiveabsoluteismainlytheretohelpthereader/hearerkeeptrackofthediscourseperspective.

Thompson & Longacre(1985:206)statethat“adverbialclausesmaybeusedtoprovidecohesionforanentirediscoursebyassistingtomaintainthediscourseperspectiveandbyhelpingtoarticulatethesectionsofdiscourse”.Theusageofsubclausesinbothtextsisinlinewiththisobservation,astheperspectiveofthethematicstructureona

relativelyhighlevelisprimarilymaintainedviasubclauses( Anabasis§16§pe‹ d' §p‹ yÊraiw ∑san …w §p‹ de›pnon pariÒntew;§21§pe‹ d¢efis∞lyon §p‹ tÚ de›pnon;§26§peidØ d¢ proux≈rei ı pÒtow;§34…w

d' ∑n ¥liow §p‹ dusma›w;§35…w d' §jªsan;Cyropaedia§3…w d' ∑lyonofl klhy°ntew §p‹ tÚ de›pnon;§6 §pe‹ d¢ §de¤pnoun;§7 §pe ‹§dedeipnÆkesan ka¤ …;40§9 §k toÊtou dØ §pe‹ Íp°pinon ).41Other

———39 See Gera (1993: 150, n. 62) for some discussionon the precise meaning of

Íp°pinon.

40 Due(1989:102)statesthat“thedescriptionof thedinner itselffallsinto fourparts,inwhichCyrusdiscussesdifferentsubjectswithdifferentguests.”Hedividesthepassageasfollows:partI:§§6-8;partII:§§9-12;partIII:§§13-23(onpage103,n.49,§24ismentioned,butjudgingfromthebodytext§23mustbemeant);partIV:§§24-27).Gera(1993:133)correctlynotesthat“theconversationatthissymposium—aseriesoflooselyconnectedexchangesbetweenCyrusandhisguests—beginsonlyafter




linguistic devices at the onset of sentences, such as the genitiveabsolute proÛÒntow (d¢) toË sumpos¤ou(Cyropaedia §13) and thetemporaladverbial metå (d¢) taËta( Anabasis§33,Cyropaedia§24)dohaveatextarticulatingfunction,albeitonadifferentstructurallevel;theyhelptoarticulateboundarieswithinathematicunit,theoverallthemeofwhichhadalreadybeendefinedbyasubclause.42Onastilllowerstructurallevel,viz.,aboundarybetweentwosentenceswithinasub-segment, we may have for instance a participial clause like

ékoÊsaw( Anabasis §25; note the ‘asyndeton’), a case of ‘reciprocal

coupling’(Thompson& Longacre,1985:213),preparingforatopicswitch.

At this point, it may be instructive tocompare the continuationafterasentenceclosingoffadiscourseunitin[11]and[12];[11]istakenfromthesamebanquet-sceneas[10]:

[11] Cyropaedia8.4.21-23¶peita d', ¶fh, simØ ên soi fisxur«w sumf°roi. PrÚw t¤ dØ aÔ toËto;ÜOti, ¶fh, sÁ grupÚw e‰: prÚw oÔn tØn simÒthta sãf' ‡syi ˜ti ≤grupÒthw êrist' ín prosarmÒseie. L°geiw sÊ, ¶fh, …w ka‹ t“ eÔ dedeipnhkÒti Àsper ka‹ §gΔ nËn êdeipnow ín sunarmÒttoi. Na‹ måD¤', ¶fh ı KËrow: t«n m¢n går mest«n grupØ ≤ gastØr g¤gnetai,

———dinner.”Iwouldarguethatalthough§6and§7aretightlyconnected(Àste;cf.theimperfect §dÒkeiin§6),aboundaryiscreatedintheorganizationofthediscourseaccordingto theindividualpartsofbanquetbythesubclause§pe‹ §dedeipnÆkesanka‹…

41 Ofcourse,subclausesneednotbeusedsolelyfordiscourseorganization;inı d'ÉArÊstaw, §pe‹ par' aÈtÚn f°rvn tÚ k°raw ı ofinoxÒow ∏ken, e‰pen fidΔn tÚnJenof«nta oÈk°ti deipnoËnta, ÉEke¤nƒ, ¶fh, dÒw: sxolãzei går ≥dh, §gΔ d¢

oÈd°pv( Anabasis§24),wherethetopicisexpressedbymeansofafullNPattheinitialposition of the sentence before the subclause, the subclause is primarily there toindicatearelationinthenon-linguisticworldbetweenthesubordinatedandthemainevent,totheeffectthatthesubordinatedeventisrelayedfromtheparticipant’spointofview.

42 Cf.Herodotus6.129:…w d¢ ≤ kur¤h §g°neto t«n ≤mer°vn t∞w te katakl¤siowtoË gãmou ka‹ §kfãsiow aÈtoË Kleisy°neow tÚn kr¤noi §k pãntvn, yÊsaw boËw•katÚn ı Kleisy°nhw eÈ≈xee aÈtoÊw te toÁw mnhst∞raw ka‹ Sikuvn¤ouw pãntaw.

, ofl mnhst∞rew ¶rin e‰xon émf¤ te mousikª ka‹ t“legom°nƒ §w tÚ m°son. p kat°xvn pollÚn toÁw êllouwı ÑIppokle¤dhw §k°leus° ofl tÚn aÈlhtØn aÈl∞sai §mmele¤hn(‘whentheappointedday came for the marriagefeastand for Cleisthenes’ declaration ofwhomhe had

chosenoutofthemall,CleisthenessacrificedahundredoxenandgaveafeasttothesuitorsandtothewholeofSicyon.A ,thesuitorsviedwitheachotherinmusicandinanecdotesforalltohear.T ,whenHippocleides,nowfaroutdoingtherest,orderedtheflute-playertoplayhimadance-tune’).Here,too,theonsetofathematicunitismarkedbyasubclause,whileagenitiveabsolutehelpstoarticulateaboundarybetweensub-segments.




ka‹ ı KËrow prÚw taËtae‰pe:

[speechbyCyrus]

,¶doj° te Ùry«w to›wparoËsi l°gein ka‹∑gen ı GvbrÊaw Àsper§k°leuse.

toÁw d° ge aÎjontaw efi mØ §pista¤mhn éntitimçn, afisxuno¤mhn ên. ka‹ ıLÊsandrow e‰pen:ÉAll' ‡svw ka‹ mçllonefikÒta sÁ poie›w μ §gΔ¶pratton. tãde oÔn moi§k toË loipoË xãrisai,˜pvw ín mÆt' afisxÊnvmai édunat«nparå so‹ mÆt' §mpod≈nsoi Œ, épÒpemcÒn po¤ me. ˜pou går ín Œ,peirãsomai §n kair“soi e‰nai.

¶doje ka‹ t“ÉAghsilãƒ oÏtvpoi∞sai, ka‹ p°mpei aÈtÚn §f' ÑEllhs-pÒntou.

Beingdistressedathis

ßsthke suntetagm°non;o„ μn §p' §ke¤nouw ≤me›w§laÊnvmen,ÍpotemoËntai ≤mçwpãlin [§ke›noi]: ≤m›n d¢oÎpv ≤ fisxÁwpãrestin. ÉAll' μn sÁ m°n˙w, ¶fh ı KËrow,ka‹ énalambãn˙w toÁwprosbohyoËntaw, fobÆsontai otoi ka‹oÈ kinÆsontai, ofl d' êgontew eÈyÁw éfÆsousi tØn le¤an,§peidån ‡dvs¤ tinaw §p' aÈtoÁw §laÊnontaw.

46 ¶doj° til°gein t“ ÉAstuãgei.ka‹ ëma yaumãzvn …wka‹ §frÒnei ka‹§grhgÒrei keleÊei tÚnuflÚn labÒnta tãjinflpp°vn §laÊnein §p‹toÁw êgontaw tØnle¤an.AsforCyrus,uponseeingmanyhorsemenoveragainstthem,heasked:“Say,grandfather”,saidhe,

“arethosemenenemieswhosittherequietlyupontheirhorses?”“Yes,indeed,theyare”,saidhe.“Arethoseenemies,too”,saidCyrus,“whoareriding upanddown?”“Yes,

———

46 efipÒntow aÈtoË:e fipΔn y(=D& F) R. As the conjunct participle is,syntacticallyspeaking,thelectiofacilior , I take the genitive absolute as the correctreading,followingthemostrecenteditionsbyMarchantandGemoll.Someoldereditions, e.g., those ofLeunclavius,Hutchinson, andSchneider,read taËt' efip≈n.Cf.Schneideradloc.(myitalics):“ExGuelf.efipÒntow aÈtoËdifficilioremmultolectionemrecepitcumZeunioWeiske.”




constituentse‰pen

/efip≈n

ofthepreposedembeddedclausesandthedependendinfinitive l°gein ) share the same subject, the finite mainverb¶dojeispersonallyused,whichleadstotheinterpretationofthemainclauseasawholeasexpressingan opiniononthepartoftheparticipantreferredtointhedativecase.

The caseof the genitiveabsolute in[14] is different. There, theverbalconstituentefipÒntowofthepreposedembeddedclauseandtheinfinitivepoi∞saihavedifferentsubjects.InthecaseofthesequenceefipÒntow d¢ taËta ¶doje ka‹ t“ ÉAghsilãƒ oÏtv poi∞sai , a

personalinterpretationof¶dojeisimpossible;furthertheka¤in¶dojeka‹ t“ ÉAghsilãƒ oÏtv poi∞saitriggersaninterpretationofthisclauseasadecisiononthepartofAgesilausratherthananopinion.48

Example[15],then,isunusual.Agenitiveabsoluteisusedwithitsown subject expressed ( aÈtoË ), which is also the subject of theinfinitivel°gein.Thesentencemaybeinterpretedaseither:

taËt' efipÒntow aÈtoË ¶doj° ti (sc. Cyrus) l°gein t“ÉAstuãgei49

oras:

taËt' efipÒntow aÈtoË ¶doj° ti l°gein (sc. aÈtÒn )t “ÉAstuãgei.

In the former case the main verb ¶doje is given its personalinterpretation (cf. [16]), in the latter ¶dojeisgivenitsimpersonalinterpretation,asin[14].ButaswearedealingwithanopiniononthepartofAstyagesin[15],itseemsmostnatural,andinaccordancewiththefindingsof[13]and[16]ontheonehand,andof[14]ontheother,tooptforthepersonalinterpretationof¶doje.50

Thefactthatthegenitiveabsolutehasitssubjectexpressed( aÈtoË ),whichisalsothesubjectof ¶dojeandl°gein makes this exampleratherodd,ifnotungrammatical.Syntacticallyspeakingitisaberrant,butIsubmitthatitissoundintermsofdiscoursepragmatics.Intheremainder of this SectionI shallclaim that the genitiveabsolute is

———48 Thisinterpretationissubstantiatedbythefactthatofthesubordinateclausesin

[13]through[17],onlyinefipÒntow d¢ taËtain[14]theanaphoricpronounoccursclause-final.

49 ThusKühner-Gerth(1898-1904:110).50 The order of the constituents in the matrix clause in [15] (¶doj° — t il°gein — t“ ÉAstuãgei ) is different from the one found in [16] (¶doje — t “Kleãrxƒ — élhy∞ l°gein ).The order of constituents canbe explained byassuming themostfocalconstituenttotakethefinalposition:‘heseemedtoClearchusthoughttobes ’vs.‘heseemedtohaveapointt ’.




¶dojen aÈt“ ≤ ıdÚw ∂n æei par' aÈtÚ tÚ te›xow f°rein(‘andwhen,as

heproceeded,hecameinsightofthecityofBabylonanditseemedtohimthattheroadwhichhewasfollowingledclosebythewalls’),thisoneinitializingadiscussionbyCyrusontheonehand,andGobryas(andGadatas)ontheother,onhowtoproceed.Ourexample §pe‹ d¢taËt' e‰pen, then, ends the discussion, and they leave the spotsuccesfully( épex≈rei ),whilein5.4.51§pe‹ d¢ poreuÒmenow oÏtvw §nta›w gignom°naiw ≤m°raiw éfikne›tai efiw tå meyÒria t«n SÊrvn ka‹MÆdvn, ¶nyenper …rmçto, §ntaËya dÆ…(‘andwhen,ashethus

proceeded,hecamein theusualnumberofdaystotheplaceontheboundariesbetweenMediaandSyriafromwhichhehadoriginallystarted;there…’)initializesanewstageatadifferentlocation.Toallappearances,then,thesubclausesperformafunctioninthethematicorganizationofthenarrative.

Secondly,itmayverywellbethecasethatasubclauseisusedin[13] to indicate that the content of the subclause, i.e., what thereporteesaid,istakenbythenarratorasthepointofdepartureforboth the reaction of the reportee’s addressee and the fact thatGobryas tookthelead, asthisis somethingCyrus’swordsaimed at.Thecontributionof§pe¤,then,ismarkingthatbothactionsshouldbeviewedinthelightofthewordsspoken.

Thirdly,andinconnectionwiththesecondpoint,weobserveadifferenceinscoping :ineachofthethreeexamplesthesubordinateclauseisfollowedbyaclausecombination,combinedonthesamesyntactic-hierarchicallevelbyka¤.Onlyinthecaseofthesubclauseist e … ka¤ used, indicating that the two combined (independent)

clausesareunderthescopeofthesubclause.In[14],[15],and[16],onlythemainverbofthefollowingclause(¶doje ) isunder thescopeoftheparticipial clause.Inthesecases,theuseofan§pe¤-clausewouldresultinarticulatingtoodeepanincisionin the text, or, alternatively, in articulating a major discourseboundary which is inexpedient in the context. Rather, the use of participialclausesintheseexamples,andtheindividualdifferencesamongthem,havetobeaccountedforintermsoftopicmanagement.

Thesituationisratherstraightforwardin[16]:wedonothavea

conversation, but an alternation of speeches. The conjunct participletaËta efip≈njustindicatesthattheonespeakerhasfinished;via¶dojet“ Kleãrxƒ (élhy∞ l°gein)thecamerashiftstoClearchus,whobecomes the next speaker. As the (conjunct) participial clause



CHAPTERTHREE

PRESENTATIONOFREALWORLDRELATIONS

Introduction

ThespeakerwhohasselectedtwoRealWorldeventsforrecordingina text and wishes to present them in two clauses, may decide toexpressoneoftheseclausesasgrammaticallysubordinate.Hethenhas to decide whether he wants the subordinate clause to be asubclauseoraparticipialclause.Inthecaseofasubclause,thechoiceofasubordinatingconjunctioniscompulsory,whereasinthecaseofaparticipial clause, he is free to use a relator or not. In Chapter 1,Section2.2.1,itwasclaimedthatifaRealWorldrelationbetweenthecontent of the clauses of which the clause combination consists isconceivableandthespeakerwishestoindicatethisrelation,heputsina subordinating conjunction, and that he may opt for relators thatrefer to (aspects) of states of affairs in reality (semantically specificrelators)orrelatorsthatoperatieonthediscourselevel(semanticallynon-specificrelators) according tohisaim.Asa referentialaspectof the communication between the speaker and his audience, thespeaker’s motivation for choosing one of the available relatorsdeservescarefulconsideration.Thespeaker’smotivationforchoosing

oneofthesubordinatingconjunctions,itwillappear,isboundupwiththekindofrelationthespeakerexperiencesintheRealWorldthatheisdescribing,aswellaswiththewayhewishestopresentthisrelationinacertaincontext.

1 SubclausesHeadedbyaSemanticallySpecificRelator

1.1 ‘Causal’Relations

Whenthespeakerusesthesubordinatingconjunction diÒtior˜tiatthe head of one ofthe (finite)clausesofa clause combination, thuscreatinga‘subclause’,therelationintheRealWorldbetweenthe



88 CHAPTERTHREE

Xenophon’s own addition. Xenophon also describes why theAthenians acted as they did: they bought their provisions at adistance, and they thought lightly of Lysander (the postposedparticipial clauses tã te sit¤a pÒrrvyen »noÊmenoi ka‹katafronoËntew dØ toË Lusãndrou ). That the Athenians thoughtlightlyofLysanderismarkedby dÆassomethingthereader/hearerwill understand without difficulty: the reasonfortheircontempt,of course, was that Lysander did not put out to meet them (˜ti oÈk

éntan∞gen ).

Thepassageiscomplicatedasfarastheattributionofopinionsisconcerned:withtheclause˜ti oÈk éntan∞genXenophongivesthereason the Athenians had for thinking lightly of Lysander, asexpressedintheparticiplekatafronoËntew,whichitselfismeanttoclarifyXenophon’sownstatementthattheAtheniansdidfarmorefreelyeverydaywhatLysandertoldhismentheywouldbedoing.Insuchacontext,thenarrator,byusingthesemanticallyspecificrelator˜tiforexpressingthe‘causal’relationbetweenLysander’snotputting out to meet them and the contempt of the Athenians, secures afelicitous communication with his reader/hearer; the narrator haseveryreasontobespecificabouttheRealWorldrelationhesignals,especially since the one Real World situation (‘Lysander o È k

éntan∞gen’)ismeanttobeviewedasthereasonnotonlyfortheothers (katafronoËntew ) but also for his own evaluation of theattitudeoftheAtheniansasonethereader/hearerwillunderstandwithoutdifficulty( dÆ ).


§pe‹ d¢ …si≈yhsan afl ≤m°rai, ka‹ ¶dei basil°a kay¤stasyai, ént°legon per‹ basile¤aw Levtux¤dhw, uflÚw fãskvn ÖAgidow e‰nai,ÉAghs¤laow d¢ édelfÒw. efipÒntow d¢ toË Levtux¤dou:Le: ÉAll' ı nÒmow, Œ ÉAghs¤lae, oÈk édelfÚn éll' uflÚn basil°vw

basileÊein keleÊei: efi d¢ uflÚw Ãn mØ tugxãnoi, ı édelfÒw ka Õw basileÊoi.

Ag: ÉEm¢ ín d°oi basileÊein.Le: P«w, §moË ge ˆntow;Ag: .Le: ÉAll' ≤ polÁ kãllion §ke¤nou efidu›a mÆthr ka‹ nËn ¶ti fhs¤n.

Ag: ÉAllå ı Poteidån …w mãla seu ceudom°nv katemÆnusen §k toË yalãmou §jelãsaw seism“ efiw tÚ fanerÚn tÚn sÚn pat°ra.sunemartÊrhse d¢ taËt' aÈt“ ka‹ ı élhy°statow legÒmenow xrÒnow e‰nai: ~éf' o gãr toi ¶fus° [se] ka‹ §fãnh §n t“yalãmƒ, dekãtƒ mhn‹ §g°nou~.



PRESENTATIONOFREALWORLDRELATIONS 89

ofl m¢n toiaËt' ¶legon.

Theprescribeddaysofmourninghadbeenreligiouslyobservedanditwasnecessarytoappointaking;Leotychides,whoclaimedtobea sonofAgis, and Agesilaus, abrotherofAgis, contended for the kingship.Leotychidessaid:“But, Agesilaus, the law directs,not thatabrother,but that a son of a king, should be king; if, however, there shouldchancetobenoson,inthatcasethebrotherwouldbeking”.“It isI,then,whoshouldbeking”.“Howso?—Iamalive”.“B

”.“Nay,butmymother,whoknowsfarbetterthanhedid,sayseventothisdaythatI

am”.“ButPoseidonshowedthatyouareentirelyinthewrong,forhedroveyourfatheroutofherchamberintotheopenbyanearthquake.Andtimealso,whichissaidtobethetruestwitness,gavetestimonythatthegodwasright;foryouwereborninthetenthmonthfromthetimewhenhefledfromthechamber”.Suchwerethewordswhichpassedbetweenthesetwo.

In this passage of direct speech, we have another example of asemantically specific ( ‘causal’) relator. When Agesilaus andLeotychides contended for the kingship of Sparta, Agesilaus is

reported to have claimed that he should be king (ÉEm¢ ín d°oi basileÊein ).Sinceaking’ssonshould succeed tothe throne,nothisbrother,Leotychidasreplies:“Howso?<Givemeanexplanationof

your statement, for> I amalive”(P«w, §moË ge ˆntow; ). Agesilausreplies:“<Iclaimthis>becausehewhomyoucallyourfathersaidthatyouwerenothisson”(ÜOti ˘n tÁ kale›w pat°ra, oÈk ¶fh see‰nai •autoË ). The semantically specific ‘causal’ relator˜tiindicatesthatthecontentoftheclauseservesasanexplanationofthecontent

of the clause expressing Agesilaus’s claim: it expresses the reasonAgesilaushadformakingtheclaim.Tothebestofmyknowledgeaquestion introduced by p«w cannot be answered by a statementintroducedbyasemanticallynon-specificrelator,sinceitisnothingbutareasonthatisaskedfor.

1.2 TemporalRelations

Whenthespeakerusesthesubordinatingconjunction ˜te,≤n¤ka,or

§n ⁄attheheadofoneofthe(finite)clausesofaclausecombination,thuscreatinga‘subclause’,therelationintheRealWorldbetweenthecontentofthetwoclausesofwhichtheclausecombinationconsists,islikely to have been experienced by the speaker, but certainly



94 CHAPTERTHREE

passageof Hellenica3.4.16,wherethesubclauseisheadedby§peidÆ

.5

Thisdifferencebetween‘autumn’and‘spring’isexplainedeasilyfromthefactthatofthesetwoseasons,‘autumn’markstheendofayear’speriod of military warfare, whereas ‘spring’ marks its opposite, thebeginning, which entails that the beginning of spring, apart frommarkinganewphaseinRealWorldtime,inadditionmarkstheonsetof a new episode onthediscourselevel :thestartingpointofthenextmilitary season as well as of a new sequence of narrative events.Indeed,theverbalactionsofthematrixclausescombinedwiththe

subclausesunderconsiderationunderscorethis:whereasthematrixclause in [6] expresses Agesilaus’s expected retreat ( épπei §k t∞w

x≈raw ), the matrix clauses of the §pe¤/§peidÆ ¶ar-clauses6allshowverbal actions of attack.7 The use of a semantically non-specificrelator,then,coincideswithathematicbreak:thebeginningofspring providesnotonlyalocalframeofreferencefortheupcomingmatrixclause, but also amore globalone foranewactionsequence tobeexpressedintheupcomingtextsegment,sothatthesemanticallynon-specificrelatorsignalstheimportanceofthecontentofthesubclauseforthediscoursestructurethatisabuilding.

Theco-occurrenceof¶arandsemanticallynon-specificrelatorsinorder to notify of the beginning of spring is mirrored in thenotificationofthebeginningofday:thebreakofdayisreportedninetimesinasubclauseinthesethreeworksofXenophon,eighttimeswith ≤m°ra8fillingthesubjectslot,oncewith ßvw9;thesubclauseis,

———5 Hellenica3.4.16:§k d¢ toÊtou §peidØ ¶ar Íp°faine(‘afterthis,atthefirstsignof

spring’≈ Agesilaus1.25:§peidØ d¢ ¶ar Íp°faine:‘atthefirstsignofspring’); Hellenica5.4.47: §pe‹ d¢ tÚ ¶ar §p°sth(‘whenthespringcame’); Hellenica5.4.59:§pe‹ ¶arÍp°faine(‘whenspringwasjustbeginning’).

6 Hellenica3.4.16:sunÆgage m¢n ëpan tÚ strãteuma efiw ÖEfeson(‘hegatheredhis whole army at Ephesus’ ≈ Ages ilaus1.25:sunÆgage pçn tÚ strãteuma efiwÖEfeson:‘hegatheredhiswholearmyatEphesus’); Hellenica5.4.47:pãlin ¶fainon

frourån ofl ¶foroi efiw tåw YÆbaw, ka‹ toË ÉAghsilãou, √per tÚ prÒsyen, §d°onto≤ge›syai(‘theephorsagaincalledoutthebanagainstThebesand,justasbefore,requestedAgesilaustotakecommand’); Hellenica5.4.59:ofl d¢ LakedaimÒnioi(…)pãlin frourãn te ¶fainon ka‹ KleÒmbroton ≤ge›syai §k°leuon (‘TheLacedaemonians(…)againcalledoutthebananddirectedCleombrotustotakecommand’).

7 Note al so the difference in the choice of verb of appearance: ¶ a rÍp°faine/§p°sthvs.§peg¤gneto tÚ metÒpvron,thedifferenceinwordorder,andtheadditionof ≥dhinthecaseof‘autumn’,alsofoundinthecaseof‘afternoon’in Anabas is 3.4.34: ≤n¤ka d' ∑n ≥dh de¤lh (‘as soon as it came to be late in theafternoon’).

8 Hellenica5.1.21:…w d¢ ≤m°ra Íp°fainen(‘asdaywasdawning’); Hellenica5.4.9:




( dihgÆsomai d¢ ka‹ tØn mãxhn

), and a sequence of introductorysentencesdescribingthesituationbothpartiesfoundthemselvesin(sunªsan m¢n gãr… tÚ eÈ≈numon e‰xon ); from this point onwards,theadversariesapproach(suniÒntvn d° ).The narrative isorganizedin two spatial segments, marked by two genitive absoluteconstructions+ d°: suniÒntvn d°and…w d¢ tri«n ¶ti pl°yrvn §n

m°sƒ ˆntvn.Whilethesegenitiveabsolutesatthesametimepropelnarrativetime forward, temporal progression of real time is indicatedwithinthefirstspatialsegmentbymeansoftheclausesheadedby

t°vw m°n…≤n¤ka d°,whichcombinationfunctionsunderthescopeoftheinitial d°(suniÒntvn d° ).Asfarasthecontentoftheseclausesisconcerned,thesuccessivetemporalphasescreatesuspense:thereisaperiodoftimeinwhichsilencewasmaintainedoneitherside,andnoaction was undertaken, which is followed by one in which theThebans raised the battle-cry and rushed forward at the double.Schematically:

Figure1:Spatio-temporalSegmentationinAgesilaus2.10

SpatialSegment1 suniÒntvn d°

Real t°vw m¢n sigØ pollØ ∑n ép'Time émfot°rvn

TemporalProgression ≤n¤ka d¢ épe›xon éllÆlvn

Narrative ˜son stãdion

Time élalãjantew ofl Yhba›oi drÒmƒımÒse §f°ronto

SpatialSegment2 …w d¢ tri«n ¶ti pl°yrvn §n m°sƒˆntvn

éntej°dramon épÚ t∞w ÉAghsilãou fãlaggow œn ÑHripp¤daw§jenãgei...

AstheorganizationoftheRealWorldsituation onthediscourselevel istakencareofbythegenitivesabsolute,thesubclauseheadedby≤n¤ka,




anaphoricpronountaËta

+∑n

,thereisaback-referencetoeventsintheprecedingdiscourse.Thesesubclauses,too,serveasatimeanchorforthesuperordinatesituation,inawaycomparableto§n ⁄ d¢ taËta§bouleÊonto in [9]above,yethere thecontentof the subclause isentirelygiven.

[15] Anabasis1.10.15-16ka‹ ı LÊkiow ≥las° te ka‹ fidΔn épagg°llei ˜ti feÊgousin énåkrãtow. sxedÚn d' ˜ ka‹ ¥liow §dÊeto. §ntaËya d'¶sthsan ofl ÜEllhnew ka‹ y°menoi tå ˜pla énepaÊonto:

AndLyciusrodeup,looked,andbroughtbackwordthattheenemywereinheadlongflight.A thesunset.ThentheGreekshalted,groundedarms,andrestedthemselves.

[16] Anabasis3.1.33-34§pe‹ d¢ pãntew sun∞lyon, efiw tÚ prÒsyen t«n ˜plvn §kay°zonto: ka‹§g°nonto ofl sunelyÒntew strathgo‹ ka‹ loxago‹ émf‹ toÁw •katÒn.

sxedÚn m°sai ∑san nÊktew. §ntaËya ÑIer≈numowÉHle›ow presbÊtatow Ãn t«n Proj°nou loxag«n ≥rxeto l°gein

œde.All had come together, and seated themselves at the front of theencampment;andthegeneralsandcaptainsthusassembledamountedinnumbertoaboutonehundred.B itwasnearlymidnight.ThenHieronymustheElean,whowastheeldestofProxenus’captains,begantospeakasfollows.

[17] Hellenica2.3.35-36(fromaspeechbyTheramenes)§gΔ d¢ épologoÊmenow …w diå tÚn xeim«na oÈd¢ ple›n, mØ ˜ti énaire›syai toÁw êndraw dunatÚn ∑n, ¶doja tª pÒlei efikÒta l°gein,§ke›noi d' •aut«n kathgore›n §fa¤nonto. fãskontew går oÂÒn tee‰nai s«sai toÁw êndraw, pro°menoi épol°syai aÈtoËw épopl°ontew’xonto. oÈ m°ntoi yaumãzv ge tÚ Krit¤an ~paranenomhk°nai~: ˜

, oÈ parΔn §tÊgxanen, éll' §n Yettal¤& metåPromhy°vw dhmokrat¤an kateskeÊaze ka‹ toÁw pen°staw Àplizen §p‹toÁw despÒtaw.

Isaidinmydefencethatonaccountofthestormitwasnotpossibleeventosail,muchlesstopickupthemen,anditwasdecidedbythestatethatmypleawasareasonableone,whilethegeneralswereclearlyaccusingthemselves.Fortheysaiditwaspossibletosavethemen,yettheysailedawayandleftthemtoperish.Idonotwonder,however,thatCritiashasmisunderstoodthematter;f

,itchancedthathewasnothere;hewasestablishingademocracy




Andnowitwasmidday,andtheenemywerenotyetinsight.W,therewasseenarisingdust,<whichappearedat

first>likeawhitecloud;sometimelater<itappeared>likeakindof blackness in the plain, extending over a great distance. W

,therewaspresentlybronzeflashing hereandthereandspearsand the hostilerankscame intosight. Andtherewere…

Inthispieceofdiscoursethespeakerisunequivocallyconcernedwithtemporalorganization(≥dh…∑n m°son ≤m°raw…oÎpv…≤n¤ka d¢

de¤lh §g¤gneto… xrÒnƒ d¢ suxn“ Ïsteron…˜te…tãxa dÆ ).Thesubclausesunderconsiderationpropelnarrativetimeforward;bothsubclausescanbereplacedbyaphraselike‘aftersometime’or‘sometimelater’.Inthefirstsentence,twostatements,closelyconnectedbyte…ka¤,arepresented:itwasmiddayandtheenemywerenotinsight. The temporal situation as expressed in this first sentence isalteredinthesecondsentence:thepreposed,sentence-initiallyplacedtemporal subclause ≤n¤ka de¤lh §g¤gneto propels narrative timeforwardandpresentsanew temporal settingforthedescriptionofthe

hostile army coming in sight. This description fills the rest of theepisode, insuccessive steps that are also temporallyorganized: note xrÒnƒ d¢ suxn“ Ïsteron, ˜te d° … tãxa dÆ. Regarding thedistributionofparticles,the d°after≤n¤kasignalsthenextstepinthemainlinetemporalsequence,whereastheinstancesof d°after xrÒnƒand ˜te presentsuccessivestepsintime,tobecontinuedwith ka‹(∑san ), under the scope of the initial d° ,inasmuchastheseeventsoccurred‘≤n¤ka de¤lh §g¤gneto’.Itshouldbenotedthattheverbalconstituentsofthetwotemporalsubclausesareimperfects(§g¤gneto,

§g¤gnonto ).Thisisbecausethecontentoftheseclausesisnotassertedas such. Throughout the passage, the presence of the narratorreducedtoaminimum,inordertocreatethenarratologicaleffectof relayingtheeventsfromthepointofviewoftheGreeks.Compareforinstance the aspect of ‘visualization’ as praised by Lendle in hiscommentaryadloc.(1995:67):“ManspürtdemBerichtXenophonsdie Spannung an, mit welcher er das eindrucksvolle Manöverbeobachtet hat, und muß die Anschaulichkeit seiner Darstellung

bewundern:mansiehtförmlichdieBlitzevorAugen,welchevondenblank geputzten Speerspitzen, Harnischen, Metallbeschlägen dergepanzerten Pferde usw. je nach ihrerBewegung vor dem dunklenHintergrund aufstrahlten und wieder verloschen.” Lendle’s




(predominantlypreposed) subclauses headedbya semanticallynon-specificrelator.

1.3 Conclusion

Whetherasubclauseispreposedorpostposedtoitsmatrixclause,andirrespective of the nature of the relator, there is always a relationbetweenthecontentoftheclausesofwhichtheclausecombinationconsists.Inthecaseofsubclausesusedinhistoriographicnarative,this

means that the historian has decided to select (at least) two RealWorld‘happenings’forrecordingatacertainpointinhistext,andtocommunicate to the recipient of the text that he has perceived orconceived a relation between them. As such, this involves a firstdecisiononpresentation.Aseconddecisiononpresentationcanbemade,thatis:thehistorianmaydecidetospecifythisrelation.Whenthis happens, the historian narrows down the possibly conceivablesemantic relations between the content of the two clauses to theappropriateone.Thus,whenforinstanceheuses ˜tior diÒti,therelationispresentedas‘causal’;when,ontheotherhand,heuses˜te,≤n¤ka,or§n ⁄,aRealTimerelationbetweeneventsisindicated,§n ⁄usuallyindicatingsimultaneityofthetwosituationsdescribed.Whilethis happens within the sentence, such subclauses may evoke otherparts of the text, and therewith add to the comprehension of thecontentofthematrixclausewithinthediscourseasawhole—whetherthe subclause is preposed ([5], [8], [9], [15], [16], [17], [18]) orpostposed ([22a]). Preposed subclauses headed by a semantically

specific relator when accompanied by d° may help to articulateindividualsegmentsofthetext.Thenatureoftherelatorinvolvedthendeterminesthenatureoftheorganizationofthediscourseatthatpoint. Thus, while the preposed subclauses headed by a temporalrelatorin[6],[7],and[21]markRealTimetemporalprogressionatthe onset of a new segment, successive temporal phases arearticulated. The articulation of unequivocally temporal boundariestypicallyoccurswithina largerthematicunit.Thismayhappenatitsclosure ([6]), or within one differently organized: see [7], which isspatially organized by a genitive absolute and a …w-clause, andespecially[21],wheretwotemporalsubclausesarticulatesmall-scaleboundariesinanepisodeabouttheenemyoftheGreekscomingintosight,asappearsfromthewordkatafane›wframingthethematicunit




Whetherthematrixclausefollowingonthesubclauseheadedbyasemanticallynon-specificrelatorpresentsanarrativestatementornot,thespeakerusesthepreposedsubclausetoinformhisaudienceaboutacertainRealWorldsituation.Byimplication,thistakestheformofafactualstatementonthepartofthespeaker.Wemaythussaythatthepreposed subclause when headed by a semantically non-specificrelatorpresentsaRealWorldsituationthereader/hearerneedstobeinformedaboutinordertobeabletocomprehendthesequel.Thisleads to different interpretations of the subclause in different

contexts.22AswasstatedinChapter1,Section2.2.1, ceterisparibustheorder in which the events are narrated mirrors the speaker’sperception of the succession in real time of the Real World‘happenings’theyrepresent.Whenboththepreposedsubclauseandthefollowingmatrixclausepresentanarrativestatement,theeventsdescribedinthetwoclausesareinterpretedas‘insequence’,inwhichcase wewill probablyuse thesubordinatingconjunction ‘when’ (or:‘after’); when we can conceive of a ‘causal’ relation between thecontent of the clauses of which the clause combination consists wemay thinkof ‘because’; however, in translation ‘when’ is still tobepreferred.Outofnecessititywebringoutourinterpretation 23ofthe

———22 Intext typesother than narratives,semantically non-specificrelatorsmaybe

put in to signal relations between the content of clauses of a different type. Forinstance in Plato, Gorgias 448 b11-c3: XAI. — Efi d° ge ∏sper ÉAristof«n ıÉAglaof«ntow μ ı édelfÚw aÈtoË ¶mpeirow ∑n t°xnhw, t¤na ín aÈtÚn Ùry«w§kaloËmen; — PVL. — D∞lon ˜ti zvgrãfon. — XAI. — NËn d' §

, t¤na ín kaloËntew aÈtÚn Ùry«w kalo›men;

(‘CHAE.—And if Gorgias were expert in the same art as Aristophon, son of Aglaophon, or his brother, what name should we rightly givehim?—POL.—Obviouslythatofpainter.—CHAE.—Butasitis,s <wewouldliketoknow>i ,bywhatnameweshouldrightlycallhim?’),notonlythematrixclause,buteventhesubordinateclauseitselfareinterrogativeclauses.ÉEpeidÆindicatesthat,accordingtothespeaker,thetwoquestionsarerelated,insofarastheanswertothesecondquestionwillbefoundwithoutdifficultyoncethefirstquestionisanswered.Cf.alsotheusageofsemanticallynon-specificrelatorsheading clausesinclausecombinationsusuallyregardedasparatactic,e.g.,Plato,Gorgias474b6-7: §gΔ d° ge oÎt' §m¢ oÎt' êllon ényr≈pvn oÈd°na. §

; (‘and I, that neither I nor anyone else in theworldbelievesit.Y

’) and Plato,Gorgias473e3-4: oÈk o‡ei §jelhl°gxyai, Œ S≈kratew, ˜tantoiaËta l°g˙w ì oÈdeiw ín fÆseien ényr≈pvn; § (‘doyounot think yourself utterly refuted, Socrates, when you make such statements asnobodyintheworldwouldassentto?Y

’).23 ThereaderwillobserveonthebasisoftheviewspresentedinthisSectionthatI




todeath,—itwasallconjectures,ofonesortandanother;andnograveofhiswaseverseen.

ThisexampleistakenfromapassagewhereClearchusisthereporterofhowOrontas’trialwasconducted(1.6.5).Thenarrativesequenceof Orontas’s report runs up toand includingOrontas having beenconducted into the tent of Artapates (efiw tØn ÉArtapãtou skhnØnefisÆxyh,expressedinthesubclause).From metå taËtainthematrixclauseonwards,ashiftinthepointofviewfromwhichtheeventsarerelayediseffected:itisthenarratorXenophonhimselfwhoisfurther

responsible for the statement that ‘from that moment noman eversaw Orontas living or dead, nor could anyone say from actualknowledgehowhewasputtodeath’;especiallytheaddition:‘itwasallconjectures,ofonesortandanother;nograveofhiswaseverseen’iscertainly Xenophon’s. An argument for this reading would be thatOrontas, who is a topical discourse participant, performs thesyntacticalfunctionofsubjectofthesubordinatedfiniteverb efisÆxyhwithout reference, but then is referred to by his full name in the

accusative(ÉOrÒntan )inthematrixclause.The shift in point of view coincides with the point at whichClearchus’s role as an eye-witness comes to an end. Xenophonhimselftakesoverassoonastheeventsweretakingplaceinthetentof Artapates,outofthereporter’ssight,andpresentsthecircumstancessurroundingOrontas’sdeathinasuggestivemanner.27InasituationwherethepresentationofRealWorldeventscoincideswithachangeinpointofview,thenon-specificrelator §pe¤offersthenarratorthepossibility of smoothly relating a narrative statement (‘he was

conductedintothetentofArtapates’)andanon-narrativestatementover a change of point of view without indicating any semanticrelationsthatare,inthecontextunderconsideration,inexpedient;forthat matter, this is just another instance where the temporalorganizationisrealizedby meansof thetemporallyspecificadverbial

metå taËtainthematrixclause,notbythechoiceofrelatorortensestem.

[27] Anabasis1.6.5-7

Kl°arxon d¢ ka‹ e‡sv parekãlese sÊmboulon, ˜w ge ka‹ aÈt“ ka‹to›w êlloiw §dÒkei protimhy∞nai mãlista t«n ÑEllÆnvn. §pe‹ d'

———27 Commentators, too, put out feelers about what happened to Orontas and

suggestthathewasburiedalive.



122 CHAPTERTHREE

questioned about a situation C, which follows on B ( metå taËta

).Beyond the shift of addressee and point of view, the non-specificrelator§pe¤indicatesarelationbetweenthecontentoftwoclausesof a different type, a declarative clause and an interrogative clause,respectively:30thesubclausepresentsfactualinformationaboutaRealWorldsituation,in the lightofwhichthequestionisasked,and thisproducesanunexpected,rhetoricaleffect(cf.[24]above).

[28] Agesilaus2.13-14§peidØ d¢ ≤ m¢n n¤kh sÁn ÉAghsilãƒ §g°neto, tetrvm°now d' aÈtÚw

proshn°xyh prÚw tØn fãlagga, proselãsant°w tinew t«n flpp°vnl°gousin aÈt“ ˜ti t«n polem¤vn ÙgdoÆkonta sÁn to›w ˜ploiw ÍpÚt“ na“ efisi, ka‹ ±r≈tvn t¤ xrØ poie›n. ı d¢ ka¤per pollå traÊmata¶xvn pãntose ka‹ panto¤oiw ˜ploiw ˜mvw oÈk §pelãyeto toË ye¤ou, éll' §çn te épi°nai ˜poi boÊlointo §k°leue ka‹ édike›n oÈk e‡a, ka‹prop°mcai §p°taje toÁw émf' aÈtÚn flppe›w ¶ste §n t“ ésfale›§g°nonto.

, par∞n dØ yeãsasyai ¶nyasun°peson éllÆloiw tØn m¢n g∞n a·mati pefurm°nhn, nekroÁw d¢keim°nouw fil¤ouw ka‹ polem¤ouw met' éllÆlvn, ésp¤daw d¢

diateyrumm°naw, dÒrata sunteyrausm°na, §gxeir¤dia gumnå kole«n,tå m¢n xama¤, tå d' §n s≈mati, tå d' ¶ti metå xe›raw.

ThevictorylaywithAgesilaus;buthehimselfwascarriedwoundedtohisbattle-line;thensomehorsemenrodeup,andtoldhimthateightyof theenemyretainingtheirarmshadtakencoverinthetemple,andtheyaskedwhattheyshoulddo.Thoughwoundedineverypartofhisbodywitheverysortofweapon,hedidnotforgethisdutytowardsthegods,butgaveordersthatthesemenshouldbeletgowhithersoevertheywished, and would not suffer them to be harmed, and charged his

escortofcavalrytoconductthemtoaplaceofsafety.,anoddspectaclemettheeye,asone surveyed the sceneofthe conflict—theearthstainedwith blood,friendandfoelyingdeadsidebyside,shieldssmashedtopieces,spearssnappedintwo,daggersbaredoftheirsheaths,someontheground,someembeddedinthebodies,someyetgrippedbythehand.

The subclause §pe¤ ge mØn ¶lhjen ≤ mãxh contains a narrativestatement about a Real World situation. The matrix clause withwhich it combines does not contain a narrative statement, but

———30 For a characterization of sentence types, see Risselada (1993: 71-2), who

proposes “to characterize declarative sentences as ‘presenting a proposition’,interrogative sentences as ‘presenting a proposition as (partially) open’, andimperativesentencesas‘presentingthecontentofapropositionforrealization’”.



124 CHAPTERTHREE

TheSciritanssawthemandfellbackata fasterpacethanawalk.Andnotoneofthemwaskilled;nevertheless,theThebanssetupatrophy,because the Sciritans who had climbed the hill had retired. As forAgesilaus, <for him to do so>, he withdrew andencamped at the veryspot where he had seentheenemydrawn up;then on the following day he led his army away by the road toThespiae.

Thesesubclauseshavethenarratologicaleffectofindicatingthataparticipant’sviewonaRealWorldsituationispresented.ThewordkairÒw in [29] is an evaluative NP: one has to judge whether it is

kairÒwornot;thesamegoesfor Àrain[30].Ineithercase,therearetwo possible candidates for the attribution of the evaluation: thenarrator or a participant. Expressions like kairÚw §dÒkei e‰naiandÀra ∑ncallforadativephraseindicatingtowhomorforwhomthisReal World situation obtains—cf. Ehrlich (1990: 18); this dativephrase is not expressed.31 In general, the dative phrase is deletedeasier if the situation is presented from the point of view of aparticipantthanifthesituationispresentedfromthepointofviewof

thenarrator.32

In [29], with the ofl m°n…ofl d°-sequence preceding the §pe¤-clause,thenarratorsimplycouldnothaveusedadativephraselike

aÈto›winthesubclause,asthereferenceastowhichpartyhemeantwould have been unclear. By means of §pe¤, the content of thesubclauseisrelatedtothecontentofitsmatrixclausewhosesubjectisofl pol°mioi,sothat,inthelinearperceptionofthetext,onlyafterreading/hearingtheclausecombinationasawholedoesitbecome

———31 Cf. also Anabasis2.3.9:ı d¢ Kl°arxow e‰pe: (…)

, épÆggellen ˜ti sp°ndoito, ka‹ eÈyÁw ≤ge›syai §k°leue prÚw tépitÆdeia(‘Clearchussaid:(…).W ,hereportedthatheaccepted the truce, and directedthem tolead the way immediatelyto theprovisions’).Therelevanceofthepresenceorabsenceof aÈt“fortheattributionof pointofviewmaybegatheredfrom Anabasis1.2.1,wherethedativephraseto§dÒkeiis expressed, in a situation where the sentence as a whole is relayed from thenarrator’spointofview,asappearsfromtÆn…prÒfasin §poie›tointhematrixclause: , tØn m¢n prÒfasin§poie›to …w Pis¤daw boulÒmenow §kbale›n pantãpasin §k t∞w x≈raw(poreÊesyai

aÈt« ênv ≥dhD: aÈt“ ≥dh poreÊesyai ênvdet.;‘w,thepretextheofferedwasthathewishedtodrivethePisidiansoutofhislandentirely’).

32 Just asa dative phrase may bedeletedundercertainconditions inordinaryconversation,oneofwhichexistswhenthedativeNPiscoreferentialwiththespeakeroraddresseeofthespeechevent(Ehrlich,1990a:18).




[34] Hellenica4.7.7§k d¢ toÊtou boulÒmenow teix¤sai froÊriÒn ti §p‹ ta›w paråKhloËsan §mbola›w, §yÊeto: ka‹ §fãnh aÈt“ tå flerå êloba. …w d¢toËto §g°neto, épÆgage tÚ strãteuma ka‹ di°luse, mãla pollå blãcaw toÁw ÉArge¤ouw, ë .

After this, desiring to fortify a garrison post at the entrance to theArgivecountrywhichleadspastMountCelusa,heofferedsacrifice;andtheliversofthevictimswerefoundtobelackingalobe.Inviewof this event, he led his army away and disbanded it—he had inflicted

verygreatharmupontheArgivesi.

Withtheaoristfiniteverbs épÆgage tÚ strãteuma ka‹ di°luse,thenarratorroundsoffhisaccountofAgesipolis’sinvasionofArgos.Byway ofafterthought,he adds thatAgesipolishad inflictedverygreatharm upon the Argives, using a postposed participial clause ( mãlapollå blãcaw toÁw ÉArge¤ouw ).Thisparticipialclause, initsturn, isfollowedbyanotherparticipialclause,headedby ëte; ëtesignalsthatthecontentoftheclausetofollowistobeviewedasanexplanationforthecontentoftheprecedingclause.

Ifweconsidertheotheroptionthespeakerhasathisdisposal—theuseofasubclause:*§pe¤ & c./˜ti éprosdokÆtvw aÈto›w §n°balen —,we must conclude that this would have been an option,38 had thenarratorwantedtopresenttheactofinvadingasanexplanation:‘heinflictedvery greatharmupon theArgivesinasmuchas/becausehehadinvadedtheirlandunexpectedly’.However,byusingaparticipialclauseinsteadofa subclause,thespeakerachievesthatthepostposed

narrative statement is interpreted as ‘he inflicted very great harmupontheArgivesinasmuchashisinvasionwas unexpected ’:usingthenon-finite participle, the speaker focusses on the adverb

éprosdokÆtvw , which is appropriate since it is precisely itsunexpectednessthatmadeAgesipolis’sinvasionsuccessful.Thechoiceofclausetypeisthereforepragmaticallymotivated.

Thequestionremainswhy ëteisaddedtotheparticiplehere.39

Usually,apostposed(aorist)participleeitherexpresseshowthemain

———38 In that case, the subclause would have a participial clause ( mãla pollå blãcaw toÁw ÉArge¤ouw )asits matrixclause,a clausecombinationthatis preferablyavoided.

39 Inthe Hellenicaand Anabasis, ëteoccurs31times( Anabasis 4.5.18isdisputedandthereforenotincluded)attheheadofaparticipialclause(withpresentstem:19x;



132 CHAPTERTHREE

eventcameabout(so-called‘coincidental’usage)orpresentsafurtherexplanation why a participant acts like he does (type: …[ FINITE

VERB ],nom¤saw…)orpresentsanindependentlyinformativepieceof information that for varying reasons is placed after the main verb.Here, ëte seems to be added because the participial clause underconsiderationdoesnotfitoneofthesecategories.Without ëte,thefinalparticipialclausewouldstillbeinterpretedbythereader/heareras expressing an explanation of the content of the preceding participial clause, which is in fact the only conceivable relation

betweenthecontentofthetwoclauses.However,especiallysincetherelationbetweenthecontentofthetwoclausesresidesprimarilyinthefactthatbyinvading éprosdokÆtvw Agesipolisinflictedverygreat( mãla pollã )harmupontheArgives,thereader/hearerissparedthecognitiveeffortofinterpretingthismessagebytheadditionof ëte.Inordertoestablishasuccessfulcomprehensionofwhathewantstosayon the part of his audience, the speaker specifies the Real Worldrelationbetweenthecontentofthetwoparticipialclausesinvolved.

[35] Hellenica7.4.39 ékoÊvn oÔn ı Yhba›ow ±pÒrei te ˜ ti xrÆsaito t“ prãgmati ka‹ éf¤hsi pãntaw toÁw êndraw. ka‹ tª Ístera¤& sugkal°saw t«nÉArkãdvn ıpÒsoi ge dØ sunelye›n ±y°lhsan, épeloge›to …w§japathye¤h. ékoËsai går ¶fh …w LakedaimÒnio¤ te e‰en sÁn to›w˜ploiw §p‹ to›w ır¤oiw prodidÒnai te m°lloien aÈto›w tØn Teg°an t«nÉArkãdvn tin°w. ofl d¢ ékoÊsantew §ke›non m°n, k

, éf¤esan: p°mcantew d' efiw YÆbaw pr°sbeiw kathgÒroun aÈtoË …w de›n époyane›n.

TheThebanaccordingly,onhearingthis,wasatalosstoknowhowheshoulddealwiththematter,andreleasedallthemen.ThenonthefollowingdayhecalledtogetherasmanyoftheArcadiansaschosetogatherandsaidinhisdefencethathehadbeendeceived.Forhehadheard,hesaid,thattheLacedaemonianswereonthebordersunderarmsandthatsomeoftheArcadiansweregoingtobetrayTegeatothem.Uponhearingthistheylethimgo,a

,buttheysentambassadorstoThebesandbroughtchargesagainsthim,sayingthatheoughttobeputtodeath.

———withaoriststem:6x;withperfectstem:6x):19xheadingaconjunctparticiple(nom.case:18x;acc.case:1x),12x headingagenitiveabsolute;preposed:19x (5timesof whichatthesentence-initialposition),postposed12x;itisnotusedinreportedspeech.Inthe Agesilausitisnotusedatall.



REALWORLDCONSTRUCTIONANDTEXTARTICULATION 147

theprecedingrecordingofheractivewarfareagainstthecitiesonthecoast—note the fact that the topical verb sunestrateÊeto,animperfect,occupiesthesentence-initialposition.9

Whereasthechange-oversfrom DU4toDU5,andfromDU5toDU6arerathersmooth,there-startofthenarrativesequenceinDU7receivesextramarking.Beforethefirstnarrativeclauses efiselyΔn

épopn›jai (l°getai)arereached,Meidiashastobeintroduced.Hisappearanceandwayofconductarepivotaltothestory;this‘discoursepeak’10ismarkedbyanunusualclusterofconjunctparticiplesand

genitive absolute constructions that all provide backgroundinformation—note the use of the aspectual forms: present stemparticiples (oÎshw, n , fulattom°nhw ,pisteuoÊshw ka‹

éspazom°nhw ) and one passive aorist participle ( énaptervye‹w ).Thepresentstemsentence-initiallyplacedgenitiveabsolute≥dh d' oÎshw

aÈt∞w §t«n pl°on μ tettarãkonta,whileexpressinginformationthatdoes not belong to the narrative assertion, helps to articulate thetransitionfromageneraldescriptiontoaspecificevent:itcreatestheexpectation that the narrative sequence will be continued. It alsoindicates that some time has passed, albeit implicitly, for theinformationthatManiawasmorethanfortyyearsoldonlyimplies,withthehelpof≥dh,thatshewasyoungerbefore.Atthesametime,thegenitiveabsolutehelpsinparticipanttracking:Maniaremainsanactiveparticipantinthestory,atapointwhereMeidiashastobeintroduced.

Thesecond SPhassixD U’s.TheboundariesbetweentheseDU’sare of unequal status. The onset of D U 9and D U 11 have been

articulated by means of the resumptive participial clause taËta d¢poiÆsawandtheadverbialexpression§k d¢ toÊtou,respectively.Astheseboundarieshavereceivedextralinguisicmarking,thepartofthetexttheyopenhasbeeninterpretedasaBUwithinaparagraph.Onahierarchicallylowerlevel,i.e.,thearticulationofanew DUwithinaBU,thetransitionsarelessheavilymarked.Atthebeginningof DU8,for instance, the verbal constituent occupying the sentence-initialpositionis,again,topical(thenotionof‘killing’iscontextuallyevokedby épopn›jaiinDU7);in DU9and10citieswhichheseizedarecontrastedtotheothercitiesthatwouldnotadmithimwithintheir

———9 SeeH.Dik,1995,Chapter7.10 SeeLongacre,1985.



154 CHAPTERFOUR

Moreover,whenachangeinthecurrenttemporalsettingoccurs,thediscourseislikelytobediscontinuouswithrespectto(oneofthe)othercoherencestrandsaswell.Further,changeswithrespecttooneofthecoherencestrandsitselfmayshowvariousdegreesof(dis-)continuity:referential(dis-)continuitymayinvolveoneormoreparticipants,andbesidesbringingaparticipanttoorremovinghimoffthestage,wefind addition ofa single participant toa topical set ofparticipants;spatial(dis-)continuitymayrangefromratherlocaltomoreglobal,andsoon.

We now state that the higher the degree of discontinuity withrespecttooneormoreofthefourcoherencestrands,thedeepertheincisioninthecompositionofthediscourse,andthehigherthedegreeoflinguisticcodingis,asillustratedinfigure3:

Figure3: ContinuumofDiscontinuityatPointsofSegmentationandLinguisticCoding

LOW DegreeofDiscontinuity HIGH

<———————————————————>LOW DegreeofLinguisticCoding HIGH

<———————————————————>

Asfarassubordinateclausesoccupyingthesentence-initialpositionatpoints of segmentation are concerned, a high degree of desententialization corresponds toa low degreeof linguistic coding,whilealowdegreeofdesententializationcorrespondstoahighdegree

oflinguisticcoding.Therefore,ahighdegreeofdesententializationcorresponds to a high degree of (thematic) continuity, while a lowdegree of desententialization corresponds to a high degree of discontinuity, i.e. discontinuity at points where segmentation takesplacenotonlyontheleveloftextarticulation,butalsoonthelevelof theRealWorldconstruction.

Whenwetakereferent(dis-)continuityforexample,thereferentof theNPthatperformsthefunctionofsubjectofaconjunctparticiplewhile also performing a syntactic function in the matrix clause, is

predictablyacontinuousDiscourseTopic;inthecaseof,forinstance,thelargenumberofclausechainingparticiplesinthenominativecasefoundin extant texts, the reader/hearerneeds littlereference totheparticipantwhoperformstheaction,ashe/sheisacontinuoustopic;



160 CHAPTERFOUR

the two clauses are perfectly coherent: there is a referential linkbetweenthem(thewoman÷ she),thereisaction-eventcoherence(decision÷ effectuation),timeiscontiguous(decision andsubsequent effectuation) and the clauses are about the same location(satrap/mistressofthesameprovince)—yetrecallfromSection2.1thatinthestoryofMania,thesecondclausewasrealizedasan§pe¤-clause(…ı Farnãbazow ¶gnv de›n tØn guna›ka satrapeÊein. ≤ d'§pe‹ kur¤a t∞w x≈raw §g°neto…),andthatwithintheframeworkof thewholestory,≤ d' §pe‹ kur¤a t∞w x≈raw §g°netowasanalysedas

theonsetnotonlyofanewDU ,butalso,partlyduetothenarratorchanginghismodeofdiscourse,ofanembeddedparagraph.

The form a certain statement takes, is, I contend, entirelydependentonhowthespeakerwantstopackagehismessage.Anysequenceofindependentclausesmaybegiventexture,orrelief,bydifferentlinguisticmeansinordertoindicateinterclausalconnectionsorrelations,or,moreglobally,discourseconnectedness.Considerthefollowingnarrativesequence:

TheGreekshadnotyetgonefifteenstadia.Theybegantomeetwithdeadbodies.Theybroughttherearoftheircolumntoapointoppositethefirstbodiesthatappeared.Theyburiedallthatthecolumncovered.Theyburiedthisfirstgroup.Theymarchedforward.Theyagainbroughttherearofthecolumnintolinewiththefirstofthebodieswhichlayfurtheron.Theyburiedallthatthearmycoveredinthesamewayasbefore.

Theyreachedtheroadleadingoutofthevillages.Therethedeadlaythick.Theygatheredthemalltogether.Theyburiedthem.

This sequence of twelve independent clauses may beregarded as athematic unit: all clauses with the exception of 10 have the samesentence-topic, introduced in the first sentence (‘The Greeks’) andreferredbacktoby‘they’.Thesequenceshowsactioncoherenceinthatthetheme,orDiscourse-topic,ofthisactionsequenceis‘burying

corpses’.ThereisnoRealTimeindication,buttheeventsexpressedinthenarrativeclausesarecontiguous:fromthefirstnarrativeclause(2)onwards,narrativetimeispropelledforwardintheclauses3,4,6,7,8,9,11,and12.Finally,nomajorspatialbreakisarticulated:all




itfunctionsasa(thematic)hingebetweentwoadjacentsub-sectionsof oneandthesamethematicunit.

ItisnoteworthythatThompsonandLongacrereport(1985:212)thatthesortofback-referenceweencounterhereinthecaseofclause5is

thestandardlinkageinnarrativeandprocedurallanguagesinatypicalPhilippinelanguage.Theunitssolinkedaresentencesorembeddedparagraphs.Instancesinvolvingthelatterareamorecomplexvariationofrelativelysimplerstructuresofthesortillustratedhereinwhicheach

successive ‘build-up’ of the paragraph is a separate sentence whichcarriesforwardtheeventline.Insuchsimplerstructuressentence jhasan initial time clause or a time phrase that is a back-reference tosentencei —muchasdescribedastail-headlinkage(…).Suchastructurecould beeither narrative or procedural and mightbetter be termedsomewhatmoreneutrallyaSEQUENCEPARAGRAPH.19

As our thematic sequence is a narrative sequence with a strong proceduraltouchtoit(i.e.thepassagealsoaddressesthequestionof how the bodies were buried), clause 5 in the passage under

consideration may be said to illustrate the kind of phenomenondescribedby ThompsonandLongacre, giventhat itis aresumptiveclause.

The §pe¤-clause 9 is not resumptive, but a progressive, and anarrative clause; however, by introducing a new phase in thedescriptionoftheRealWorld,thisclause,too,helpstoarticulatetheboundarybetweentwosub-sectionsofthesequenceparagraphontheleveloftheRealWorldconstruction.

4.2 TheArticulationofThematicUnits

Thearticulationofanewphaseinthedepictedworldmay,undercertainconditions,occurwithinanepisodethat,asawhole,showsthematicunity.Generally,however,thearticulationofabreakinthecontinuityofthedepictedworldtakesplaceataboundarybetweentwoseparatethematicunits.Hereisanexample:

———19 They alsoreport (ibid .,213)that“back-referenceofthissortisendemicfor

many Philippine languages” and that “such regularity of back-reference is alsocharacteristicofmanystructuresinNewGuinea”.



164 CHAPTERFOUR

[6] Anabasis1.1.1-3

Dare¤ou ka‹ Parusãtidow g¤gnontai pa›dew dÊo, presbÊterow m¢nÉArtaj°rjhw, ne≈terow d¢ KËrow:

[ A ], §boÊleto tΔ pa›de émfot°rv pare›nai. ı m¢n oÔn

presbÊterow parΔn §tÊgxane: KËron d¢ metap°mpetai épÚ t∞w érx∞w ∏w aÈtÚn satrãphn §po¤hse, ka‹ strathgÚn d¢ aÈtÚn ép°deijepãntvn ˜soi §w KastvloË ped¤on èyro¤zontai. énaba¤nei oÔn ıKËrow labΔn Tissaf°rnhn …w f¤lon, ka‹ t«n ÑEllÆnvn ¶xvnıpl¤taw én°bh triakos¤ouw, êrxonta d¢ aÈt«n Jen¤an Parrãsion.

[ B ]

, Tissaf°rnhw diabãllei tÚn KËronprÚw tÚn édelfÚn …w §pibouleÊoi aÈt“. ı d¢ pe¤yetai ka‹sullambãnei KËron …w épokten«n: ≤ d¢ mÆthr §jaithsam°nh aÈtÚn épop°mpei pãlin §p‹ tØn érxÆn.

DariusandParysatishadtwosonsborntothem,ofwhomtheelderwasArtaxerxesandtheyoungerCyrus.

,hewishedtohavebothhissonswithhim.Theelderhappenedtobewith him already; but Cyrus he summoned from the province over

which he had made him satrap, and he had also appointed himcommanderofalltheforcesthatmusterintheplainofCastolus.Cyrusaccordinglywentuptohisfather,takingwithhimTissaphernesasafriendand accompanied bythree hundred Greekhoplites,under thecommandofXeniasofParrhasia.

,TissaphernesfalselyaccusedCyrustohisbrotherofplottingagainsthim.HebelievedtheaccusationandarrestedCyrus,withtheintentionofputtinghimtodeath;hismothermadeintercessionforhim,andsenthimbackagaintohisprovince.

Example[6]presents,ofcourse,thebeginningofthe Anabasis:thefirstsentenceisdevotedtotheintroductionofmajorparticipants;sinceitisthefirstsentenceofthework,thepresentindicativeg¤gnontai isproperly speaking without time reference. The subclause §pe‹ d¢±sy°nei Dare›ow ka‹ Íp≈pteue teleutØn toË b¤ou does severalthingsatthesametime:itintroducesthefirsttimereference(±sy°nei… ka‹ Íp≈pteue:pasttenses);moreover,itintroducesthefirstthemeofthenarrative:Darius’illness.Thecontentofthesubclauseisnot

presentedasanarrativestatementofitsown;byvirtueoftheverbalconstituents being imperfects, it is to be understood in connectionwiththeverbalactionofthematrixclause:‘Inviewofthefactthat helaysickandsuspectedthattheendofhislifewasnear , Dariuswishedtohaveboth




his sons with him’ . Then, for several sentences, the sequence isorganized around its participants:ı m¢n oÔn presbÊterow parΔn§tÊgxane … KËron d¢ metap°mpetai … énaba¤nei oÔn ı KËrow.Thethematicsetting remainsthesame,inasmuchasanythematicbreakdefined intermsofdiscontinuityof participants, time,placeand/oraction-sequenceisabsent.

Then,withthesubclause§pe‹ d¢ §teleÊthse Dare›ow ka‹ kat°sthefiw tØn basile¤an ÉArtaj°rjhw,thereisdiscourseturbulence.Theold theme is no longer current, and the thematic setting changes:

Darius disappears from the stage, as he dies, and Artaxerxes isestablishedasthenewking.Thistime,aoriststemsareusedinthesubclauseinasmuchasthesubordinatedverbalactionbelongstothenarrativeassertion.

From the point of view of discourse organization, the two§pe¤-clausesmarked[ A ]and[ B ] are used for thematic segmentationbetween thematic structures. After the introductory first sentence,subclause [ A ] opens a thematic unit about what happened whenDariuslaysickandexpectedthattheendofhislifewasnear.ThesecondthematicunitisabouttheeventsthatoccurredafterDariushaddiedandArtaxerxeshadsucceededhimtothethrone.Subclause[ B ] createsacontent-oriented boundary between the unit concernedwiththeoldkingabouttodie,andtheunitabouteventsaftertheoldkinghadbeensucceedednaturallybytheeldersonArtaxerxes.Atthesametime,subclause[ B ]makesfordiscoursecoherencebycontinuing the chain of events initiated by subclause [ A ]: (§pe‹ d° ) §teleÊthseDare›ow ka‹ kat°sth efiw tØn basile¤an ÉArtaj°rjhw follows

naturallyon(§pe‹ d° ) ±sy°nei Dare›ow ka‹ Íp≈pteue teleutØn toË b¤ou : a discourse-level strategy known as Head-Head linkage (seeThompson& Longacre,1985:211).

Theseboundariesarearticulatedbymeansofpreposed,sentence-initiallyplacedsubclauses,as the segmentationof thetext coincideswithamomentofdiscontinuityintheRealWorld.

Inthefollowingexamplethetextissegmentedatapointwhereanewparticipantisintroduced,whilethereisthematicdiscontinuitywithrespecttotimeandplace;here,firstafinitemainclause20isused,

———20 Thereaderisremindedoftheendoftheepisodeinthe AgesilausonAgesilaus’s

activities in Asia (Chapter 2, Section 4). This involves a major relocation andtherefore a thematicbreak because ofthe spatial discontinuity inthe Real Worldsituationdescribed,anditisheavilymarkedbyafinitemainclause(1.38):theclosing-



168 CHAPTERFOUR

introductorysentence—notethathisnameoccupiesthesentence-finalposition—, who becomes the topic of the backgrounded passage(otow d° ).The mainverbs of these four backgrounded sentencesareimperfects.Thereturn tothemainnarrative lineismarkedinthreeways.Firstly,Xenophonrepeatsthepronoun otowattheimmediatebeginningofthesentence.Secondly,heusestheparticle oÔn,whichmarks the preceding as introductory to the point the author finallywantedtomake, viz.,Polydamas’sstatement[ e‰pe:A 2 ]. Thirdly, wefindasubclause,withverbalrepetitionof A 1.Thesubclauseobviously

istheretocreatecoherencebetweenthedifferentpartsofthisepisode.The subclause §pe‹ éf¤keto efiw tØn Lakeda¤monaisaback-

reference to the phrase éfikne›tai prÚw tÚ koinÚn t«nLakedaimon¤vn —note that back-reference need not be an exactsemanticreproductionoftheoriginalmaterial,butmayinvolvelossorgain of information; a case in point is provided by the differencebetween prÚw tÚ koinÚn t«n Lakedaimon¤vn and efiw tØnLakeda¤mona : in the independent main clause it is stated thatPolydamas presented himself before the general assembly of theLacedaemonians, so that A 2: e‰pehasalreadybeenprepared.Theresumed information itself forms the start of a new sequenceparagraph;thereisthematicdiscontinuitywithrespecttoparticipants(Poludãmaw FarsãliowisintroducedintothediscourseaftertheLacedaemonians were dealing with Phocian ambassadors in 6.1.1,hence ka¤ in the first sentenceof thisepisode) andtime (sxedÚn …per‹ toËton tÚn xrÒnon ); perhaps the use of the historical present isalso indicative of a boundary between two larger thematic

structures.21Theresumptivesubclause,asitwere,recapitulatesthethematicboundaryalreadyexpressedbyamainfiniteclause,andthenew sequence paragraph is re-started after Polydamas has beenproperly introduced. All in all, the high degree of thematicdiscontinuity at the point of segmentation between two largerthematicstructuresisreflectedbythehighdegreeoflinguisticcoding:a finite main clause expressing the new Real World situation thediscoursewillbeconcernedwith,thecontentofwhichisreferredbacktoina—onlyweaklydesententialized—subclausethat,atthesame

———21 Cf. for instance Hell en ic a 3.1.16: ı d¢ Derkul¤daw §n toÊtƒ t“ kair“

éfikne›tai…(‘ItwasatthisjuncturethatDercylidasarrived…’),whichfollowsonthefinalsentenceofthestoryofManiaproper,presentedinfigure1above.




time,helpsthereaderkeeptrackofthediscourseperspectiveafteradigression.

4.2.1 TemporalDiscontinuityIn example [7] of the preceding Section, the discourse isdiscontinuouswithrespecttoseveralcoherencestrandsatthesametime at a point where the narrator decided to use the finite mainclause,thecontentofwhichisreferredbacktolateroninasubclause.Asubclauseseemstobepreferredover its alternative—a conjunct22

participle—as the recapitulated verbal action should remind thereader/heareroftheentirediscontinuity-markingfinitemainclausethesequencestartedwith,sothatherethehighdegreeoflinguisticcodinghelpsthereadertotakenoticeoftheratherdeepincisionordiscourseboundary.However,noteverycoherencestrandneedbeinvolvedatthe sametime:a discourse boundarymay bearticulatedbyahighdegreeoflinguisticcodingwhenthereisdiscontinuitywithrespecttoasinglecoherencestrand,buttoafairlyhighdegree.Hereisanexample:

[8] Anabasis5.3.4-7§ntaËya ka‹ dialambãnousi tÚ épÚ t«n afixmal≈tvn érgÊriongenÒmenon. ka‹ tØn dekãthn, ∂n t“ ÉApÒllvni §je›lon ka‹ tªÉEfes¤& ÉArt°midi, di°labon ofl strathgo‹ tÚ m°row ßkastow fulãttein to›w yeo›w: ént‹ d¢ XeirisÒfou N°vn ı ÉAsina›ow ¶labe.Jenof«n oÔn tÚ m¢n toË ÉApÒllvnow énãyhma poihsãmenow énat¤yhsin efiw tÚn §n Delfo›w t«n ÉAyhna¤vn yhsaurÚn ka‹§p°grace tÒ te aÍtoË ˆnoma ka‹ tÚ Proj°nou, ˘w sÁn Kleãrxƒ ép°yanen: j°now går ∑n aÈtoË. tÚ d¢ t∞w ÉArt°midow t∞w ÉEfes¤aw,

˜t' épπei sÁn ÉAghsilãƒ §k t∞w ÉAs¤aw tØn efiw BoivtoÁw ıdÒn,katale¤pei parå MegabÊzƒ t“ t∞w ÉArt°midow nevkÒrƒ, ˜ti aÈtÚwkinduneÊsvn §dÒkei fi°nai, ka‹ §p°steilen, μn m¢n aÈtÚw svyª, aÍt“ épodoËnai: μn d° ti pãy˙, énaye›nai poihsãmenon tª ÉArt°midi ˜ tio‡oito xarie›syai tª ye“.

, katoikoËntow ≥dh aÈtoË §nSkilloËnti ÍpÚ t«n Lakedaimon¤vn ofikisy°ntow parå tØnÉOlump¤an éfikne›tai Megãbuzow efiw ÉOlump¤an yevrÆsvn ka‹ épod¤dvsi tØn parakatayÆkhn aÈt“. Jenof«n d¢ labΔn xvr¤on»ne›tai tª ye“ ˜pou éne›len ı yeÒw.

———22 Notethatagenitiveabsolutewouldhavebeenana-typicalalternativehere,as

theclausesintheclausecombinationsharethesamesubject:otow … §pe‹ éf¤keto… e‰pe.




Onefeatureremainstobeaddressed:inthecaseoftheconjunctparticiple éfikÒmenow d¢ §p‹ GerastÒn the preposed subordinateclause and its following matrix clause sharethe samesubject,as thefirst§pe¤-clauseanditsfollowingmatrixclausedo,whereasinthecaseoftheclauses …w d' §ke› §g°netoand§pe‹ d¢ §ke›se éf¤ketotheobvious alternative for a subclause would have been a genitiveabsolute.However,asboththeconjunctparticipleandthegenitiveabsolutebelongtothegrammaticalclassofparticipialclauses,thisobservationwouldseemtobeoflittleimportanceasfarastheuseof

§pe¤-/…w-clausesinthispassageisconcerned;toallappearances,itshighdegreeoflinguisticcodingmakesthesubclausetheappropriatesubordinate clause to be used at points of discourse turbulencereflecting a high degree of spatial discontinuity in the Real World(considerespeciallythehighinformationstatusof§pe‹ d¢ yusãmenow˜sa ¶dei ka‹ tîlla ka‹ tå diabatÆria §j∞lyeincombinationwiththe complex structure to follow: ta›w m¢n pÒlesi diap°mcaw

égg°louw proe›pen … aÈtÚw d' ).Whenspatialdiscontinuityisatissue,thediscourseisusuallyalso

(slightly) discontinuous with respect to time. ÉEpe¤-, §peidÆ-,and…w -clauses found at such point should not be simply labelled‘temporal’clauses,forthiswouldnotdojusticetotheircontributiontotheorganizationofthediscourse.Thisis,ofcourse,nottodenythatsometimestheonlypossibleinterpretationofthe,strictlyspeaking,notsemanticallyspecifiedrelationbetweenthecontentofthesubclauseandthematrixclauseisindeedtemporal.Hereisanexample:

[10] Anabasis3.5.3-7

§ntaËya Tissaf°rnhw ka‹ ofl sÁn aÈt“ ka¤ein §pexe¤rhsan tåwk≈maw. ka‹ t«n ÑEllÆnvn mãla ±yÊmhsãn tinew, §nnooÊmenoi mØ tå§pitÆdeia, efi ka¤oien, oÈk ¶xoien ıpÒyen lambãnoien. ka‹ ofl m¢n émf‹Xeir¤sofon épªsan §k t∞w bohye¤aw: ı d¢ Jenof«n §pe‹ kat°bh,parelaÊnvn tåw tãjeiw ≤n¤ka épÚ t∞w bohye¤aw épÆnthsan oflÜEllhnew ¶legen:

ırçte, Œ êndrew ÜEllhnew, Ífi°ntaw tØn x≈ran ≥dh ≤met°rane‰nai; ì går ˜te §sp°ndonto dieprãttonto, mØ ka¤ein tØn basil°vw x≈ran, nËn aÈto‹ ka¤ousin …w éllotr¤an. éll' §ãn pou katale¤pvs¤

———However,thesubclausemakesforthematicsegmentationinadifferentway:afterapassageaboutTeleutiascollectinganarmy(§§37-38),whichisbroughttoanendwiththeclosuretaËta d¢ poi«n, mãla pollØn ¶xvn stratiån éf¤keto efiw tØn •aut«nsummax¤da,the subclausemarksthetransition fromTeleutias’spreparationstohisinvadingtheenemy’scountry.Notealsotheuseof∑lyen,not éf¤keto.




narrative,the§pe¤

-clauseservesthefunctionofsegmentationontheleveloftheRealWorldconstruction,inthecaseathandatapointwherethereisthematic(viz.,spatial)discontinuity:ashiftoccursfromthe location where the speech delivered by Xenophon and hisconversation with Cheirisophus had taken place, which phase hadstartedatı d¢ Jenof«n §pe‹ kat°bh,tothelocationoftheirquarters.Infact,from §ntaËya onwards, the place of action had not beendescribedveryprecisely,andwith§pe‹ d¢ §p‹ tåw skhnåw ∑lyonthespatial setting is redefined. The §pe¤-clause is therefore better

regarded as a thematic boundary-marker than as a ‘temporal’subclause.Moreover,asregardsthenotionof‘time’inthecaseof subclauses of this type, they do not indicate a specific temporalrelationbetweenthecontentofthesubclauseandthecontentofitsmatrixclause,butratherpropelnarrativetimeforward.

4.2.3 ReferentialDiscontinuity[11] Hellenica2.3.23-24ofl d' §mpodΔn nom¤zontew aÈtÚn (sc. tÚn Yhram°nhn )e‰nai t“ poie›n ˜

ti boÊlointo, §pibouleÊousin aÈt“, ka‹ fid¤& prÚw toÁw bouleutåw êllow prÚw êllon di°ballon …w lumainÒmenon tØn polite¤an. ka‹paragge¤lantew nean¤skoiw o„ §dÒkoun aÈto›w yrasÊtatoi e‰naijif¤dia ÍpÚ mãlhw ¶xontaw paragen°syai, sun°lejan tØn boulÆn.

, énaståw ı Krit¤aw ¶lejen œde.

TheThirty,thinkingthatTherameneswasanobstacletotheirdoing whatevertheypleased,plottedagainsthim,andkeptaccusinghimtoindividual councillors, one to one man and another to another, of injuringthegovernment.Andtheypassedthewordtosomeyoung

men,whoseemedtothemmostaudacious,tobeinattendancewithdaggershiddenundertheirarms,andconvenedtheCouncil.,Critiasaroseandspokeasfollows…

ThediscourseprecedingthesubclauseclosesoffwiththestatementthattheThirtyconvenedtheCouncil;thenextsceneisatthemeeting oftheCouncil.Therefore,thediscourseistemporallyandspatiallydiscontinuous when the subclause is used. However, the subclausemay be said to primarily signal referential discontinuity, asTheramenesisre-introducedasadiscourseparticipantwhenwelearnthat Theramenes was present at the meeting. The relevance of Theramenes’spresenceforCritiastodeliveraspeechis,asfarasthelinktotheprecedingunitisconcerned,nottobeviewedasaback-referencetoaspecificactionintheprecedingdiscourse,butasaback-



178 CHAPTERFOUR

reference to the content of the preceding unit as a whole, theprecedingunitbeingabout the ThirtyplottingagainstTheramenes.Critias,ofcourse,isoneoftheThirty,andindeedattheendofthemeetingTherameneswilldie.27

4.3 ThematicContinuity

Inexample[11]oftheprecedingSection,Theramenes’sarrivalonthesceneinitselfconstitutesabreakinthethematic,predominantly

participant/action-orientedcontinuityofthepassageinvolved.Withrespecttothecontinuityofparticipants,itistobenotedthatanyappearance,re-appearance,ordisappearanceofaparticipantinitself involves (a certain degree of) discontinuity in the text. Whether,however,thisdiscontinuityistobeviewedbythereader/hearerasathematicbreak,dependsonthecontextinwhichitoccurs,andtheaimsandpurposesofthestory-telleratacertainpointinthetext.

———27 Ofcourse,the information that ‘someonewaspresent’canbe expressedbya

differentclausetypeaswell,aswitnessedbythefollowing(preposed)genitiveabsolute( Hellenica7.4.1):ka‹ tå m¢n per‹ EÎfronow e‡rhtai: §gΔ d¢ ¶nyen efiw taËta §j°bhn§pãneimi. ¶

ÉVrvpÚw ÍpÚ t«n feugÒntvn katelÆfyh(‘thestoryof Euphron has been told, and I returnto the point from which I digressed to thissubject:t ,whenOropuswasseizedbythosewhohadbeenexiledtherefrom’).Here,Chareswas(still)presentwhenOroposwasseizedbythosewhohadbeenexiled.Buttheessentialdifferencebetweentheuseofthisgenitiveabsoluteandofasubclausein[11]isthatthegenitiveabsoluteisusedafteroneofthegreatesttextualboundariesonecanconceiveof: the intrusionof thenarratorwithafirstpersonreference:“thestoryof

Euphron has been told; I return to the point from which I digressed”. Thus acompletely new start is made with the genitive absolute, and the immediatelyprecedingdiscourse-contentisirrelevant(notethatthedigressiononEuphronhadstartedin7.3.4,andthatwith¶ti går teixizÒntvn t«n Fleias¤vn tØn Yuam¤an,thenarratorrefersbacktothepointwherehe left thePhliasiansfortifyingThyamia, in7.2.23),whereasinthecaseofthe§pe¤-clausein[11]theinformationexpressedinthesubclauseelaboratesonthethemeoftheprecedingunit,sothattherethesubclauseispreferred to encode discontinuity of participants at a point of text segmentation,whereasin Hellenica7.4.1,themainclauseska‹ tå m¢n per‹ EÎfronow e‡rhtai: §gΔ

d¢ ¶nyen efiw taËta §j°bhn §pãneimiexpressathematicboundary,andthetextcanbecontinuedwithamorestronglydesententializedsubordinateclause(cf.Chapter2,Section4:[7]).Thecomparisonrevealsthatthechoiceamongalternatives—inthis

case:asubclauseandagenitiveabsolute—isessentiallyachoicethatismadeanddetermined by the context inwhichtheexpressionisused.In[11],thefactthatTherameneswaspresent,apieceofinformationnecessarytocomprehendthesequelofanarrativesequenceinprogress,isexpressedbymeansofasubclauseinordertoarticulatea thematicboundarywithinanon-goingnarrativesequenceorganized inphases.



182 CHAPTERFOUR

ThefactthatAgesilauscametotherescueofhishorsemenisentirelynewinformationbroughtintothediscourseatthebeginningofanewDU,whichmovesthediscourseforwardto anewpoint:assuch,thegenitiveabsolute bohyÆsantow d¢ ÉAghsilãou sÁn to›w ıpl¤taiwisaprogressive clause. It is used at a point of only slight referentialdiscontinuity: the addition of a participant to a topical set of participants. Agesilaus’s appearance is the turning point in thedescriptionof the cavalrybattle: before he came to the rescue, theGreek horsemen were turned to flight (§tr°fyhsan ofl ÜEllhnew

flppe›w ),but once Agesilaus re-appeared on the stage, the barbarianswithdrew(pãlin épex≈roun ofl bãrbaroi ). Viewed within a largerstructure,however,thesentence-initiallyplacedgenitiveabsolutejustarticulates a sub-section ofa thematic unit: thelarge-scale thematiccaesuraismarkedbythesubclause…w d' efiw xe›raw ∑lyon,whichdividestheentirepassageintotwothematicparagraphs:apreparatorystage,closedoffbytheindependentclausewithanaoristverbform¶peita m°ntoi prÒsyen Àrmhsan ofl bãrbaroi, and the battleencounteritself,whoseonsetismarkedbythe…w -clause …w d' efiw

xe›raw ∑lyon.30

The preposed genitive absolute, then, may be regarded as aminimallynecessarylinguisticdevicetosecuretextualcoherencetobeinsertedatthepointwherethespeaker/narratorwishestoprovideahingebetweentwopartsofhistextwithoutarticulatingathematicboundary.Inexample[14],[15]and[16]suchagenitiveabsoluteisused incomparable but not identical passages that at leasthave incommonthateverypassageiscontinuouswithrespecttoaction-event

coherence:

———30 Atfirst sight, the use ofa resumptivegenitiveabsolute(genom°nhw d¢ taÊthw

t∞w flppomax¤aw )tomarktheendofthebattle-sequenceseemstobeinconsistentwiththeobservationspresentedsofar:whyusea…w-clauseattheboundaryatwhichanewthematicsequencestarts,andagenitiveabsoluteataboundarywherethebattlesceneisclosedoff?Thiscanbeexplainedbypointingoutthat,whereasthe…w-clauseisusedtomarkoffthepreparatorystagefromthebattleitself,thegenitiveabsoluteclosesoff onlyonepartoftheprecedingthematicsequence(‘thefightinghadstopped’),butatthesametimecontinuesa themethathas justbeendeveloped (‘Agesilaus’smilitary

achievements’);inthesequel,thisthemeiselaboratedon:Agesilauswillturnouttobenotassuccesfulaswemightthinkheis,foronthenextday,duringasacrifice-session§p‹ proÒdƒ, the liversappear to be lacking a lobe—highlighted by the use of ahistoricalpresent(g¤gnetai )—;asaresultofthisheturnsandmarchestotheseaandrecognizestheimportanceofanadequatecalvalryforce;seealsoChapter2,Section3.




ofthestoryofCinadonreached.36Astheverbalconstituentinthegenitiveabsoluteneitherhasfocusfunctionnorperformsafunctioninestablishing relative chronology (this job having been done by thetemporal adverbial §ntÚw p°nyÉ ≤mer«n, for which the genitiveabsolute provides a point of reference), the absolute participle, theminimallynecessarylinguisticdevicetosecuretextualcoherence,justestablishesasmoothtransitionfromtheoneDUtotheotherwithinathematically continuous sequence, here at a point where a newparticipantisintroduced.

4.3.2 TheConjunctParticipleThemajorityofconjunctparticiplesinnarrativediscourseisfoundinthe nominative case, preposed to their matrix clause: the actionexpressed in the participial clause and the action expressed in thematrixclauseareperformedbythesameparticipant,andthetwoclausessharethesamesubject.Thisentails,intermsofcoherence,referentialcontinuityoveratwo-clausespan.Butthereseemstobemoretoit:anaphorically,moreoftenthannot,referentialcontinuityisalsofound.Ortoputitdifferently:preposedconjunctparticiplesinthenominativecasearepreferrablyusedwhenacertaintopichasbeenintroducedandperformsaseriesofsequentialactionswithinareferentially continuous thematic sequence, the finite main verbprimarily contributing to the thematic organization of the text inprovidingthemostbasicmaterialforthestructureofthepassageasa

———36

Cf.Jehne,1995:168-169(myitalics):“DieEpisode(sc.thestoryofCinadon)istaber nicht nur insgesamt recht unzureichend eingebettet in denGesamtzusammenhang derxenophontischenDarstellung, sondern siezerfälltauchinsichinzweiTeile ,dienurlosezusammengebundensind.Xenophonschildertzunächst,daß der noch kein Jahr amtierende König Agesilaos opferte und ihm der Sehereröffnete,daß ihmdieGötteretwasSchrecklichesanzeigten;beider Wiederholung desOpferserschienendieZeichennochschlimmer,undalsAgesilaosdasdritteMalopferte,sagtederSeher:»Agesilaos,obwirmittenunterdenFeindenselbstwaren,sooffenbartessichmir«.Daraufhinopfertemaneifrigst denapotropäischenunddenrettendenGöttern,bismanmühseligstgünstigereZeichenerhielt;fünfTagenachdemEndederOpferzeigtejemandbeidenEphorendenVerratundseinenUrheberKinadonan.

Dieser VorspannistfürdieweitereGeschichte,dieSchilderungderBefragungdesVerräters und der gegen die Verschwörung getroffenen Gegenmaßnahmen,weitgehendüberflüssig”.Withregardtothelatter,Iwouldrathersaythatitistheintroductiontothestoryasawhole,writteninthebesttraditionofstory-telling.However,whetherornotthe mãntiw mentionedin3.3.11isthesameastheonementionedin3.3.4ishardtodetermine.



190 CHAPTERFOUR

previous paragraph. Besides providing information necessary for agoodunderstandingofitsmatrixclause(cataphoricgrounding),thesubclauseisusedtoseparatethetwoparagraphs(andatthesametimelinkthelattertotheformeranaphorically).

[18] Hellenica2.1.10-12t“ d' §piÒnti ¶tei [§p‹ ÉArxÊta m¢n §foreÊontow, êrxontow d' §nÉAyÆnaiw ÉAlej¤ou,] LÊsandrow éfikÒmenow efiw ÖEfeson metep°mcato ÉEteÒnikon §k X¤ou sÁn ta›w naus¤, ka‹ tåw êllawpãsaw sunÆyroisen, e‡ poÊ tiw ∑n, ka‹ taÊtaw t' §peskeÊaze ka‹

êllaw §n ÉAntãndrƒ §nauphge›to. § xrÆmata ætei: ı d' aÈt“ e‰pen ˜ti tå m¢n parå basil°vw énhlvm°nae‡h, ka‹ ¶ti ple¤v poll“, deiknÊvn ˜sa ßkastow t«n nauãrxvn¶xoi, ˜mvw d' ¶dvke. l , §p‹tåw triÆreiw trihrãrxouw §p°sthse ka‹ to›w naÊtaiw tÚn ÙfeilÒmenon misyÚn ép°dvke. pareskeuãzonto d¢ ka‹ ofl t«n ÉAyhna¤vnstrathgo‹ ~prÚw tÚ nautikÚn §n tª Sãmƒ.

Inthefollowingyear—Archytaswasnowephor,andAlexiasarchonatAthens—LysanderarrivedatEphesusandsentforEteonicustocomethitherfromChioswiththeships,andhealsogatheredtogetheralltheothershipsthatwereanywheretobefound,andheoccupiedhimself withrefittingthesevesselsandbuildingmoreatAntandrus.H

and asked for money; and Cyrus told him that the fundsprovided by the King had been spent, in fact much more besides,showinghimhowmucheachoftheadmiralshadreceived;neverthelesshedidgivehimmoney.L andappointedtoeachtriremeitscaptain,andpaidhissailorsthewagesthatwere duethem. And the Athenian generals also were getting their fleet inreadiness,atSamos.

Atthebeginningofthequotationanewepisodeisstartedatapointof thematic discontinuity in every respect. However, we are quicklyinformed: a temporal (re-)orientation is provided by t“ d' §piÒnti¶tei;LÊsandrow is re-introduced as the main protagonist; theparticipialclause éfikÒmenow efiw ÖEfesonisthereinordertoprovideaspatialorientationandexpressthefirstactionoftheaction-sequencethat has been started simultaneously; the rest of the sentence isdevotedtoactionsthatsetthethemeofthepassage:‘refittingthefleet’.Bytheendofthefirstsentenceweknoweverythingarecipientofapieceofnarrativeneedstoknowtokeeptrackofthesituation:time,place,participant,andaction.Thenextsentenceexpressesthenextstepinthesequence,assignaledby d°.ItturnsoutthatLysanderwenttoCyrusandaskedformoney(§lyΔn d¢ parå KËron xrÆmata



194 CHAPTERFOUR

Sãrdesi PrÒjenon ka‹ KËron m°llontaw ≥dh ırmçn tØn ênv ıdÒn,ka‹ sunestãyh KÊrƒ.

Xenophon read Proxenus’ letter and conferred with Socrates, theAthenian,abouttheproposedjourney;andSocrates,suspectingthathisbecoming a friend ofCyrus mightbe a cause for accusationagainstXenophononthepartoftheAtheniangovernment,forthereasonthatCyruswasthoughttohavegiventheLacedaemonianszealousaidintheir war against Athens, advised Xenophon to go to Delphi andconsultthegodwithregardtothisjourney.Xenophonw andaskedApollotowhomofthegodsheshouldsacrificeandprayinorderbest

andmostsuccessfullytoperformthejourneywhichhehadinmindandtomeetwithgoodfortuneandreturnhomeinsafety;andApolloinhisresponsetoldhimtowhatgodshemustsacrifice.

WhenhehadreturnedfromDelphi,Xenophonreportedtheoracleto Socrates; upon hearing about it Socrates found fault with himbecausehedidnotfirstputthequestionwhetheritwerebetterforhimtogoorstay,butdecidedforhimselfthathewastogoandthenaskedthegodastothebestwayofgoing.“However,”headded,“inasmuchasyoudidputthequestioninthatway,youmustdoallthatthegoddirected.”Xenophonaccordinglyofferedthesacrificestothegodsthat

Apollo’soracleprescribed,andsetsail;andheovertookProxenusandCyrusatSardisastheywereonthepointofbeginningtheupwardmarch,andwasintroducedtoCyrus.

XenophondiscussestheproposedjourneywithSocrates,whoadviseshim to go to Delphi and consult the god (§lyÒnta efiw DelfoÁw

énakoin«sai t“ ye“ );Xenophongoes (§lyΔn d' ); the (continuative)participial clause continues an event sequence set in motion bySocrates’s advice. His return is expressed in the subclause §pe‹ d¢

pãlin ∑lye,whichmarkstheboundarybetweenthetwophasesof thispassage:theirfirstencounter(Socrates’sadviceandXenophon’sresponse to that) and their second encounter (evaluation of Xenophon’s way of putting the question). The subclause §pe‹ d¢pãlin ∑lye,then,operatesontheleveloftextarticulation and theleveloftheRealWorldconstruction:itmarksaratherdeepcaesura,asthetextualboundaryisatthesametimeathematicboundary.Theparticipialclause §lyΔn d'merelyarticulatesatextualboundaryattheonsetofasub-sectionofthefirstthematicunit.

[21] Hellenica6.2.31per‹ d¢ tÚn Mnas¤ppou yãnaton §tÊgxanen Ãn t∞w Lakvnik∞w per‹tåw Sfag¤aw. e ka‹ parapleÊsawtÚ toË ÉAlfeioË stÒma ÍpÚ tÚn ÉIxyËn kaloÊmenon …rm¤sato. tª d'



CLAUSECOMBININGANDINFORMATIONPROCESSING 203

§pisfãjasyai.Secondly,andmoreimportantly,thenaturalorderof

events is suspended because it is an unusual situation: whereasdrawingadaggerandslayingoneselfwouldbenaturalinthecaseof,e.g., a soldier, a skhptoËxow is supposed to bear a staff—not apointedweapon.Placingtheparticiplebehinditsmatrixclausegivesfocus to the action, and facilitates the link with the sequel: thatArtapatesactuallyhadadaggerofgoldispresentedina gãr-clause,whichthenpavesthepathfordescribinghisotherpossessions.


metå d¢ taËta oÈ polla›w ≤m°raiw Ïsteron ∑lyen §j ÉAyhn«n Yumoxãrhw ¶xvn naËw Ùl¤gaw: ka‹ eÈyÁw §naumãxhsan aÔyiwLakedaimÒnioi ka‹ ÉAyhna›oi, §n¤khsan d¢ LakedaimÒnioi

. met' Ùl¤gon d¢ toÊtvn DvrieÁw ıDiagÒrou §k ÑRÒdou efiw ÑEllÆsponton efis°plei érxom°nou xeim«nowt°ttarsi ka‹ d°ka naus‹n ëma ≤m°r&.

Afterthis,notmanydayslater,ThymocharescamefromAthenswithafew ships; and thereupon the Lacedaemonians and the Atheniansfoughtanothernavalbattle,andtheLacedaemonianswerevictorious,

. Shortly after this, at thebeginning ofthewinter,Dorieus,theson ofDiagoras, sailed into theHellespontfromRhodeswithfourteenships,arrivingatdaybreak.

If the genitive absolute had preceded its matrix clause here, theinformationthatAgesandridaswastheleaderwouldbeprocessedbythe reader/hearer before he knew the Lacedaemonians werevictorious; this order of clauses is possible only when there issomethingintheprecedingdiscoursethatcallsfortheinformation

thatAgesandridaswasinchargetobeprocessedfirst—whichisnotthecasehere.Thematrixclauseofthepostposedgenitiveabsoluteimmediatelyinformsuswhowerevictoriousinthebattlementionedintheprecedingsentence.Thehistorianthenaddsaclauseinordertoinformusabouttheirleader.

Inexamples[1]and[2]thereisnothinginthecontextthatwouldleadonetoa‘causal’interpretationofthepostposedparticiple.Inthesequel, I shall concern myself with postposed §pe¤-/§peidÆ-clausesand participial clauses which may receive a ‘causal’ interpretation.

Onemayaskwhathappenswhenthereader/hearerarrivesattheendofaclauseandispresentedwithasubclausethatispostposedtoit.Hewillreconsiderthecontentofthematrixclause,withtheaidof theinformationcontainedintheaddedclause,beitasubclauseora




especially the fact that it isoÈd¢n yaumastÒn

, which triggers a justification.Therewith,the§peidÆ-clauseisamotivatingclause.3

[4] Anabasis1.3.9 êndrew strati«tai, tå m¢n dØ KÊrou d∞lon ˜ti oÏtvw ¶xei prÚw ≤mçwÀsper tå ≤m°tera prÚw §ke›non: o te går ≤

, § , oÎte §ke›now ¶ti≤m›n misyodÒthw.

(Clearchus:)“Fellow-soldiers,itisclearthattherelationofCyrustousis precisely the same as ours to him; that is, w

,i ,andlikewiseheisnolongerourpaymaster”.

Thefirststatementofthespeechmaybesomewhatenigmaticforthestrati«tai,andthereforeitisexplainedinthefollowinggãr-clause.ThefirstmemberoftheoÎte…oÎte-pair,inwhichitisstatedthatClearchusandhissoldiersarenolongerCyrus’ssoldiersisfollowedbyan§pe¤-clause.Clearchus’saudiencedoesconsistofsoldiers;infact,they are addressed as such at the onset of the speech ( êndrew

strati«tai ). The message is that they are no longer Cyrus’s soldiers,inasmuchastheydonotfollowhim.Thelogicbehindthissequencecanbedescribedasfollows:(1)a

soldierisapersonwhofollowstheleader;(2)wedonotfollowCyrus;ergo (3)wearenothissoldiers.Premiss(1)isleftunexpressed;premiss(2)ispresentedasafactualstatementinthepostposed §pe¤-clause,whileinference(3)isthepointmadeinthematrixclause.Moreover,itistobenotedthatthismatrixclauseisnegatedandcontainstheword¶ti:wearenolonger hissoldiers.Thisisimportantinthatthechangeof

situationhastriggeredClearchus’sspeechinthefirstplace.Further,andmoreimportantlyforourinvestigation,itisthe changeofsituationthat has triggered the §pe¤-clause.Thecontentofthe §pe¤-clause

justifies the claim made by Clearchus, whose rhetorical aim in thispassageitistotryandpersuadehismenthattheydonothavetoobeyCyrus any longer. The speaker thus invites his audience tocomprehendtheprecedingutteranceinthelightoftheRealWorld

———3 Motivatingclauses have oftenbeendisregarded inthe descriptionof AncientGreeksyntax:“Lespropositionsdemotifnesontgénéralementpasmentionnéesparlesmanuelsscolaires,etquantauxgrammairesscientifiques,ellesnelessignalentquerarement,defaçonnonsystématique,etsouvantsansleurdonneruneappellationspéciale”(Muchnova,1991:135;seealsoRijksbaron,1976).



206 CHAPTERFIVE

situationthat,accordingtohim,istrue:‘<whateverwedo,whatisrelevantinthisconnectionisthat(ge )>wedonotfollowhim’.

[5] Anabasis5.6.35-36§nteËyen §pe‹ ¶gnvsan ofl ÑHrakle«tai ˜ti §kple›n dedogm°non e‡hka‹ Jenof«n aÈtÚw §pechfikΔw e‡h, tå m¢n plo›a p°mpousi, tå d¢ xrÆmata ì Íp°sxonto Timas¤vni ka‹ Y≈raki §ceusm°noi ∑san [t∞w misyoforçw]. §ntaËya d¢ §kpeplhgm°noi ∑san ka‹ §ded¤esan tØnstratiån ofl tØn misyoforån Ípesxhm°noi. paralabÒntew oÔn otoika‹ toÁw êllouw strathgoÁw oÂw énekeko¤nvnto ì prÒsyen ¶pratton,pãntew d' ∑san plØn N°vnow toË ÉAsina¤ou, ˘w XeirisÒfƒ

ÍpestratÆgei, Xeir¤sofow d¢ oÎpv par∞n, ¶rxontai prÚwJenof«nta, ka‹ l°gousin ˜ti metam°loi aÈto›w, ka‹ d, § , ka‹

katasxe›n tØn Fasian«n x≈ran.

Afterthat,whentheHeracleotslearnedthatithadbeendecidedtosailaway, and thatXenophon himself had put the question to the vote,they did send the ships, but in the matter of the money they hadpromisedtoTimasionandThoraxtheyturnedouttobedeceivers.Consequentlythemenwhohadpromisedthepaywerepanic-stricken,

andstoodinfearofthearmy.Theythereforetookwiththemtheothergenerals to whom they had communicated their earlierdoings—namely,allthegeneralsexceptNeontheAsinaean,whowasactingaslieutenantforCheirisophusbecauseCheirisophushadnotyetreturned—and came to Xenophon, with the message that they hadchanged their minds and t ,

, and seize the land of thePhasians.

The words of the speakers are presented by optatives in the oratio

obliqua( metam°loi, doko¤h ), but the mood of the finite verb of the§pe¤-clauseisunaffected:itisstillanindicative.Followingonanon-factual, evaluating, and, for the addressee, possibly debatableutterance( doko¤h krãtiston e‰nai ),the§pe¤-clausepresentsafactualstatement:theshipsarethere(cf.tå m¢n plo›a p°mpousiinline2);this is literally what the speakers said. In this instance of indirectspeech, then, the addressee Xenophon should comprehend thespeaker’sevaluatingclaim krãtiston e‰nai ple›n efiw Fçsininviewof the Real World situation presented in the postposed subclause.

Heretoo,thepostposedsubclausemotivates(partof)thepreceding utterance.




The question of whether a market was provided is therefore adiscourse topic, and is referred to throughout:pare›xon går ≤m›n

égorãn…§pe‹ égorån oÈ pare›xon…§pe‹ égorån o·an §dÊnantopare›xon . In the sentence ka‹ KardoÊxouw ka‹ TaÒxouw ka‹Xalda¤ouw ka¤per basil°vw oÈx ÍphkÒouw ˆntaw ˜mvw ka‹ mãla

foberoÁw ˆntaw polem¤ouw §kthsãmeya diå tÚ énãgkhn e‰nailambãnein tå §pitÆdeia, §pe‹ égorån oÈ pare›xon,weareinformedthat the Greeks have made enemies of the Carduchians, theTaochiansandChaldaeans,althoughthesepeoplewerenotsubjects

oftheKingandeventhoughtheywereformidable.Thiswascausedbythenecessityoftaking provisions( diå tÚ énãgkhn e‰nai lambãneintå §pitÆdeia )—instead of the alternative, buying them(»n°omai,cf.»noÊmenoi ).Inhispresentationofevents thespeaker stresses the factthattheydidnotactoutofÏbriw.

The postposed subclause §pe‹ égorån oÈ pare›xon does notpresent a ‘cause’ of the content of the preceding clause; it is amotivatingclauseusedbythespeakertojustifythepreceding,possiblydebatable, evaluating utterance: ‘we made enemies of theCarduchians,theTaochiansandChaldaeansoutofthenecessityof takingprovisions, andIcanactuallyclaimthatitwasnecessarytotakethem, giventherealitysituation:theydidnotprovideamarket’.The§pe¤-clausepresents the relevant data from the Real World; compare thefollowing§pe‹ égorån o·an §dÊnanto pare›xonasthemotivefortheGreeks’friendlybehaviortowardsthe—barbarian—Macronians.

The evaluating utterance tÚ énãgkhn e‰nai resembles the,doubtlesslydebatable,evaluation±nagkãsyhn di≈keinofChapter3,

Section2,example[25],whichisreproducedhereas[8]:[8] Anabasis3.3.11-13¶nya dØ pãlin éyum¤a ∑n. ka‹ Xeir¤sofow ka‹ ofl presbÊtatoi t«nstrathg«n Jenof«nta ºti«nto ˜ti §d¤vken épÚ t∞w fãlaggow ka‹ aÈtÒw te §kindÊneue ka‹ toÁw polem¤ouw oÈd¢n mçllon §dÊnato blãptein. ékoÊsaw d¢ Jenof«n ¶legen ˜ti Ùry«w afiti“nto ka‹ aÈtÚtÚ ¶rgon aÈto›w marturo¤h. “ ”, ¶fh, “

, §. §peidØ d¢

§di≈komen, élhy∞, ¶fh, Íme›w l°gete: kak«w m¢n går poie›n oÈd¢n mçllon §dunãmeya toÁw polem¤ouw, énexvroËmen d¢ pagxal°pvw…”

Hereagaintherewasdespondency.AndCheirisophusandtheeldestof thegeneralsfoundfaultwithXenophonforleavingthemainbodyof




principleexplains, afortiori ,thepredominanceofparticiplesshowing case-agreement;thesuspensionofagreementisthereforeasignof discourse complexity. In this respect, it is significant that most‘irregular’genitiveabsoluteconstructionsarepreposedtotheirmatrixclause and, within this category, mostly occupy the sentence-initialposition.

Discoursecomplexity isofavaryingnature.Syntacticcomplexity,or,intermsofinformationprocessing,the‘processingload’ofaclausecombination,isamongthefirstfactorsinvolved.Syntacticcomplexity

ensues, e.g., when another clause is dependent on the participialclause.Textualcomplexityoccurswhenagreatamountofentirelynewinformationhastobeprocessedinon-goingdiscourse.ThefactthatsomecasesinAncientGreekarelesssuitabletobeusedinsuchsituations because of the great demands they make on theconcentrationofthereader/hearermayleadthespeaker/narratortouse a genitive absolute where, syntactically, a conjunct participlewould have been possible. This predictably occurs when theconstituentwithwhichtheparticlecouldhavebeenjoinedisadative:in information processing, the dative case is too marked to be theanchorofasyntacticallyorpragmaticallycomplexparticipialclause.

Whilewewouldexpecttofindsyntacticshiftsespeciallyinspokendiscourse,13 it should be noted that even in the case of writtendiscourse the genitive absolute is often preferred over a conjunctparticiple to facilitate participant-tracking and topic management.Thegenitiveabsolute,whichusuallyhasitsown(expressed)subjectandthereforeisaltogetheroutsidethemainconstruction,offersthe

speaker the opportunity to present a piece of information withoutinterferingwiththeconstructionofthesentenceasawhole.

2.1 The‘Regular’ConjunctParticiple

If,forinstance,aparticipanthasalreadybeenfirmlyestablishedasatopicintheprecedingdiscourse,thisparticipantisreferredbacktobymeansofaformofthepronoun aÈtÒw(orotow )inthecasefittingthemain construction; a conjunct participle can be joined with the

———13 Slings (1994 a: 420) proposes the following hierarchy of cases for an oral

grammarofHomer,inwhichshiftsoccurfromthemostmarkedcaseontherightsideofthescaletotheleft:

nominative<?vocative<accusative<genitive<dative




thematiccontinuityofthecharacter-orientedparagraph,despitetheshiftinthetemporalsituation.15Followingonthegenitiveabsolute,thefirmlyestablished,continuoustopicofthepreceding DUispickedupbythepronoun aÈto›w, with which the present stem conjunctparticipialclause poreuom°noiw parå basil°aisjoined,providingaspationalorientation.Compare[20]:

[20] Anabasis3.4.1-2 me¤nantew d¢ taÊthn tØn ≤m°ran tª êll˙ §poreÊonto prƒa¤teron énastãntew: xarãdran går ¶dei aÈtoÁw diab∞nai §f' √ §foboËnto mØ

§piyo›nto aÈto›w diaba¤nousin ofl pol°mioi. dpãlin f Miyradãthw, ¶xvn flpp°aw xil¤ouw,tojÒtaw d¢ ka‹ sfendonÆtaw efiw tetrakisxil¤ouw:

That day they remained quiet, and the next morning they setforth,—theyroseearlierthanusual;fortherewasagorgetheyhadtocross,andtheywereafraidthattheenemymightattackthemastheywerecrossing.Itwaso that Mithradates

again,accompaniedbyathousandhorsemenandaboutfourthousandbowmenandslingers.

Thedativepronoun aÈto›w with which the perfect stem conjunctparticiple diabebhkÒsiisjoined,functionsasanoptionalargumentto

fa¤netai;thecontinuoustopicispickedupafterabackgroundedgãr-clauseinathematicallycontinuouspassage;cf.thelexicaloverlapin diab∞nai … diaba¤nousin… diabebhkÒsi. Narrative time ispropelledforwardtothemomentwhentheyhadmetthedifficultyof crossing the gorge without the enemy showing up, who aresubsequentlybroughttothestage.

Whenparticipanttrackingisrelativelyeasy,markedcasesinGreekcan function as the anchor of a conjunct participle. While, forinstance, the participle of éfikn°omaiisusedpredominantlyasaconjunct participle in the nominative case relocating a firmlyintroducedactiveparticipant,theaccusativecaseispossibleincertaincontexts;see[21]:

[21] Hellenica1.1.9 metå d¢ taËta Tissaf°rnhw ∑lyen efiw ÑEllÆsponton: é

ÉAlkibiãdhn j°niã te ka‹ d«ra

———15 Since,however,thepassageischaracter-oriented,nottemporallyorganized,it

isnotdiscontinuousenoughwithrespecttotimetowarrantasubclausetosignalacontent-orientedboundaryatthepointwherethereisneedtoindicatethetransitiontoanewyear.



230 CHAPTERFIVE

Oneofthemostinterestingfeaturesofthegenitiveabsoluteisthatitcontainsthenon-specifictemporaladverbial pote,whichvaguelyindicatesanewtemporalsettinginreferringtoamomentintimeinsidetheover-alltemporalsettingindicatedby diexe¤maze;assuch,ashiftintemporalsettingisestablishedfromthe general (‘thewinter’)tothespecific(‘ononeoccasionduringthatwinter’).

Inspiteofthearticulationofanew DU andtheslighttemporalshift,thesituationisotherwisehighlycontinuous:as potetakescareof the temporal organization, the present tense stem participle

(lambanÒntvn )itselfdoesnotpropelforwardnarrativetime;rather,ittakes up the action of getting provisions already introduced in theconjunct participial clause ka‹ aÈtÒyen ka‹ sÁn pronoma›w tå§pitÆdeia tª stratiò lambãnvn, postposed to §ntaËya m¢n dØ

diexe¤maze. It thus secures coherence at a point of segmentationthroughlinkagevialexicaloverlap.Thepresentstemparticiple,whichis off the main narrative line, simultaneously serves an orienting functiontowardsthesequel.

BecauseofthehighthematiccontinuityofthenewDUwithrespecttotheprecedingDU,asubclausewouldbeoutofplaceatthispoint:itwouldhaveindicatedathematicboundaryattheleveloftheRealWorldconstructionwhichisuncalledforhere.Also,onemaywonderwhetheranarratorwouldconsiderindicatingaRealWorldrelationbetweenthefactthatthesoldiersweregettingtheirprovisionsandthefact that Pharnabazus showed up. We may hypothesize that theirgetting provisions indisdainfulandcarelessfashion made Pharnabazuscome upon them, but the lexical meaning of §p°tuxen(‘tofallin

with’,‘meetwith’,LSJ)seemstomakethisinterpretationimpossible;rather,Pharnabazusmetthemwhentheywerescatteredacrosstheplain because of this way of conduct. Factors concerning both thepresentation ofRealWorldrelationsand the articulationofthe textareoperativeonthechoiceofaparticipialclausehere,totheextentthatthepossibilityofasubclauseisexcluded.

While a smooth change-over between the two adjacent DU’sisestablished,atthesametimethereisaslightshiftasfarasthetopicalparticipantsisconcerned.Uptothegenitiveabsolute,Agesilausisthedominant discourse topic. Then, with the genitive absolute, thesoldiersofhisarmybecomethenew,inferrabletopic.Thereisnomajor discontinuity in the cast of participants, as the soldiers werealreadypresentintheprecedingpostposedconjunctparticipialclause



232 CHAPTERFIVE

Koriny¤vn o· te êlloi pãntew sÊmmaxoi ka‹ ofl ÉAyhna›oi., eÈyÁw m¢n

Lepreçtai épostãntew t«n ÉHle¤vn p , eÈyÁw d¢ Mak¤stioi, §xÒmenoi d' ÉEpitalie›w. d

Letr›noi ka‹ ÉAmf¤doloi ka‹ Margane›w.§k d¢ toÊtou §lyΔn efiw ÉOlump¤an ¶yue t“ Di‹ t“ ÉOlump¤ƒ:kvlÊein d¢ oÈde‹w ¶ti §peirçto. yÊsaw d¢ prÚw tÚ êstu §poreÊeto,kÒptvn ka‹ kãvn tØn x≈ran, ka‹ Íp°rpolla m¢n ktÆnh, Íp°rpolla d¢ éndrãpoda ≤l¤sketo §k t∞w x≈raw:

Inthecourseoftheyeartheephorsagaincalledoutthebanagainst

Elis,and with the exception ofthe Boeotians and the Corinthiansalltheallies,includingtheAthenians,tookpartwithAgisinthecampaign.

,andtheLepreansatoncerevoltedfromtheEleansandcameovertohim,theMacistianslikewiseatonce,and after them the Epitalians. ,whentheLetrinians, Amphidolians, and Marganians c .Thereupon he went to Olympia and offered sacrifices to OlympianZeus, and this time no one undertook to prevent him. After hissacrificeshemarcheduponthecityofElis,layingthelandwastewithaxeandfireashewent,andvastnumbersofcattleandvastnumbersof

slaveswerecapturedinthecountry.

Theverb prosxvr°v requiresa dativecomplement,thisslotbeing filled by the pronoun aÈt“ the first time the verb occurs in thispassage. By means of the pronoun the narrator refers back to thesubject of the preceding genitive absolute (toË ÖAgidow ). With§mbalÒntow d¢ toË ÖAgidow di' AÈl«now,Agis,theleaderoftheexpedition,becomesanactivediscourseparticipant;however,hehadalready been present in the discourse, but not as an agent

(sunestrateÊonto t“ ÖAgidi ). Typically, new topics are introducedbyanominalconstituentinthenominativecase(thenameprecededbythearticle),usuallyaccompaniedby d°.Afterfirstbeingmentionedinthedative,Agisstillhastobeestablishedasatopic;althoughAgisperformsasyntacticfunctionintheleadingclause,agenitiveabsoluteis chosen; the alleged sentence-based demand of case-agreement issuspended in order to introduce Agis as an active discourseparticipant.Thesecondtimetheverbprosxvr°visused,Agishasbeenfirmlyestablishedasatopic,andaconjunctparticipleinthedativecase,evenwithoutitsnominalheadbeingexpressed,canbeused ( diaba¤nonti … tÚn potamÚn prosex≈roun ). From theviewpoint of information flow, the distribution of the aorist andpresent stem, in that order (prosex≈rhsan…prosex≈roun:plus




onhearingofthistheCorinthiansspeedilysentadequategarrisonsof theirowntoeveryplacewhereAthenianswereonguardandtoldthelattertodepart,sayingthattheynolongerhadanyneedofgarrisons.

Schwyzer(ibid.:101)notes:

Beispielemit einemgen. abs.,dessenParticipium von einemverbumdicendikommt,bietetauchXenophon;dochwirdindiesenandersalsbei Herodot II162 geradedieses Participium späterals selbständigesverbumdicendiaufgenommen(…).18Einkorrekter,aberlangerSatzwäredurchWeglassungvon ¶fhunddieFortsetzung ékoÊsantew ofl

Kor¤nyioi ktl. noch

t«n ÉAy. zu bilden gewesen; durch die

VerselbständigungdesGliedesmitto›w m°ntoimittels¶fhentstehteinAnakoluth;nachherwirdmit einemneuen Satze( ék. d¢ taËta ktl. )fortgefahren.StellenmitverbadicendisindauchXen.hist.Gr.IV8,9undCyr.<VI1,37>mitgen.abs.nebenNom.,Cyr.<I6,14>nebenAkk. Ausserdem vgl. Xen. Cyr. I 5, 5 dejam°nou toË KÊrou ofl bouleÊontew gera¤teroi aflroËntai aÈtÚn (scil.KÊron ) êrxonta.

In my opinion, the point is not that it would havebeen possibletocompose“einkorrekter(…)Satz”;apartfromthesuggestionofferedbySchwyzer,thealternativeherewouldhavebeenusingafinitemainclause(*Dhmot¤vn e‰pen,alsowithout¶fh ).Thequestion iswhythegenitiveabsolutehasbeenpreferred,andhowthecontinuationwith¶fh should be explained. Sentence-grammar fails to anwer thesequestions.

InthepassageprecedingDemotion’sstatementintheAssembly,weweretoldabouttheaccomplishmentofthealliancebetweentheAtheniansandtheArcadians;thispartofthetextisclosedoffwiththesentencekéke›now m¢n oÏtvw époynπskei, ≤ m°ntoi summax¤a ˆntvw

§pera¤neto,whichshowsallthecharacteristicsofaclosing-line:closeconnectiontotheprecedingasindicatedbyka¤, m°n… m°ntoiclosing offthecurrentepisode,anaphora(§ke›now,oÏtvw )andrecapitulation( époynπskei ).AfterLycomedeshasbeenabandonedasatopic,anewparticipantisintroduced( d° ),andat the same time the camera shiftstotheAthenianAssembly(efipÒntow d¢ Dhmot¤vnow §n t“ dÆmƒ t«nÉAyhna¤vn ): another exampleof slightdiscontinuity in an otherwise

———

18 Here thesentenceefipÒntow d¢ Dhmot¤vnow… deÒmenoi frour«nfollows,mistakenlycitedas“hist.Gr.VIII4,4”.Thefollowingcorrectionshavealsobeenmade:ForSchwyzer’scitationswithouttheGreek:“Cyr.VI3,7”and“Cyr.VI6,14”,Iread‘Cyr.VI1,37’and‘Cyr.I6,14’,respectively(cf.Kühner-Gerth,1898-1904:II,110-111);Schwyzer’s“Cyr.I41,20”(taËt' efipÒntow aÈtoË ¶doj° ti l°geint“ ÉAstuãgei )iscorrectedas‘Cyr.I4,20’.



238 CHAPTERFIVE

turnedabout,andsailingaway,cametoanchoratPhoenicusintheislandofCythera.

ThosewhoheldpossessionofthecityoftheCytheriansabandonedtheirwalls through fearofbeing capturedbystorm. ThesepeopleheallowedtodeparttoLaconiaunderatruce;hehimselfrepairedthewallof the Cytherians and left in Cythera a garrison of his own andNicophemus,anAthenian,asgovernor.ThereuponhesailedtotheIsthmusofCorinthandexhortedthealliestocarryonthewarzealouslyand show themselves men faithful to the King, and leftthem all themoneythathehadandsailedoffhomeward.NowC thatifhe(sc.Pharnabazus)wouldallowhimtohavethefleet,hewouldmaintain

itbycontributionsfromtheislandsandwouldputinatAthensandaidthe Athenians in rebuilding their long walls and the wall aroundPiraeus;h thatheknewnothingcouldbeaheavierblowtotheLacedaemoniansthanthis.“Andbythisact,therefore”,h ,“youwill have conferred a favor upon the Athenians and have takenvengeance uponthe Lacedaemonians,inasmuch asyou willundo forthemthedeedforwhoseaccomplishmenttheyunderwentthemosttoilandtrouble”.Pharnabazus,uponhearingthis,eagerlydispatchedhimtoAthensandgavehimadditionalmoneyfortherebuildingofthewalls.

Again, a genitive absolute has been chosen for reasons of topicmanagement.Cononhadbeenanactivediscourseparticipant,buthad disappeared as such. He was last mentioned in 4.8.7 inconnection with Pharnabazus (sumplhr≈saw ka‹ …prosmisyvsãmenow ¶pleusen ı FarnãbazÒw te ka‹ ı KÒnvn met'

aÈtoË;notealsothethirdperson singular verbforms).ThediscourseremainsconcernedwithPharnabazus,who,inresponsetoConon’swords,sendshimtoAthens(ı d¢ Farnãbazow ékoÊsaw taËta ).The

factthatCononhadtobeproperlyre-introduced(textualcomplexity),before being send off 21 triggers the use of a genitive absoluteconstruction,andtherelativelylargeamountofinformationthatisprovided by the preposed, sentence-initially placed subordinateconstruction—syntactic complexity ensuing from the dependentclausepresentingConon’sstatement,thistimeincludingashiftfromindirecttodirectdiscourse—accountsforthe‘slippage’toarelativelylesscomplexstructurewith¶fh(twice).

———21 Conondoes not disappear immediatelyfromview: hestill has toperforma

syntacticfunctioninthesequel(ı …Farnãbazow… ép°steilen aÈtÒn );comparethegenitiveabsoluteusedinChapter2,Section6,[14]anddiscussion.



242 CHAPTERFIVE

ka‹ prosel°syai diakos¤ouw t«n ımot¤mvn, t«n d' aÔ diakos¤vn•kãstƒ t°ttaraw ¶dvkan prosel°syai ka‹ toÊtouw §k t«n ımot¤mvn:g¤gnontai m¢n dØ otoi x¤lioi: t«n d' aÔ xil¤vn toÊtvn •kãstƒ¶tajan §k toË dÆmou t«n Pers«n d°ka m¢n peltaståw prosel°syai, d°ka d¢ sfendonÆtaw, d°ka d¢ tojÒtaw: ka‹ oÏtvw §g°nonto mÊrioi m¢n tojÒtai, mÊrioi d¢ peltasta¤, mÊrioi d¢ sfendon∞tai: xvr‹w d¢toÊtvn ofl x¤lioi Íp∞rxon. tosaÊth m¢n dØ stratiå t“ KÊrƒ §dÒyh.§pe‹ d¢ Ωr°yh tãxista, ≥rxeto pr«ton épÚ t«n ye«n:kallierhsãmenow d¢ tÒte pros˙re›to toÁw diakos¤ouw: §pe‹ d¢prose¤lonto ka‹ otoi dØ toÁw t°ttaraw ßkastoi, sun°lejen aÈtoÁwka‹ e‰pe tÒte pr«ton §n aÈto›w tãde.

Cyaxares—heheardoftheplotandofthewarlikepreparationsofthenations all ied against him—without delay made what counterpreparations he could himself and also sent to Persia both to thegeneralassemblyandtohisbrother-in-law,Cambyses,whowaskingof Persia.AndhesentwordtoCyrus,too,askinghimtotrytocomeascommander of the men, in case the Persian state should send anytroops.ForCyrushadbythistimecompletedhistenyearsamongthe

youthsalsoandwasnowintheclassofmaturemen.S,andtheeldersincouncilchoseh commanderofthe

expeditiontoMedia.Andtheyfurtherpermittedhimtochoosetwohundredpeerstoaccompanyhim,andtoeachoneofthetwohundredpeersinturntheygaveauthoritytochoosefourmore,thesealsofromthe peers. Thatmade a thousand. And each one of the thousand intheirturntheybadechooseinadditionfromthecommonpeopleofthePersianstentargeteers,tenslingers,andtenbowmen.Thatmadetenthousand bowmen, ten thousand targeteers, and ten thousandslingers—notcountingtheoriginalthousand.SolargewasthearmygiventoCyrus.Nowassoonashewaschosen,hisfirstactwastoconsultthegods;andnottillhehadsacrificedandtheomenswere

propitious,didheproceedtochoosehistwohundredmen.Thenthesealso chose each his four, and he called them all together and thenaddressedthemforthefirsttimeasfollows.

The genitive absolute is preposed and occupies the sentence-initialposition. Its subject (toË KÊrou ) is referred back to later in thesentencebytheanaphoricpronounintheaccusativecase aÈtÒn.25

Thesequenceconsistingofsendingtheinvitationandacceptingitisinterrupted by a gãr-clause informing the reader/hearer of the

positionthatCyrushadreachedatthetime.AsoÏtv dÆindicates,thereader/hearershouldbereadyforcontinuationofthenarrativeline

———25 Thepreposedparticipialclausecouldhavebeena conjunctparticipialclause;

forthepossibility,see[21]above.




whereithadbeenbrokenoff,nowthathehasbeeninformedbytheintervening gãr -clause. A conjunct participle would suffice;26 buthere,thisnextsteponthenarrativelineisbetterpresentedinaclausethat is syntactically detached from its matrix clause: a genitiveabsolute.

Cyrus,althoughalreadypresentinthediscourse,becomesan activediscourseparticipantonlyatthepointwherethegenitiveabsoluteisused:thefactthatheacceptedtheinvitationiscrucialinthatfromthat point onwards, Cyaxares, with whom the citation starts, is

completely disregarded, and the text continues with Cyrus: he ischosen commander of the expedition to Media ( aflroËntai aÈtÒn );subsequently, information off the main narrative line is presented,startingwith¶dosan d¢ aÈt“ (sc.Cyrus,whoisagainreferredtobyananaphoricpronoun),uptoandincludingtheclosinglinetosaÊth m¢n

dØ stratiå t“ KÊrƒ §dÒyh. Then the sentence-initially placedsubclause §pe‹ d¢ Ωr°yh tãxistatypicallytakescareofthethematicorganizationofthetext:itrestatesCyrus’election,picksupthepointinthenarrativethathadalreadybeenreachedbeforethedigression,andtherewithlaunchesthenewsequenceparagraph27inwhichCyrusis the continuous topic, without further reference (Ωr°yh,≥rxeto,kallierhsãmenow … pros˙re›to ),uptoandincluding Cyrus’ speech(sun°lejen … ka‹ e‰pe ).Aswehave seen before, agenitiveabsoluteconstruction is used where a conjunct participle would have beenpossiblefromasyntacticalpointofviewinordertofirmlyestablishaparticipantasanactivediscoursetopic.

According to Schwyzer, the emphatic function of the genitive

absoluteisespeciallyclearwhenitoccupiesapositionearlyinthesentenceandisfollowedimmediately,oralmostimmediately,bythe(possiblyunexpressed)nominalconstituentwith whichitcouldhavebeenjoined.28TheexamplesfromtheworksofXenophonwillnowbereconsideredwithinawidercontext.29

———26 Compare[20]above.27 Cf.Chapter4,Section4.2,[7].28 Ibid.:102:‘DeutlichistdieAbsichtderHeraushebung,wenndergen.abs.und

dasBezugswort,dasauchineinerVerbalformenthaltenseinkann,nurdurcheinWortoderauchgarnichtgetrenntsind’.29 Ibid.:102:‘XenophonhatCyr.<I4,20>taËt' efipÒntow aÈtoË ¶doj° (der

Sprecher) ti l°gein t“ ÉAstuãgeioderI4,2 ésyenÆsantow aÈtoË (despãppow )oÈd°pote ép°leipe tÚn pãppon, an. II 6, 3 ofl ¶foroi ≥dh ¶jv ˆntow [aÈtoË]

épostr°fein aÈtÚn §peir«nto; mit doppeltem gen. abs. Xen. An. V 2, 24



244 CHAPTERFIVE

[29] Cyropaedia1.4.1-2toiaËta m¢n dØ pollå §lãlei ı KËrow: t°low d¢ ≤ m¢n mÆthr ép∞lye,KËrow d¢ kat°meine ka‹ aÈtoË §tr°feto. ka‹ taxÁ m¢n to›w≤liki≈taiw sunek°krato Àste ofike¤vw diake›syai, taxÁ d¢ toÁwpat°raw aÈt«n énÆrthto, prosiΔn ka‹ ¶ndhlow Ãn ˜ti ±spãzeto aÈt«n toÁw ufle›w, Àste e‡ ti toË basil°vw d°ointo, toÁw pa›daw§k°leuon toË KÊrou de›syai diaprãjasyai sf¤sin, ı d¢ KËrow, ˜ ti d°ointo aÈtoË ofl pa›dew, diå tØn filanyrvp¤an ka‹ filotim¤an per‹pantÚw §poie›to diaprãttesyai, ka‹ ı ÉAstuãghw d¢ ˜ ti d°oito aÈtoË ı KËrow oÈd¢n §dÊnato ént°xein mØ oÈ xar¤zesyai. ka‹ går

oÈd°pote ép°leipe t oÈd¢kla¤vn pot¢ §paÊeto, éllå d∞low ∑n pçsin ˜ti Íperefobe›to mÆ oflı pãppow époyãn˙: ka‹ går §k nuktÚw e‡ tinow d°oito ÉAstuãghw,pr«tow ºsyãneto KËrow ka‹ pãntvn éoknÒtata énepÆdaÍphretÆsvn ˜ ti o‡oito xarie›syai, Àste pantãpasin énektÆsatotÚn ÉAstuãghn.

InthiswayCyrusoftenchatteredon.Atlasthismotherwentaway,butCyrusremainedbehindandgrewupinMedia.Soonhehadbecomesointimatelyassociatedwith otherboysofhisown yearsthathewason

easytermswiththem.Andsoonhehadwontheirfather’sheartsbyvisiting them and showing that he loved their sons; so that, if theydesiredanyfavoroftheking,theybadetheirsonsaskCyrustosecureitforthem.AndCyrus,becauseofhiskindnessofheartandhisdesireforpopularity, made every effort to secure for the boys whatever theyasked. And Astyages couldnot refuse any favor that Cyrus asked of him.Forw ,Cyrusneverlefth norceasedtoweepbutplainlyshowedtoallthathegreatlyfearedthathisgrandfathermightdie.Forevenatnight,ifAstyageswantedanything,Cyruswasthefirsttodiscoveritandwithgreateralacritythananyone

elsehewouldjumpuptoperformwhateverservicehethoughtwouldgivehimpleasure,sothathewonAstyages’heartcompletely.

Thepeculiarityofthisexampleresidesinthefactthatthesubjectof thegenitiveabsoluteistheanaphoricpronoun aÈtÒw,tobespecifiedlateroninthesentenceintheaccusative(tÚn pãppon ). Because thesubjectofthegenitiveabsoluteistheanaphoricpronoun aÈtÒw,oneof the two active discourse participants is continued as a discoursetopic.Attheonsetofthissentence, ésyenÆsantow aÈtoËmightbeconsidered ambiguous: is the referent of the anaphoric pronoun

——— maxom°nvn d' aÈt«n ka‹ époroum°nvn ye«n tiw aÈto›w mhxanØn svthr¤aw d¤dvsin’.TheoccurrenceofthegenitiveabsoluteinCyropaedia1.4.20hasalreadybeendiscussedinChapter2,Section6,NMP6,[15].




aÈtoËAstyages—thesubjectoftheprecedingmainclause—orCyrus,

whowaslastmentionedbyname?AsthesequelshowsAstyagesismeant.The subjectofthepreceding mainclause isusuallythemostlikely candidate, and here it has been marked linguistically (ka‹ ıÉAstuãghw d°;note d° ). As a corollary, Cyrus is continued as thesubjectoftheupcomingmainverbs( ép°leipe… oÈd°…§paÊeto,

éllã…∑n:zero-anafora).Remains the fact that thegenitive absolute is‘irregular’, sinceits

subjectdoesperformasyntacticfunctioninthematrixclause.When

thecontextinwhichthesentenceunderconsiderationoccursistakeninto consideration, it is plain why the demand of case-agreementrequiring a conjunct participle is disregarded here. The genitiveabsoluteisusedattheonsetofanembeddedparagraphexemplifying Cyrus’devotiontowardsAtyageswhichservesasanexplanation(ka‹gãr ) for the preceding statement that Astyages could not refuse anyfavor that Cyrus asked of him. At this point, the text has to bereorganised. The relatively less strongly desententialized genitiveabsolutehastheadvantageofhavingitsownsubjectexpressedandtherewithofferstheopportunityofcontinuingAstyagesasadiscoursetopicacrossthisdiscourseboundary.

[30] Anabasis2.6.2-4ka‹ går dØ ßvw m¢n pÒlemow ∑n to›w Lakedaimon¤oiw prÚw toÁwÉAyhna¤ouw par°menen, §peidØ d¢ efirÆnh §g°neto, pe¤saw tØn aÍtoËpÒlin …w ofl Yròkew édikoËsi toÁw ÜEllhnaw ka‹ diaprajãmenow …w§dÊnato parå t«n §fÒrvn §j°plei …w polemÆsvn to›w Íp¢rXerronÆsou ka‹ Per¤nyou Yr&j¤n. §pe‹ d¢ metagnÒntew pvw ofl¶foroi ≥ épostr°fein a §peir«nto §jÉIsymoË, §ntaËya oÈk°ti pe¤yetai, éll' ’xeto pl°vn efiw Ñ EllÆsponton. §k toÊtou ka‹ §yanat≈yh ÍpÚ t«n §n Spãrt˙ tel«n…w épeiy«n.

For,inthefirstplace,aslongastheLacedaemonianswereatwarwiththeAthenians,heborehispartwiththem;then,onthearrivalofpeace,hepersuadedhisstatethattheThracianswereinjuringtheGreeks;hegainedhispointasbesthecouldfromtheephors,andsetsailwiththeintention of making war upon the Thracians who dweltbeyond theChersonese and Perinthus. When the ephors having changed their

mindsforsomereasonorother—h —tried to turnhimbackfromtheIsthmusofCorinth:atthatpointhedeclinedtorenderfurtherobedience,butwentsailingofftotheHellespont.AsaresulthewascondemnedtodeathbytheauthoritiesatSpartaonthegroundofdisobediencetoorders.



246 CHAPTERFIVE

Fromthepointofviewofinformationprocessing,thereader/hearerneedstobeinformedthatClearchushadalreadygoneinorderforthesequence‘theephorschangedtheirmindsforsomereasonorotherandtriedtoturnhimbackfromtheIsthmusofCorinth’tobefullycomprehensible. Properly speaking, this information is not to beconnected to the person of Clearchus himself ( aÈtÒn ), but to theactionsperformedbytheephors.Theparticipialclausedescribesthesituationinwhichtheephorschangedtheirminds,andtriedtorecall him. Therefore, a conjunct participle is disregarded here, and aparticipialclausethatissyntacticallydetachedfromitsleadingclausehasbeenused.Thisgenitiveabsolute,servingasakindofparenthesisoutsidethemainconstruction,illustratesthecorrespondencebetweenthegrammaticalformofaclauseanditsinformativefunction.

[31] Anabasis5.2.21-25§ntaËya pareskeuãzonto tØn êfodon, ka‹ toÁw m¢n stauroÁw ßkastoitoÁw kay' aÍtoÁw diπroun, ka‹ toÁw éxre¤ouw ka‹ fort¤a ¶xontãw te§jep°mponto ka‹ t«n ıplit«n tÚ pl∞yow, katalipÒntew ofl loxago‹

oÂw ßkastow §p¤steuen. §pe‹ d¢ ≥rjanto époxvre›n, §pej°yeon¶ndoyen pollo‹ g°rra ka‹ lÒgxaw ¶xontew ka‹ knhm›daw ka‹ krãnhPaflagonikã, ka‹ êlloi §p‹ tåw ofik¤aw én°bainon tåw ¶nyen ka‹¶nyen t∞w efiw tØn êkran feroÊshw ıdoË: Àste oÈd¢ di≈kein ésfal¢w∑n katå tåw pÊlaw tåw efiw tØn êkran feroÊsaw. ka‹ går jÊla megãla§perr¤ptoun ênvyen, Àste xalepÚn ∑n ka‹ m°nein ka‹ épi°nai: ka‹ ≤nÁj foberå ∑n ≤ §pioËsa. m

ye«n tiw a mhxanØn svthr¤aw d¤dvsin.§jap¤nhw går én°lamcen ofik¤a t«n §n dejiò ˜tou dØ §nãcantow. …w d' aÏth jun°pipten, ¶feugon ofl épÚ t«n §n dejiò ofiki«n. …w d¢¶mayen ı Jenof«n toËto parå t∞w tÊxhw, §nãptein §k°leue ka‹ tåw§n éristerò ofik¤aw, a„ jÊlinai ∑san, Àste ka‹ taxÁ §ka¤onto.¶feugon oÔn ka‹ ofl épÚ toÊtvn t«n ofiki«n.

Thentheymadepreparationsforthewithdrawal:theytoredownthepalisades,eachdivisiontakingthoseonitsownfront,andsentoffthemenwhowereunfitforserviceorwerecarryingburdens,andlikewisethegreaterpartofthehoplites,thecaptainskeepingbehindonlythosetroopsthattheyeachreliedupon.Theynosoonerbegantoretirethanthererushedoutuponthemfromwithinagreatcrowdofmenarmedwithwickershields,spears,greaves,andPaphlagonianhelmets,whileotherssetaboutclimbingtothetopsofthehousesthatwereoneithersideoftheroadleadinguptothecitadel.Theresultwasthatevenapursuitinthedirectionofthegatesthatledintothecitadelwasunsafe;fortheywouldhurldowngreatlogsfromabove,sothatitwasdifficult




eithertoremainortoretire.Andtheapproachofnightwasalsoacauseforfear.I somegodgavet ameansof salvation. Forof asudden oneof thehousesontherightbrokeintoablaze,setonfirebysomebodyorother.Thenthishousebegantofallin,andageneralflightfromtheotherhouses onthe rightside of the road ensued.Xenophon graspedthislessonwhichchancehadgivenhim,andgaveorders tosetfireto thehousesontheleftalso,whichwereofwoodandsofelltoburningveryquickly.Theresultwasthatthepeopleinthesehouseslikewisetooktoflight.

Thisexampleisaratherfamiliarone: 30attheonsetofanewDU,thetextcontinuesatthepointwhereithadstopped.Narrativetimeisnotpropelled forward, but maintained. A sentence-initially placedgenitive absolute resumes the situation described in the preceding unit;thesituationthatwasverystronglyimpliedthere(theirfightandawkward situation) is turned into the frame of reference for whatfollows: the opportunity for salvation. A textual boundary isarticulated by means of the (resumptive) genitive absolute, but thecurrenttopiciscontinued( aÈt«n )31whileanew‘participant’(ye«ntiw ) is added to the set. Topic management at a point of referentdiscontinuityandthetext-organizingfunctionofsecuringcoherencebyresuminginformationinordertopreparethereader/hearerforthepivotofthestorytriggertheuseofthegenitiveabsolute.

The remaining instances deal with postposed genitive absoluteconstructions.In thepartof Schwyzer’sarticle thatopenswith“derillegitimegen.abs.kannaberauchnachdemangeblichenBezugswortstehen, oft am Satzende” (Schwyzer, ibid.: 102ff.), he starts from

Herodotus 1.178: ke›tai (BabulΔn) §n ped¤ƒ megãlƒ, m°gayow§oËsa m°tvpon ßkaston e‡kosi ka‹ •katÚn stad¤vn §oÊshwtetrag≈nou(‘itliesinagreatplain,andisinshapeasquare,eachsidefifteenmilesinlength’;Schwyzer:“stattanscheinendkorrektem§oËsa tetrãgvnow”)and2.134:puram¤da d¢ ka‹ otow katel¤petopollÚn §lãssv toË patrÒw, e‡kosi pod«n katad°ousan k«lon

———30 Cf.thedescriptionof[12]ofChapter4,Section4.3.1.

31 The function of continuing the current topic as performed by the anaforicpronoun aÈtÒw,thesubjectofthepreposed,sentence-initiallyplacedgenitiveabsolutemusthavebeensowellknowntotheGreekreader/hearerthatXenophoncouldwrite maxom°nvn d¢ aÈt«n ka‹ époroum°nvn ye«n tiw(insteadofthealsowell-knownexpression tiw ye«n ),without therebeingtheriskofanyone interpretingye«nasthesubjectof époroum°nvn.



248 CHAPTERFIVE

ßkaston tri«n pl°yrvn, §oÊshw tetrag≈nou (Schwyzer: “statt

§oËsan tetrãgvnon”), l¤you d¢ §w tÚ ¥misu AfiyiopikoË(‘thisking,too,leftapyramid,butfarsmallerthanhisfather's,eachsidetwentyfeetshortofthreehundredfeetlong,squareatthebase,andasmuchashalfitsheightofEthiopianstone’),andargues:“[intheseinstances]liegt es nahe, an ein Anakoluth in umgekehrter Folge wie die zuAnfangbesprochenenzudenken.AberdieStrukturderbeidenSätzeistfeiner:diebeidengleichlautendengen.abs.gehörennichtzudenHauptverba und gehen nicht den vorhergehenden konjunkten

Participien §oËsa und §oËsan parallel, sondern sind diesenuntergeordnet,wasandersnurdurcheineParentheseoderhierbesserOpisthothese nach meiner Terminologie hätte ausgedrückt werdenkönnen. Also ist hier der gen. abs. einfach eine Opisthothese inkonzentrierterForm”.Thegeneralgistofhisargument,itseems,isthat these postposed genitive absolutes form an independentinformationunit,andarethereforesyntacticallydetachedfromtheirmatrixclause.Assuch,theymaybeseenaspresentingsomekindof ‘afterthought’.

SchwyzergivestwoexamplesfromtheworksofXenophon,whichwillnowbediscussed([32]and[33]).

[32] Cyropaedia6.1.36-38ı oÔn KËrow katamayΔn taËta §kãlesen aÈtÚn ka‹ mÒnow mÒnƒ¶lejen: ÑOr« se, ¶fh, Œ ÉArãspa, foboÊmenÒn te §m¢ ka‹ §n afisxÊn˙ dein«w ¶xonta. paËsai oÔn toÊtvn: §gΔ går yeoÊw te ékoÊv¶rvtow ≤tt∞syai, ényr≈pouw te o‰da ka‹ mãla dokoËntaw fron¤mouwe‰nai oÂa pepÒnyasin Íp' ¶rvtow: ka‹ aÈtÚw d' §mautoË kat°gnvn mØ ín karter∞sai Àste sunΔn kalo›w émele›n aÈt«n. ka‹ so‹ d¢ toÊtoutoË prãgmatow §gΔ a‡tiÒw efimi: §gΔ gãr se sugkate›rja toÊtƒ t“ émãxƒ prãgmati. ka‹ ı ÉArãspaw ÍpolabΔn e‰pen: ÉAllå sÁ m°n, ŒKËre, ka‹ taËta ˜moiow e‰ oÂÒsper ka‹ tîlla, pròÒw te ka‹suggn≈mvn t«n ényrvp¤nvn èmarthmãtvn: §m¢ d', ¶fh, ka‹ ofl êlloi ênyrvpoi katadÊousi t“ êxei. …w går ı yroËw di∞lye t∞w §m∞wsumforçw, ofl m¢n §xyro‹ §fÆdonta¤ moi, ofl d¢ f¤loi prosiÒntewsumbouleÊousin §kpodΔn ¶xein §mautÒn, mÆ ti ka‹ p ÍpÚ soË,

. ka‹ ı KËrow e‰pen: EÔ to¤nun ‡syi,Œ ÉArãspa, ˜ti taÊt˙ tª dÒj˙ oÂÒw t' e‰ §mo¤ te fisxur«w xar¤sasyaika‹ toÁw summãxouw megãla »fel∞sai. Efi går g°noito, ¶fh ı

ÉArãspaw, ˜ ti §g≈ soi §n kair“ ín geno¤mhn [aÔ xrÆsimow].SoCyrus,uponlearningofthis,sentforhimandhadatalkwithhiminprivate. “I see, Araspas”, said he, “that you are afraid of me andterriblyovercomewithshame.Donotfeelthatway,pray;forIhave




heardsaythatevengodsarevictimsoflove;andasformortals,Iknowwhatevensomewho are consideredverydiscreet havesufferedfromlove. And I have myself acknowledged that I would not have thestrength tobe thrown incontactwithbeautyand beindifferent toit.Besides,Iammyselfresponsibleforyourcondition,foritwasIthatshutyouupwiththisirresistiblecreature”.“Aye,Cyrus”,saidAraspas,interruptinghim,“youareinthis,justasineverythingelse,gentleandforgivingofhumanerrors.Othermenmakemereadytosinkwithmyshame;foreversincethereportofmyfallgotout,myenemieshavebeenexultingoverme,whilemyfriendscometomeandadvisemetokeepoutoftheway,forfearthatI fromyouf

”.“Letmetellyouthen,Araspas”,saidCyrus,“thatbyreasonofthisveryreportwhichpeoplehaveheardinregardtoyou,

you are in a position to do me a very great favor and to be of greatassistance to our allies”. “Would that some occasion might arise”,answeredAraspas,“inwhichIcouldbeofservicetoyou”.

Thesubjectofthegenitiveabsoluteisthesameasthesubjectofitsmatrixclause,sothataconjunctparticiplewouldhavebeenpossiblefromasyntacticalpointofview.32Apartfromthefactthatthegenitiveabsoluteformsanindependentinformationunittothematrixclauseasawhole,ratherthantoitssubjectalone,thereisalsothefactthatitsmatrixclause( mÆ ti ka‹ pãyv ÍpÚ soË ) presents a situation that isfocalizedbythefriends(ofl … f¤loi )ofAraspas,whereasthesituationpresented in the genitive absolute is focalized by Cyrus (soË ); thisanalysisissubstantiatedbythefactthatthegenitiveabsoluteisheadedby …w.Thisprovidesanadditionalreason,nottosaythedecisivefactor,forthechoiceofagenitiveabsolutehere.

[33] Anabasis1.4.11-13

§ntaËya ¶meinen ≤m°raw p°nte. ka‹ KËrow metapemcãmenow toÁwstrathgoÁw t«n ÑEllÆnvn ¶legen ˜ti ≤ ıdÚw ¶soito prÚw basil°a m°gan efiw Babul«na: ka‹ keleÊei aÈtoÁw l°gein taËta to›wstrati≈taiw ka‹ énape¤yein ßpesyai. ofl d¢ poiÆsantew §kklhs¤an épÆggellon taËta: ofl d¢ strati«tai §xal°painon to›w strathgo›w,ka‹ ¶fasan aÈtoÁw pãlai taËt' efidÒtaw krÊptein, ka‹ oÈk ¶fasanfi°nai, §ån mÆ tiw aÈto›w xrÆmata did“, Àsper t

parå tÚn pat°ra toË KÊrou, k

. taËta ofl strathgo‹ KÊrƒ épÆggellon:

———32 Note that this is one of the instances meant by Smyth, quoted above:

“Exceptionally,thesubjectofthegenitiveabsoluteisthesameasthatofthemainclause.Theeffectofthisirregularconstructionistoemphasizetheideacontainedinthegenitiveabsolute”.




reportingafurtherconsiderationtobeaddedforrhetoricalpurposestotheprecedingcomparisonasawhole.34

Finally,anexamplefromthe MemorabiliaismentionedbySchwyzerinafootnote(ibid.: 104):“Unberücksichtigt sind imVorangehendenStellenmitgen.abs.nebenabsolutgebrauchtemVerb,dasansichDat.hat,z.B. mou §pixeiroËntow front¤sai t∞w prÚw toÁw dikaståw

épolog¤aw ±nanti≈yh tÚ daimÒnion Xen. comm. IV 8, 5; mehrBeispielebeiKÜHNERa.a.O.111Anm.”

Thisisthepassageinfull:

[34] Memorabilia4.8.4-6l°jv d¢ ka‹ ì ÑErmog°nouw toË ÑIppon¤kou ≥kousa per‹ aÈtoË. ¶fhgãr, ≥dh MelÆtou gegramm°nou aÈtÚn tØn grafÆn, aÈtÚw ékoÊvn aÈtoË pãnta mçllon μ per‹ t∞w d¤khw dialegom°nou l°gein aÈt“ …w xrØ skope›n ˜ ti épologÆsetai. tÚn d¢ tÚ m¢n pr«ton efipe›n: OÈgår dok« soi toËto melet«n diabebivk°nai; §pe‹ d¢ aÈtÚn ≥reto˜pvw, efipe›n aÈtÚn ˜ti oÈd¢n êllo poi«n diageg°nhtai μ diaskop«n m¢n tã te d¤kaia ka‹ tå êdika, prãttvn d¢ tå d¤kaia ka‹ t«n éd¤kvn épexÒmenow, ¥nper nom¤zoi kall¤sthn mel°thn épolog¤aw e‰nai. aÈtÚw d¢ pãlin efipe›n: OÈx ıròw, Œ S≈kratew, ˜ti ofl ÉAyÆnhsi dikasta‹ polloÁw m¢n ≥dh mhd¢n édikoËntaw lÒgƒ paraxy°ntew ép°kteinan, polloÁw d¢ édikoËntaw ép°lusan; ÉAllå nØ tÚn D¤a, fãnai aÈtÒn, Œ ÑErmÒgenew, ≥

tÚ daimÒnion.ka‹ aÈtÚw efipe›n: Yaumastå l°geiw. tÚn d°, Yaumãzeiw, fãnai, efi t“ye“ doke› b°ltion e‰nai §m¢ teleutçn tÚn b¤on ≥dh; oÈk o‰sy' ˜ti m°xri m¢n toËde toË xrÒnou §gΔ oÈden‹ ényr≈pvn Ífe¤mhn ín oÎte b°ltion oÎy' ¥dion §mautoË bebivk°nai;

IwillrepeatwhatHermogenes,sonofHipponicus,toldmeabouthim.

“Meletus had actually formulated hisindictment”,hesaid.“Socratestalkedfreelyinmypresence,butmadenoreferencetothecase.Itoldhimthatheoughttobethinkingabouthisdefence.Hisfirstremarkwas,‘Don’tyouthinkthatIhavebeenpreparingforitallmylife?’Then I asked him how, and he said that he had been constantlyoccupiedintheconsiderationofrightandwrong,andindoingwhatwasrightandavoidingwhatwaswrong,whichheregardedasthebestpreparationforadefence.ThenIsaid,‘Don’tyousee,Socrates,thatthejuriesinourcourtsareapttobemisledbyargument,sothattheyoften put the innocent to death, and acquit the guilty?’ ‘Ah, yes,

Hermogenes’, he answered, ‘but I———

34 Schwyzer notes (ibid.: 103): “Die ans Ende gestellten illegitimen genitiviabsoluti sind gelegentlich durch eine besondere Einführung als «Schleppe»gezeichnet”.



252 CHAPTERFIVE

…Theonlythingisthatthedeityr ’.‘Strangewords’,saidI;andhe,‘Doyouthinkitstrange,ifitseemsbettertoGodthatIshoulddienow?Don'tyouseethattothisdayIneverwouldacknowledgethatanymanhadlivedabetterorapleasanterlifethanI?…”

Here, the deity did not resist Socrates’ undertakingtheattempt ( mo u§pixeiroËntow )ofthinkingouthisdefencetothejury,buttotheideaitselfofdefendinghimselfatall,asappearsfromSocrates’replytoHermogenes (tÚn d°, Yaumãzeiw, fãnai, efi t“ ye“ doke› b°ltion

e‰nai §m¢ teleutçn tÚn b¤on ≥dh; ). Therefore, the information thatSocratesdidtrytothinkouthisdefenseissyntacticallydetachedfromthe matrix clause, in order that the two information units arepresentedseparately.This,forthatmatter,createsamomentof(albeitslight)discontinuityinthewaythesubsequentinformationunitsareprocessed, which, in thiscontext,createsa certain tension. In itself,the fact that a genitive absolute can even combine with a matrixclause whose finite verb usually governs a dative, provides a verystrong sign indeed that the genitive absolute may be used for

pragmaticreasonsininformationprocessing.

2.3 SummaryandConclusion

Conjunct participles and genitive absolute constructions constitutetwodifferent,andwell-distinguished,grammaticalclasses.Whentheparticiplecan bejoinedwithaconstituent thatperformsasyntacticfunctioninthematrixclause,wespeakofaconjunctparticiple;when,

ontheotherhand,theparticiplecannotbejoinedwithaconstituentofthematrixclause,theparticipialclauseissaidtobe‘absolute’,andassuchisaltogetheroutsidethemainconstruction.

Extant Greek shows a relatively minor, but still considerablenumber ofinstances inwhichthe genitiveabsolute isusedwhile theparticiple could have been joined with a constituent performing asyntacticfunctioninthematrixclause.Intheseinstances,thesamereferentismentionedtwiceinonesentence:theonetimethroughthegenitiveasthesubjectoftheabsoluteconstruction,theotherthrough

oneoftheothercasesfitting thematrixclause.Moreoftenthannotthe pronoun aÈtÒworotow fills one of these slots. The genitiveabsolutemaybepreposedorpostposed.




Understandably,suchinstancesaredescribedas‘irregular’inoursentence-based grammars. The discourse-analytical approach toexamplesfromtheworksofXenophonaspresentedinSection2hasshown that the use of a genitive absolute ‘instead of’ a conjunctparticipleistobeunderstoodfromthesyntaxoftextgrammar,fromwhich point of view the usage of such genitive absolute is quiteunderstandable.

Thegenitiveabsolutehastheadvantageoveraconjunctparticipleoftypicallyhavingitsown(expressed)subject.Asithasitsownsubject

andverbalconstituent,thisclausetypeisdesententializedtoalesserdegreethanaconjunctparticiple,andtherewithoffersthespeakertheopportunitytopresentapieceofinformationwithoutinterferingwiththeconstructionofthesentenceasawhole.Bythesefeatures,thegenitiveisparticularlysuitableforcarryinginformationinthecaseof discoursecomplexity.Thegenitiveabsolute,buildingasyntacticallyindependentinformationunit,occurs‘insteadof’aconjunctparticiplewhenthe‘processingload’ofaclausecombinationisheavy,andtheconjunctparticipleisatoostronglydesententializedconstructiontobeused,especiallyintheobliquecases.

Acaseinpointisprovidedbythosepassagesthatarecomplexwithregard to the distribution of discourse topics. A referent (‘topic’) isexpressedasthesubjectofthepreposedgenitiveabsoluteandthenreferred back to by means of the pronoun aÈtÒworotowintheappropriatecasewhenthisreferent:• isabouttobecomeanactivediscourseparticipant([23]-[24];

[28])

• isjustaminorparticipantwhoneedstobeintroduced([22];[25])• hasdisappearedfromviewandneedstobere-introduced([26])Alternatively,whenthesentence-baseddemandofcase-agreement

issuspendedbecauseatopicneedstobecontinuedacrossadiscourseboundary, the pronoun aÈtÒw fills the subject slot of the genitiveabsolute([29];[31]).

The discussion of the different clause types that are central toSection2,aselsewhereinthisbook,hasshownthattherearegoodgroundstousetextgrammarforthedescriptionofthedifferentclausetypesinAncientGreek.



CHAPTERSIX

CONCLUSION

Although the grammar of Ancient Greek may have been welldescribed as a result of scholarly efforts, which were initiated by

Ancientgrammariansandboostedinthelatenineteenthcenturyandbeyond, there are stil l features that have been described lessconvincinglythanisdesirable.Thisappliestosyntacticphenomenainparticular.Inpart,thisisacorollaryofthefactthattheAncientGreeklanguage has been described so thoroughly, to the extent thatdescriptionsofferedbystandardgrammarsandreferencebooksarestillwidelyinuse,whilenewviewsandinsightstendtotakemoretimetoreplaceold-fashionedonesthanisdue.

One of the features to which this particularly applies is clausecombininginnarrativediscourse.Thisbookattemptstocontributetotheunderstanding of thedistributionofparticipialclauses (conjunctparticiples and genitives absolute) and subclauses (especially §pe¤-,§peidÆ-,and …w-clauses) using the linguistic method of DiscourseAnalysis.

Chapter 1 presents the status quaestionis: although the differencebetweentheclausetypesunderconsiderationhasbeenrecognizedinthesubjectliterature,ithasbeendescribedratherunconvincingly,due

to the sentence-based approach adopted by the authors of thestandardgrammars. Basing themselveson intuition, they claimthatsubclausesarepreferredtoparticipialclausesifthespeaker/narratorwishes to emphasize the content of the embedded predication.Moreover, some subject literature is primarily concerned with the‘temporal’relationoftheembeddedpredicationtoitsmatrixclause,takingthetense/aspectofthesubordinateverbalconstituentasthebasisforthisallegedrelation.Whilesuchatemporalrelationmay,and usually does, exist, as temporal organization is one of thecharacteristicsofnarrativediscourse,exampleshaveshownthatthisrelation is not necessarily coded by grammar, but resides in thecontextand/orsituation—thekindofactions involved, theorderinwhicheventsarepresented—,andthereader/hearer’sknowledgeof



CONCLUSION 255

theworld.Thetraditional,sentence-basedapproachfailstodescribethe distribution of the different clause types satisfactorily. Itsimportance notwithstanding, we should profit from text grammar,whichwasdevelopedincross-linguisticstudieswithpromisingresultsover roughly the last thirty years, but has only relatively recentlybeguntobearfruitinthedescriptionofAncientGreek.

This book focusses on the organization of texts and thecontribution of subclauses and participles to text coherence. As aresult of this text grammar-oriented approach, a more diversified

imageofthedistributionofsubclauses,genitiveabsoluteconstructionsand participial clauses emerges. We should distinguish, among subclauses, between clauses headed by the type of semanticallyspecificrelatorsthatisfoundgenerallyinlanguagesoftheworldandthose headed by the language-specific semantically non-specificrelator(§pe¤,§peidÆ,and…w )—thelatterformingthereal‘alternative’foraparticipialclause.Weshouldalsotryanddescribethedifferencein effect of the different embedded predications on thereader/hearer’scomprehensionofthetextinwhichtheyoccur.Whilesubclausesheadedby thesemanticallyspecific(‘causal’)relators diÒtiorby˜tiareusedtoilluminatetheinformationalcontentofanotherclauseandsemanticallyspecificrelatorssuchas ≤n¤ka,˜te,and§n ⁄areusedtoprovidethereader/hearerwithatimeanchorfortheRealWorld situation described in the matrix clause, the relators §pe¤,§peidÆ,and…waresemanticallynon-specificandsubclausesheadedbytheserelatorscanthereforebeusedtoagreatervarietyofendsthanjustrelatingthecontentofthesubclausetothecontentofthe

matrixclause.TheypresentafactualstatementabouttheRealWorldwhichthereader/hearerneedsforasuccessfulcomprehensionofthediscoursethatisintheprocessofbeingorganized.Semanticallynon-specificrelatorswilltypicallybeusedwhenthepointofviewfromwhich the events are relayed is involved, or when a Real Worldsituationhastoberelatedtothespeaker’sconstitutionofthetext.Theyareusedespeciallywhenthesubclauseoccupiesthesentence-initialpositionandintroducesanewphaseinanon-goingdescriptionofthe RealWorld,predictablyat a thematicbreak.Conversely,theparticipialclausethatoccupiesthesentence-initialposition,especiallywhen the conjunct participle or genitive absolute is resumptive orcontinuative, makes for textual coherence at points where a small-scaletextualboundaryisarticulatedwithinanotherwisecontinuous



256 CHAPTERSIX

thematic whole. Postposed ‘causal’§pe¤

-and§peidÆ

-clauses weredistinguishedfrompostposedparticipialclausesthatreceivea‘causal’interpretation in that the former occur predominantly in non-narrative episodes (speech situations), where they are used asmotivatingclauses.

During the investigation of the distribution of subclauses andparticipial clauses in Xenophon’s Hellenicaand Anabasis, withadditionalexamplesfromthe Agesilaus,theCyropaediaandothertextsby Xenophon, it turned out to be useful to link the degree of

desententializationoftheclause,whichcorrespondstothedegreeof linguisticcodingoftheallegedlyalternativetypes,totheirfunctionindiscourse. With an eye on the specific characteristics of—often—alargerstretchofcontext,morefactorsthatinfluencethedistributionof preposedsubclausesandparticiples,aswellastheusageofgenitivesabsolute‘insteadof’aconjunctparticiplecouldbemapped,andamoresatisfactoryaccountofclausecombininginAncientGreekcouldbeofferedthanthatofthetraditionalgrammarianswhoattemptedtodescribedifferencesbetweenclausetypes,insentencelevelsemanticterms,asdifferencesinthedegreeof‘emphasis’thatisputonthecontentoftheclause.

Throughoutthisbook,twolevelsrelevanttothestudyofnarrativesweredistinguished:the leveloftheRealWorldconstructionandthelevelof text articulation. Furthermore, three main categories of factorsinfluencingthechoicebetweendifferentclausetypeswereproposed: presentationofRealWorldrelations,textarticulation,andinformationprocessing .

Chapter2consistsofaselectionofsixparallelpassagesfromthe

worksofXenophon,whichIhavecalled‘near-minimalpairs’.Theyare ‘near-minimal’ inthat theydescribethe exactsameRealWorldeventsindifferenttexts(the Hellenicaandthe Agesilaus ) or the sametype of events according to a comparable scheme (such as thedescriptionofabanquet-sceneinthe AnabasisandtheCyropaedia ), orpresent comparable expressions or embedded predications with acomparable content in differently constituted texts. In the case of AncientGreek,wherenativespeakerscanobviouslynotbeconsulted,thisprocedurecangetoneascloseaspossibletoanativespeaker’sactualusageofthedevicesstudied.

OnthebasisoftheselectionofexamplesinthisChaptertheusageofdifferentclausetypes(independentmainclauses,subclauses,andparticipialclauses,bothconjunctandabsoluteparticiples)wasfound



CONCLUSION 257

tovaryaccordingtoboththe context andthetexttypeinwhichtheyareused.Itappearedfromtheobservationsandanalysesincomparablecontextsthateventsthataremostsalientinviewoftheconstitutionof aspecificdiscourseepisodewillgenerallybelinguisticallycodedina(finite)independentclause.Subclausesandparticipialclausesareusedto present events that are less central to the narrative backbone,althoughtheymaybe,andindeedoftenare,usedtopresenteventsbelonging to the narrative assertion. This applies to embeddedpredications with aorist verbal constituents in particular, but not

solely:acertainverbalactionmaybeconnectedtootheractionsinthecontextbymeansofapresentstemwhileanarrativeeventisbeing described. Conversely, the text-organizational function of anembeddedpredicationmayrequireittobeexpressedintheaoristwhile, strictly speaking, it does not propel narrative time forward.Resumptive embedded predications provide a case in point. Thediscussion of different clause types and clause combinations in thenear-minimalpairsconfirmedthehypothesisintroducedinChapter1that the type of clause used in clause combining is influenced byfactors concerning the presentation of Real World relations, textarticulation and/or information processing (including hierarchy of informationandtopicmanagement).Thesefactorsturnedout,eitherindependentlyofone anotherorin combination, tobe operative onthechoiceamongalternativeclausetypesintheirrespectivecontexts,andwere takenas thebasisforfurther investigationasdevelopedinthenextChapters.

ThepresentationofRealWorldrelationswasdiscussedinChapter

3.Inprinciple,aspeaker/narratorisfreetopresentaRealWorldsituationanywayhewants;hewillmakethisdecisionaccordingtohiscommunicativeaimsinthecontextinwhichRealWorldhappeningsare to be presented. This is, for that matter, the reason for mypreferencefortheterm presentation —instead of, for instance, theoftenfoundnotionofrepresentation:intheend,anyrepresentationofRealWorldhappeningsintheformoftextispresentation;ifwecompareRembrandt’spaintingofhissonTitusandPicasso’sofhissonPaulo,wemightbetemptedtoconcludethatRembrandt’spaintingoffersa

more natural representation of a boy, but both painters, in cognitivelyarrangingRealWorldexperienceandselectingpersonalimpressionswhileleavingothersaside,finallydecideonhowtheirsonsare presented tothespectator.



258 CHAPTERSIX

Ifthespeaker/narratorwishestoindicatearelationinthe(non-linguistic) Real World between the content of the embeddedpredicationandthecontentofitsmatrixclause,hewillpresentthisrelationbyusingarelator.InAncientGreek,relatorscomeinvariousshapes,withdifferentfunctions.Unlikemodernwesternlanguages,AncientGreekofferstheopportunitytocombineasubclauseandamainclausewithoutspecifyingtherelationinparticularbyasetof semanticallynon-specific relators: §pe¤, §peidÆ,and …w . This posesdifficultiesintranslation,asthelackofsuchsemanticallynon-specific

relatorsinmodernwesternlanguagescausesthetranslatortobemorespecificthantheGreeklanguageis;atthesametime,thisexplainsthediscussionwhetheragiven §pe¤,§peidÆ,or…wis‘temporal’or‘causal’(orinthecaseof§pe¤even‘concessive’),andthetendencyto‘identify’analleged‘temporal’relationonthebasisofthetense/aspectofthesubordinate verbal constituent. In fact, this relation is not at allexpressed.Moreover,therelators§pe¤,§peidÆ,and…wareoftenusedin contexts where specificity about a certain semantic relation isinexpedient, or even impossible. They are used to combinedeclarativeandnon-declarativeclauses,tocreateashiftofpointof view across sentences, and, probably the situation most frequentlyfoundinextantGreektexts,tomakeacertainsituationasawholetheframeof reference for thecontentof aclause (ordiscourseepisode),rather than to indicate one single relation. If, however, thespeaker/narratorwishestoindicateasinglerelation(eithertemporalor‘causal’),otherrelatorsareavailable:themainonesare˜te,≤n¤ka,and §n ⁄ for temporal relations, and ˜ti and diÒti for ‘causal’

relations.Text articulation was the subject of Chapter 4. Apart from thepresence orabsence ofrelators,another formal characteristicoftheembeddedpredicationsunderconsiderationshouldbeobserved;theyare either finite (subclauses) or non-finite (conjunct participles andgenitivesabsolute).Acontinuumofdesententializationwassetupbasedon the formal characteristics of embedded predications. Thesubclause, thegenitiveabsolute,and the conjunctparticiple, in thatorder,showincreasingdegreesofdesententialization.Thedifferencesin form were then related to the difference in functionofpreposed,especiallysentence-initiallyplaced,subclauses,genitivesabsoluteandparticipialclauses.Theupshotofthisapproachisthatwecannowunderstand their respective contribution to the constitution of



CONCLUSION 259

discourseepisodes(textarticulation).Thelinguisticdevicebywhichapoint of segmentation is marked is determined by the kind of segmentationinvolved,asthearticulationofthetextfollowsonthe(re-)constructionandpresentationofRealWorldevents.Whenthediscourse is discontinuous to a fairly high degree with respect toreferents,time,place,oraction-eventsequence,thearticulationofthetext will predictably show a high degree of linguistic coding; theheaviertheboundarypertainingtothethematicorganizationoftheReal World in the formofnarrative text, the heavier the linguistic

codingthetextreceiveswillbe,andviceversa.Thesubclause,then,showingthemostheavylinguisticcoding,tendstobeusedatpointsof highthematicdiscontinuity,andtheconjunctparticiple,whichis themost strongly desententialized clause type and therefore shows theleastheavylinguisticcoding,isusedwhenthetextissegmentedwithinan on-going thematic structure. The genitive absolute takes theintermediateposition;itoffersthespeakeranopportunitytopresentinformation that contributes to a correct understanding of themessagewithoutinterferingwiththeconstructionofthesentenceasawhole.Therefore,itispreferablyusedbythespeakerasaminimallynecessary linguistic device for securing textual coherence at pointswheretwotextualunitsaresegmented(and/orlinked).

In the caseof postposed ‘causal’ clauses, the degree of linguisticcoding of the clause corresponds to its contribution to textcomprehension.Postposed‘causal’ §pe¤-and§peidÆ-clausesareusedtomotivateaprecedingutterance,actionorpositionthatcallsforsubstantiation in its linguistic or non-linguistic context. Postposed

participles that receive a ‘causal’ interpretation just provide‘additional’ information, without which the preceding utterancewouldstillremaincomprehensible.ThiswasdescribedinChapter5,Section1,whichwasdevotedtoaspectsofinformationprocessing.

Section 2 of Chapter 5 discussed the usage of genitive absoluteconstructions where, from the point of view of sentence-syntax, aconjunctparticiplewouldhavebeenanalternativeoption.Here,apragmaticapproachtotheflowofinformationinon-goingdiscourse,withtheuseofviewsonthesyntaxoftextgrammaroutlinedintheforegoingChapters,isadoptedinadiscussionofeverysingleinstancementionedinanarticlebySchwyzerdatingasfarbackas1942,andserving here as an exponent of the now out-dated sentence-basedapproach.Fromtheviewpointoftextgrammar,theusageofsuch



260 CHAPTERSIX

genitiveabsoluteconstructionsisquiteunderstandable.Therelativelylessstronglydesententializedgenitiveabsolutehastheadvantageoveraconjunctparticipleoftypicallyhavingitsown(expressed)subject.Itcan build a syntactically independent information unit, and isthereforetheclausetypetouseinthecaseofdiscoursecomplexity,whenthe ‘processingload’ofa clausecombinationisheavyandtheconjunctparticipleisatoostronglydesententializedconstructiontobeused,especiallyintheobliquecases.Further,thegenitiveabsoluteisfound‘insteadof’aconjunctparticiplewhenthereisneedof‘topic

management’.Participantswhoperformasyntacticfunctioninthematrixclause are nonethelessreferredto asthesubjectofa genitiveabsolute both when the discourse requires them to be overtly(re-)introducedandwhentheyneedtobecontinuedacrossadiscourseboundary.

The discussion brings to light that a text grammar for AncientGreekisindispensablefortheunderstandinganddescriptionofthesyntactical function of l inguistic features such as embeddedpredications and their contribution to text production and textcomprehension.Ihavenot,however,presentedthefindingsofthisbook in a matrix, for even if texts are seen not as a sequence of sentences, but as coherent discourse, a description of syntacticphenomenawillalwayshavetotakeintoaccountthattheonefactorcannotbeconsideredwithouttheother.Thisisnotaset-back—itisonly natural when the object of research is language, which causesonetodescriberegularitieswhichwillhavetobereconsideredineachindividual context. At the same time, the existence of semantically

non-specific relators in Ancient Greek offers an interest ing opportunity for cross-linguistic study on the subject as well as anexplanationoftheinadequatedescriptionintheliteraturethusfar.



BIBLIOGRAPHY

Apter,D.E.,1993.‘Yan’anandtheNarrativeReconstructionofReality’, Daedalus122Spring’93,207-232.

Armstrong,D.,1981.‘TheAncientGreekAoristastheAspectofCountableAction’,

in:Tedeschi& Zaenen,eds.(1981),1-12.Bakker,E.J.,1990.‘HomericDiscourseandEnjambement.ACognitiveApproach’,

TAPA120,1-21. ——, 1991 a. ‘Linguistische verhaalanalyse. Temporele bijzinnen in Herodotus

2.121’, Lampas24,82-96. ——, 1991 b. ‘Foregrounding and Indirect Discourse. Temporal Subclauses in a

HerodoteanShortStory’, JournalofPragmatics16,225-247. ——,1993. ‘Boundaries,TopicsandtheStructureofDiscourse. AnInvestigationof

theAncientGreekParticle DÉ ’,StudiesinLanguage17-2,275-311. ——,1994.‘Voice,AspectandAktionsart:MiddleandPassiveinAncientGreek’,in:

Fox& Hopper,eds.(1994),23-47. ——, ed. 1997. Grammar as Interpretation. Greek Literature in its Linguistic Contexts,

MnemosyneSupplement 171,Leiden. ——,1997a.‘VerbalAspectandMimeticDescriptioninThucydides’,in:Bakker,ed.

(1997),7-54. ——,1997b. ‘Storytelling intheFuture:Truth,Time,andTense inHomericEpic’,

in:Bakker& Kahane,eds.(1997),11-36. —— & Kahane,A.,eds.1997.WrittenVoices,SpokenSigns.Tradition,Performance,andthe

EpicText ,Cambridge,Massachusetts/London,England. ——, unpublished paper. ‘Whose Point of View? Linguistics and Narratology in

GreekHistorians’.Bartsch,R.,1985.‘OnAspectualPropertiesofDutchandGermanNominalizations’,

in:LoCascio& Vet,eds.(1985),7-39.

Basset,L.,1997.‘ éll' §jÒloisy' aÈt“ koãj.Réexamendesemploisde éllã àlalumièredel’énonciationdanslesGrenouillesd’Aristophane’,in:Rijksbaron,ed.(1997),75-99.

Bauman, R., 1986. Story, Performance, and Event. Contextual Studies of Oral Narrative,Cambridge,CambridgeUniversityPress.

Berent,G.P.,1973.‘AbsoluteConstructionsas‘SubordinateClauses’’,in:Corum et al .,eds.(1973),147-154.

Biraud,M.,ed.1991. Étudesdesyntaxedugrecclassique.Rechercheslinguistiquesetapplicationsdidactiques.Actesdu premier Colloque InternationaldeDidactique de laSyntaxe duGrecclassique17,18,19avril1991UniversitédeNice,PublicationsdelaFacultédesLettreset ScienceshumainesdeNice7,Paris.

Bolkestein, A.M., 1991. ‘Causally Related Predications and the Choice between

ParataxisandHypotaxisinLatin’,in:Coleman,ed.(1991),427-451. ——, 1993. ‘Cohesie in tekst en syntactische alternatieven’, in: Goedegebuure, ed.(1993),161-174.

——,1996. ‘Zand zonder kalk:Cohesie enhet proza vanSeneca’, Lampas19,298-308.



262 BIBLIOGRAPHY

Bolling,G.M.,1960.‘Descriptionof§pe¤’sSyntaxinHomer’,Glotta38,18-33.

Boogaart,R.& Noordegraaf,J.,eds.1994. Nauwebetrekkingen.VoorTheoJanssenbijzijnvij ftigste verjaardag , Stichting Neerlandist iek VU, Amsterdam/NodusPublikationen,Münster.

Brown,G.& Yule,G.,1983. DiscourseAnalysis,Cambridge.Brugman,C.& Macaulay,M.,eds.1984. ProceedingsoftheTenthAnnualMeetingofthe

BerkeleyLinguisticsSociety,Berkeley,California.Bruner,J.,1991.‘TheNarrativeConstructionofReality’,CriticalInquiry18,1-21.Buijs, M., (unpublished). ‘Aspectual Differences and Narrative Technique:

Xenophon’s Hellenica& Agesilaus’.Bybee,J.L.& Dahl,Ö.,1989.‘TheCreationofTenseandAspectSystemsinthe

LanguagesoftheWorld’,StudiesinLanguage13,51-103.Calboli,G.,ed.1989.SubordinationandOtherTopicsinLatin(StudiesinLanguageCompanion

Series17),Amsterdam/Philadelphia.Chafe,W.L.,1976.‘Givenness,Contrastiveness,Definiteness,Subjects,Topics,and

PointofView’,in:Li,ed.(1976),25-55. ——, 1984. ‘How People Use Adverbial Clauses’, in: Brugman & Macaulay,eds.

(1984),437-449. ——,1987. ‘CognitiveConstraintsonInformationFlow’,in: Tomlin,ed. (1987),21-

51. ——,1988.‘LinkingIntonationUnitsinSpokenEnglish’, in:Haiman& Thompson,

eds.(1988),1-27.Chausserie-Laprée,J-P.,1969. L’expressionnarrativechezleshistorienslatins,Paris.Collinge,N.E.,1988.‘ThoughtsonthePragmaticsofAncientGreek’, PCPhS 214

( NewSeries34),1-13.Cole,P.,ed.1981. RadicalPragmatics,NewYork.Coleman,R.,ed.1991.NewStudiesinLatinLinguistics(StudiesinLanguageCompanionSeries

21),Amsterdam/Philadelphia.Comrie,B.,1976. Aspect ,Cambridge.Conte,M-E.,Petöfi,J.S.& Sözer,E.,eds.1989.TextandDiscourseConnectedness(Studies

inLanguageCompanionSeries16),Amsterdam/Philadelphia.Corum,C.,Smith-Stark,T.C.& Weiser,A.,eds.1973.YouTaketheHighNodeandI'll

TaketheLowNode,Chicago.Coulthard,M.,19852. AnIntroductiontoDiscourseAnalysis,LondonandNewYork.Dahl,Ö.,1985.TenseandAspectSystems,Oxford.Danes,F.,1989.‘“FunctionalSentencePerspective”andTextConnectedness’,in:

Conte,Petöfi& Sözer,eds.(1989),23-31.Deese,J.,1978.‘ThoughtintoSpeech’, AmericanScientist 66,314-321.Denniston,J.D.,19705.TheGreekParticles,Oxford.Dik,H.,1995.WordOrderinAncientGreek.APragmaticAccountofWordOrderVariationin

Herodotus( AmsterdamStudiesinClassicalPhilology5),diss.Amsterdam.Dik,S.C.,1989.TheTheoryofFunctionalGrammar.Part1:TheStructureoftheClause ,

Dordrecht.Dover,Sir.Kenneth,1997.TheEvolutionofGreekProseStyle,Oxford.Due,B.,1989.TheCyropaedia.Xenophon’sAimsandMethods,Århus.Duhoux,Y.,1997.‘Grecécritetgrecparlé.Uneétudecontrastivedesparticulesaux

Ve-IVesiècles’,in:Rijksbaron,ed.(1997),15-48.Ehrlich,S.,1990a. PointofView.ALinguisticAnalysisofLiteraryStyle,LondonandNew

York. ——, 1990 b. ‘Referential Linking and the Interpretation of Tense’, Journal of

Pragmatics14,57-75.Eriksson,K.,1943. DasPräsens historicum indernachklassischengriechischenHistoriographie,

Lund.



BIBLIOGRAPHY 263

ErpTaalmanKip,A.M.van,1997.‘ ∑ gãrinQuestions’,in:Rijksbaron,ed.(1997),

151-156.Exter Blokland, A.F. den, 1995. In Search of Text Syntax. Towards a Syntactic Text-

SegmentationModelforBiblicalHebrew,Diss.Amsterdam(FreeUniversity).Fleischman,S., 1989. ‘Temporal Distance:ABasicLinguisticMetaphor’,Studiesin

Language13,1-50. ——,1990.TenseandNarrativity.FromMedievalPerformancetoModernFiction,London.Fox,B.,1983.‘TheDiscourseFunctionoftheParticipleinAncientGreek’,in:Klein-

Andreu,ed.(1983),23-41. ——& Hopper, P.J., eds. 1994. Voice: Form and Function, (Typological Studies in

Language27)Amsterdam/Philadelphia.Gera,D.L.,1993. Xenophon’sCyropaedia.Style,Genre,andLiteraryTechnique,Oxford.Giora, R., 1983. ‘Segmentation and Segment Cohesion: On the Thematic

OrganizationoftheText’,Text 3,155-181.Givón,T.,ed.1979.SyntaxandSemantics,Volume12:DiscourseandSyntax,NewYork.

——, 1983. ‘Topic Continuity in Discourse: The Functional Domain of Switch-reference’,in:Haiman& Munro,eds.(1983),51-82.

——,1990.SyntaxAFunctional-TypologicalIntroduction,volII,Amsterdam/Philadelphia. ——, 1993. En gli sh Grammar . A Functi on-Bas ed In tro du ct io n (2 Vols.),

Amsterdam/Philadelphia.Goedegebuure,J.,ed.1993. Nieuwewegen in taal- en literatuurwetenschap.Handelingenvan

hetEenenveertigsteFilologencongres,Tilburg.Gray,V.,1989.TheCharacterofXenophon’sHellenica,Baltimore.Grimes,J.E.,1975.TheThreadofDiscourse,TheHague/Paris.

——,ed.1986.SentenceInitialDevices,SummerInstituteofLinguistics,Dallas.Haiman,J.,1978.‘ConditionalsareTopics’, Language54,564-589. — — & Munro, P., eds. 1983. Swi tch Ref er ence and Unive rsal Grammar ,

Amsterdam/Piladelphia. — — & Thompson, S.A., 1984. ‘“Subordination” in Universal Grammar’, in:

Brugman& Macaulay,eds.(1984),510-523. — — & Thompson, S.A., eds.1988. Clause Combining in Grammar and Discourse

(TypologicalStudiesinLanguage18),Amsterdam/Philadelphia.Hatzfeld,J.,ed.1936. XénophonHelléniques,Paris.Havers,W.,1926.‘Dersog.‘ Nominativuspendens’’, IF 43,207-257.Helttula, A., 1987. Studies on the Latin Accusative Absolute (CommentationesHumanarum

Litterarum81),Helsinki.

Hettrich,H.,1976. KontextundAspektinderaltgriechischenProsaHerodots,Göttingen.Hinds,J.,1979.‘OrganizationalPatternsinDiscourse’,in:Givón,ed.(1979),135-157.

Hoff,F.,1989.‘Lesablativesabsolusirréguliers:unnouvelexamenduproblème’,in:Calboli,ed.(1989),401-423.

Hofmann,H.,ed.1991. Fragmenta Dramatica. Beitrage zur Interpretation der griechischenTragikerfragmenteundihrerWirkungsgeschichte,Göttingen.

Hohler, T.P., 1987. ‘Storytelling and Human Experience’ (review of P. Ricoeur:TempsetrécitI,II,III ), ResearchinPhenomenology17,291-303.

Holland,G.B.,1986.‘NominalSentencesandtheOriginofAbsoluteConstructionsinIndo-European’,ZeitschriftfürVergleichendeSprachforschung 99,163-193.

Hopper,P.J.,1979.‘AspectandForegroundinginDiscourse’,in:Givón,ed.(1979),213-241.

Horrocks,G.,1995.‘OnCondition…:AspectandModalityintheHistoryofGreek’, PCPS 41,153-173.

Hwang,S.J.J.,1994.‘RelativeClauses,AdverbialClauses,andInformationFlowinDiscourse’, LanguageResearch30-4,673-706.



264 BIBLIOGRAPHY

Instone,S.,1996. Pindar:SelectedOdes:OlympianOne,PythianNine,NemeansTwo&Three,

IsthmianOne,Warminster.Ireland, S., 1977. ‘Sentence Structure in Aeschylus and the Position of the

PrometheusintheCorpusAeschyleum’, Philologus121,189-210. Jacquinod, B.,1997. ‘Sur lerôlepragmatiquedeka¤toi’,in:Rijksbaron,ed.(1997),

131-149. Janson, T.,1964. ‘TheProblems ofMeasuring Sentence-Length inClassical Texts’,

StudiaLinguistica18,26-36. Jehne, M., 1995. ‘Die Funktion des Berichts über die Kinadon-Verschwörung in

Xenophons Hellenika’, Hermes123-2,166-174. Jong, I.F. de, 1997. ‘ Gãr Introducing Embedded Narratives’, in: Rijksbaron, ed.

(1997),175-185.Klein-Andreu,F.,ed.1983. DiscoursePerspectivesonSyntax,NewYork.Klug, W., 1992. Erzählstruktur als Kunstform. Studien zur künstle rischen Funktion der

ErzähltemporaimLateinischenundimGriechischen,Heidelberg.Koller, H., 1951. ‘Praesens historicum und erzählendes Imperfekt. Beitrag zur

AktionsartderPraesensstammzeitenimLateinischenundGriechischen’, Museum Helveticum8,63-99.

Komen, J.A.M., 1994. Over de ontwikkeling vanabsolute constructies, diss. Amsterdam(UvA),Amsterdam[withabibliographyonabsoluteconstructions].

Kortmann,B.,1991. FreeAdjunctsandAbsolutesinEnglish,London/NewYork.Knebel,G.,1960.‘Bemerkungenzu“Descriptionof §pe¤’sSyntax”vonG.Melville

Bolling’,Glotta38,34-43.Kraus,W.,1970.‘Einkonzessiv-adversatives§pe¤?Vorstudiefüreinengriechischen

Thesaurus’,in:Wimmel,ed.(1970),146-150.Krieken,R.van,1980. Actio, Aspect en hetWerkwoord. De temporele zinsconstructies in het werkvanHerodotus,diss.Leiden.

Kroon, C., 1995. Discourse Particles in Latin. A Study of nam, enim, autem, vero and at ( AmsterdamStudiesinClassicalPhilology4),diss.Amsterdam.

Kuppevelt,J.van,1991.TopicenComment.Explicieteenimplicietevraagstellingindiscourse ,diss.Nijmegen.

Labéy,D.,1950. Manueldesparticulesgrecques,Paris.Labov, W., 1972. ‘The Transformation of Experience into Narrative Syntax’, in:

LanguageintheInnerCity,Philadelphia.Lakoff,R.,1984.‘ThePragmaticsofSubordination’,in:Brugman& Macaulay,eds.

(1984),481-492.

Lang,M.L.,1995.‘ParticipialMotivationinThucydides’, MnemosyneXLVIII-1,48-65.Lehmann, C., 1988. ‘Towards a Typology of Clause Linkage’, in: Haiman &

Thompson,eds.(1988),181-225.Levinsohn, S.H., 1987. Textual Connections in Acts(The Society of Biblical Literature

MonographSeries31),Atlanta,Georgia.Li,C.N.,ed.1976.SubjectandTopic,NewYork.Lo Cascio, V. & Vet, C., eds. 1985. Temporal structure in Sentence and Discourse,

Dordrecht.Longacre,R.E.,1979.‘TheParagraphasaGrammaticalUnit’,in:Givón,ed.(1979),

115-134. ——,1983.TheGrammarofDiscourse,NewYork. ——,1985. ‘SentencesasCombinations ofClauses’, in:Shopen,ed.(1985, II),235-

286. ——,1985.‘DiscoursePeakasZoneofTurbulence’,in:Wirth,ed.(1985),83-98.Lücke,O.,1876. AbsoluteParticipiaimGotischenundihrVerhältnisszumgriechischenOriginal,

mitbesondererBerücksichtigungderSkeireins,Magdeburg.



BIBLIOGRAPHY 265

McKay,K.L.,1988.‘AspectualUsageinTimelessContextsinAncientGreek’,in:

Rijksbaron,Mulder,Wakker,eds.(1988),193-208.Marchese,L.,1987.‘OntheRoleofConditionalsinGodiéProceduralDiscourse’,in:

Tomlin,ed.(1987),263-280.MartínezVázquez,R.,1995.‘Tiemporelativoengriegoantiguo’, PhilologíaHispalensis

10,175-197.Matthiessen, C. & Thompson, S.A., 1988. ‘The Structure of Discourse and

‘Subordination’’,in:Haiman& Thompson,eds.(1988),275-329.Morpurgo Davies, A., 1997. ‘Particles in GreekEpigraphical Texts. The Case of

Arcadian’,in:Rijksbaron,ed.(1997),49-73.Muchnova,D.,1991.‘Lespropositionscausalesditesdemotif’,in:Biraud,ed.(1991),

135-143.Myhill, J.& Hibiya, J., 1988. ‘The Discourse Function of Clause-chaining’, in:

Haiman& Thompson,eds.(1988),361-398.Nickau,K.,1993.‘ZumVerbalaspektderImperativeinAristophanes’Acharnern

1097-1142’,Glotta71,158-166.Nickel, R., 1979. Xenophon(WissenschaftlicheBuchgesellschaft,ErträgederForschungBand

111),Darmstadt.Oosten,J.van,1984.‘AUnifiedViewofTopic’,in:Brugman& Macaulay,eds.

(1984),372-385.Ophuijsen,J.M.van,1993.‘ OUN ,ARA,DH,TOINUN:TheLinguisticArticulationof

Arguments in Plato’s Phaedo ’,TwoStudiesinAtticParticleUsage.Lysias&Plato(C.M.J.Sicking& J.M.vanOphuijsen), MnemosyneSupplement 129,Leiden,67-164.

Pelliccia,H.,1989.‘Pindar, Nemean7.31-36andtheSyntaxofAetiology’, HSPh92,71-101.Pfeijffer,I.L.,1999. FirstPersonFuturesinPindar ( HermesEinzeilschriften81),Stuttgart.Pinkster,H.,1990. LatinSyntaxandSemantics,LondonandNewYork.Polanyi,L.& Scha,R.J.H.,1983.‘TheSyntaxofDiscourse’,Text 3,261-270.Poulakos,T.,1987.‘Isocrates’sUseofNarrativeinthe Evagoras:EpideicticRhetoric

andMoralAction’,QuarterlyJournalofSpeech73,317-328.Prince,E.F.,1981.‘TowardaTaxonomyofGiven-NewInformation’,in:Cole,ed.

(1981),223-255.Rajam,V.S.,1985.‘TheDurationofanAction—RealorAspectual?TheEvolution

ofthePresentTenseinTamil’, JAOS 105.2,277-291.Ramsey,V.,1987.‘TheFunctionalDistributionofPreposedandPostposed“if”and

“when”ClausesinWrittenDiscourse’,in:Tomlin,ed.(1987),383-408.Ricoeur,P.,1983-1985.Tempsetrécit ,Paris.Riedinger, J-C., 1991. Étude sur les Helléniques. Xénophon et l'Histoire(Coll.d'Études

Anciennes120),Paris.Rijksbaron,A.,1976.Temporal&CausalConjunctionsinAncientGreekwithSpecialReference

totheUseof§pe¤ and…w inHerodotus,diss.Amsterdam. ——,1986. ‘Taalkundeendestructuur vanHerodotus’ Historiën’, Lampas19,220-

230. ——, 1988. ‘The Discourse Function of the Imperfect’, in: Rijksbaron, Mulder,

Wakker,eds.(1988),237-254. ——, Mulder, H.A. & Wakker,G.C.,eds.1988. In the Footsteps of Raphael Kühner ,

Amsterdam. ——, 1991.GrammaticalObservationsonEuripides’Bacchae( AmsterdamStudies inClassical

Philology1),Amsterdam. ——,1997.‘Adverbor Connector?Thecase ofka‹... d°’,in:Rijksbaron,ed.(1997),

187-208. ——, ed. 1997. New Approaches to Greek Particles. Proceedings of the ColloquiumHeld in



266 BIBLIOGRAPHY

Amsterdam,January4-6, 1996, toHonourC.J. Ruijgh on theOccasionof hisRetirement

( AmsterdamStudiesinClassicalPhilology7),Amsterdam. ——,2001a.‘DeXenophon-fabriek’, Lampas34.2,114-139. ——,2001b.Overbepaaldepersonen,Amsterdam. ——,20023.TheSyntaxandSemanticsoftheVerbinClassicalGreek ,Amsterdam.Risselada, R., 1993. Imperatives and other Directive Expressions in Latin. A Study in the

Pragmatics of a Dead Language( Amsterdam Studies in Classical Philology 2),diss.Amsterdam.

Rossum-Steenbeek,M.E.van,1993.‘RashConclusionsonfourXenophonPapyri’,ZPE Band97,57-66.

——,1994.‘NotesonXenophon’, MnemosyneVol.XLVII, Fasc.4,454-465.Ruijgh,C.J.,1971. Autourde“ teépique”.Étudessurlasyntaxegreque,Amsterdam.

——,1991. ‘Apropos dupronom · (Soph.Fr.471Radt)’,in:Hofmann,ed.(1991),61-78.

RuizYamuza,E.,1998.‘Sintaxis,semánticaypragmáticadelaasignacióndesujetoenpasivaengriegoclásico’, RevistadelaSociedadEspañoladeLingüística28.2,333-358.

Sadock,J.M.& Zwicky,A.M.,1985.‘SpeechActDistinctionsinSyntax’,in:Shopen,ed.(1985,I),155-196.

Sansone,D.,1990.‘TheComputerandthe HistoriaAugusta:aNoteonMarriott’, JRS 80,174-177.

Schwyzer,E.,1942. ‘Zumsog.genitivusabsolutusstatt participiumconiunctumimGriechischen’, Emerita10,98-104.

Shopen,T.,ed.1985. LanguageTypologyandSyntacticDescription,VolumeI:ClauseStructure

&VolumeII:ComplexConstructions,Cambridge.Sicking,C.M.J.,1991.‘TheDistributionofAoristandPresentTenseStemFormsinGreek,EspeciallyintheImperative’,Glotta69,14-43and154-170.

——, 1993. ‘Devices for Text Articulation in Lysias I and XII’,TwoStudiesinAttic Particle Usage. Lysias&Plato(C.M.J.Sicking& J.M.vanOphuijsen), MnemosyneSupplement 129,Leiden,1-66.

——,1994. ‘TaalkundigeobservatiesbijhetonderwijsinhetGrieks’, Lampas27.1,4-27.

——,1996. ‘AspectChoice.TimeReferenceorDiscourse Function?’,TwoStudiesintheSemanticsoftheVerbinClassicalGreek (C.M.J.Sicking& P.Stork), MnemosyneSupplement 160,Leiden,1-118.

—— & Stork,P.,1996.‘TheSyntheticPerfectinClassicalGreek’, TwoStudiesinthe

SemanticsoftheVerbinClassicalGreek (C.M.J. Sicking& P.Stork), Mnemo syneSupplement 160,Leiden,119-298. ——,1997.‘ParticlesinQuestionsinPlato’,in:Rijksbaron,ed.(1997),157-174. —— & Stork,P.,1997.‘TheGrammaroftheSo-CalledHistoricalPresentinAncient

Greek’,in:Bakker,ed.(1997),131-168.Slings,S.R.,1992.‘WrittenandSpokenLanguage:anExerciseinthePragmaticsof

theGreekLanguage’,CPh87,95-109. ——,1994a.‘Eentandjelager.AanzettenvooreenoralegrammaticavanHomerus’,

Lampas27.5,411-427. ——, 1994 b. ‘Geschiedenis van het perfectum in het oud-Grieks’, in: Boogaart&

Noordegraaf,eds.(1994),239-247. ——,1997.‘AdversativeRelatorsBetweenPUSHandPOP’,in:Rijksbaron,ed.(1997),

101-129. ——,2002. ‘‘En devrouw, dieziethemweggaan’. Orale strategieën inde taalvan

Herodotus’, Lampas35.1,25-43.Sluiter,I.,1990. Ancient GrammarinContext. Contributions to theStudy ofAncient Linguistic

Thought ,diss.Amsterdam.



BIBLIOGRAPHY 267

——,1997.‘ParapleromaticLucubrations’,in:Rijksbaron,ed.(1997),233-246.

——,2000. ‘SevenGrammariansontheAblativeAbsolute’, HistoriographiaLinguistica27.2/3,381-416.

Smith,C.S.,1983.‘ATheoryofAspectualChoice’, Language59-3,479-501. ——,1991.TheParameterofAspect ,Dordrecht.Spieker,E.H.,1885.‘Ontheso-calledGenitiveAbsoluteanditsUseespeciallyinthe

AtticOrators’, AJPh6,310-343.Stadter, P. 1997. ‘Herodotus and the North Carolina Oral Narrative Tradition’,

Histos1997,www.dur.ac.uk/Classics/histos/1997/stadter.htmlStehouwer, A.P., 1995. SeverianvonGabala in Illud: PoneManum Tuam, et inDiversa

Testimonia(CPG4198),diss.Amsterdam(FreeUniversity)[withanAppendixon:PostposedCausal§pe¤-,§peidÆ-and˜ti-clausesinSeverianandChrysostom].

Sterling,N.,unpublishedpaper1997.‘ALittleFriendlyPersuasion:thePerformanceofAgesilaus,Lysander,andXenophonin Hellenika3’.

Sternberg,M.,1982.‘PointofViewandtheIndirectionsofDirectSpeech’, LanguageandStyle15.2,67-117.

Stork,P.,1982.TheAspectualUsageoftheDynamicInfinitiveinHerodotus ,diss.Leiden,Groningen.

——, 1988. ‘Aspectual Variant Readings in Herodotus’, in: Rijksbaron, Mulder,Wakker,eds.(1988),265-289.

——,unpublishedpaper.‘“Gettingsomewhere”metg¤nomaibijHerodotus?’.Stump,G.T.,1985.TheSemanticVariabilityofAbsoluteConstructions,Dordrecht.Tedeschi, P.J.& Zaenen,A.,eds.1981.SyntaxandSemantics,Volume14:Tenseand

Aspect ,NewYork.

Terwelp,G.,1873. DeAgesilai,quiXenophontisnominefertur,auctore,diss.Monasterii.Thompson,S.A.,1985. ‘Grammar andWrittenDiscourse:Initialvs.FinalPurposeClausesinEnglish’,Text5,55-84.

—— & Longacre,R.E.,1985.‘AdverbialClauses’,in:Shopen,ed.(1985,II),171-234.

——,1987. ‘Subordination and Narrative Event Structure’, in: Tomlin, ed. (1987),435-454.

Tomlin,R.S.,ed.1987.CoherenceandGroundinginDiscourse(TypologicalStudiesinLanguage11),Amsterdam/Philadelphia.

Vasilaros, G.N., 1993. Der Gebrauch des Genitivus absolutus bei Apollonios Rhodios imVerhältniszuHomer ,Athens.

Virtanen, T., 1992. ‘Given and New Information in Adverbials: Clause-initial

AdverbialsofTimeandPlace’, JournalofPragmatics17,99-115.Waanders, F.M.J., 1997. ‘Particulars of Some Proto-Indo-European Particles’, in:Rijksbaron,ed.(1997),269-274.

Wakker,G.C.,1994.ConditionsandConditionals.AnInvestigationofAncientGreek ( AmsterdamStudiesinClassicalPhilology3),diss.Amsterdam.

——, 1997. ‘Emphasis and Affirmation. Some aspects of mÆn in Tragedy’, in:Rijksbaron,ed.(1997),209-231.

Wathelet, P., 1997. ‘Les particules ke(n)et ê n dans les formules de l’épopéehomérique’,in:Rijksbaron,ed.(1997),247-268.

Weinrich,H.,19773(1964).Tempus.Besprochene und erzählteWelt , Stuttgart, Berlin,Köln,Mainz.

Wimmel,W.,ed.1970. ForschungenzurrömischenLiteratur.Festschriftzum60.Geburtstagvon KarlBüchner I,Wiesbaden.

Wirth, J.R., ed. 1985. Beyond the Sentence: Discourse and Sentential Form,AnnArbor,Michigan.

Zycha,J.,1885.‘Gebrauchvon§pe¤§pe¤per;§peidÆ§peidÆper’,WienerStudien7,82-115.



268 BIBLIOGRAPHY

Editions,Commentaries,andTranslationsUsed

Breitenbach,L.,1873-1876. Xenophon’sHellenica,Berlin.Brownson,C.L.,ed.19614. XenophonHellenica,BooksI-V ,London.

——,ed.19504. XenophonHellenica,BooksVI&VII;Anabasis,BooksI-III ,London.Büchsenschütz,B.,1880-1891. XenophonsgriechischeGeschichte,Leipzig.Gemoll, W., 1968 (ed. correctiorem curavit J. Peters). Xenophontis Institut io Cyri

( BibliothecaScriptorumGraecorumetRomanorumTeubneriana ),Lipsiae.Hude,C.,1934(repr.1969). XenophontisHistoria Graeca( Bibliotheca ScriptorumGraecorum

etRomanorumTeubneriana ),Stutgardiae. ——,1972 (ed.correctioremcuravitJ.Peters). XenophonAnabasis( Bibliotheca Scriptorum

GraecorumetRomanorumTeubneriana ),Lipsiae.Hug,A.,1905. Xenophontis Institutio Cyri ( Bibliotheca ScriptorumGraecorum etRomanorum

Teubneriana ),Lipsiae.Hutchinson,Th.,17383, 17565. XENOFVNTOS KUROU PAIDEIAS BIBLIA OKTV,

Londini. ——,1774. XenophontisCyropaedia,Lipsiae.Keller,O.,1908. XenophontisHistoriaGraeca( BibliothecaScriptorumGraecorumetRomanorum

Teubneriana ),Lipsiae.Krentz,P.,1989. XenophonHellenica I-II.3.10.EditedwithanIntroduction,Translation and

Commentary,Warminster. ——, 1994. XenophonHellenicaII.3.11-IV.2.8.EditedwithanIntroduction,Translationand

Commentary,Warminster.

Lendle,O.,1995. KommentarzuXenophonsAnabasis(Bücher1-7),Darmstadt.LeunclaviusAmelburnus,J.,1729,1736.XENOFVNTOS KUROU PAIDEIAS BIBLIA HÄ,Londini.

Marchant,E.C.,ed.198011. XenophontisOperaOmnia (Tomus I:HistoriaGraeca&cet.),Oxford.

Rehdantz,C.& Carnuth,O.,1905-1912. XenophonsAnabasis,Berlin.Schneider,J.G.,1819. XenophonHistoriaGraeca(&cet.),Oxonii.Sélincourt,A.de,19615. Herodotus.TheHistories,Middlesex.Underhill,G.E.,1906. XenophonHellenica,Oxford.

GrammarsUsed

Bornemann,E.& Risch,E.,1973.GriechischeGrammatik ,FrankfurtamMain.Gildersleeve,B.L.,1980(reprinted,withanindexofpassagescitedcompiledbyP.Stork).SyntaxofClassicalGreekfromHomertoDemosthenes,Groningen.

Goodwin,W.W.,1889.SyntaxoftheMoodsandTensesoftheGreekVerb,London.Humbert,J.,19603.SyntaxeGrecque,Paris.Hummel,P.,1993. LaSyntaxedePindare,Paris.Kühner, R.& Gerth,B.,1898-1904. Ausführl iche Grammatik der griechischen Sprache.

ZweiterTeil:Satzlehre,Hannover.Schwyzer,E.& Debrunner,A.,1950. GriechischeGrammatik.II:Syntaxundsyntaktische

Stilistik ,München.Smyth,H.W.,1920(rev.ed.1956).GreekGrammar ,Cambridge,Massachusetts.

Forfurtherblibliographicreferences,see:

Buijs,M. ABibliographyofAncientGreekLinguistics:www.let.uu.nl/hist/goac/bgl



INDEXOFLINGUISTICTERMS

Back-reference Instance where an item in the preceding discourseisevoked

Clausetype The grammatical class a certain clausebelongsto:• an independent finite clause

(mainclause)• a dependent finite clause(subclause)• adependentnon-finiteclause:• – inflection marking

(infinitive)• +inflectionmarking participle agreeing in thegenitive with its ownsubject(genitiveabsolute)• +inflectionmarking participle showing case-a g r e e m e n t wi t h aconstituent of its matrixclause(conjunctparticiple)

Closure-marking High density of linguistic marking (anaphoric deictic elements, particleclusters)oftenco-occurringwithcontent-oriented marking of the terminus of a

narritiveepisode,e.g.,whenaparticipantistakenoffthestage,oralapseoftimeorrelocationisindicated

Complementation-marking m°n… d° Instance of text articulation where in aclause combination co-ordinated by m°n… d°,the d°-memberisanticipatedbythepreceding m°n;theparticle d° doesnotmark the onset of a new D U (cf.Transition-marking m°n… d° )

Contextuallypreparedinformation A pieceof information that is inferrable

from other entities in the preceding discourseorthespeechsituation



270 INDEXOFLINGUISTICTERMS

Continuativeclause Aclausecontainingcontextuallyprepared

information and continuing an eventsequencealreadysetinmotion

Desententialization Thereductionprocess inwhichaclauselosescertainpropertiesandistransformedintoacategoryofa lowersyntactic level;in the caseof the Ancient Greek clausetypesunderconsiderationthecontinuumof desententia lizat ion consi st of anindependent clause’ asubclause’ agenitiveabsolute’aconjunctparticiple,inthatorder

DevelopmentUnit(DU )1 Oneormoresentencesoftext,markedbythe speaker as the next phase in thedevelopment of the discourse by(‘solitary’) d° atitsonset.Withinsuchaunit,sentencesmaybecloselyconnectedto the initial sentence by means of theparticleka¤or‘asyndeton’

Discourseperspective Thecognitivev iew,on thepartofboththe speaker and his audience, of the

speaker’s aims and objectives inproducingagivenpieceofdiscourse

Discourseturbulence The occurrence of a thematic breakdefined in terms of discontinuity of participants, time, place and/or action-sequence, usual ly marked by a highdegreeoflinguisticcoding

Embeddedpredication Asyntacticallydependentclausesuchasasubclause, a genitive absolute or aconjunctparticipialclause

Entirelygiveninformation Apieceofinformationalreadyprocessedintheprecedingdiscourseorpresentinthespeechsituation

Entirelynewinformation Apieceofinformationnotyetprocessedintheprecedingdiscourseornotpresentinthespeechsituation

———1 ThetermDevelopmentUnitwas first coinedbyLevinsohn (1987:83-85,and

179); his definition runs: “one ore more sentences of narrative, introduced by adevelopmentalconjunction(e.g., de )andassociatedbykai or te,whichpresentsanewdevelopmentinthestory”.



INDEXOFLINGUISTICTERMS 271

Expectancychain Aseriesof twoormoreverbalactions the

first of which gives rise to a predictionabouttheaction(s)tofollowonthebasisofone’sknowledgeoftheworld

Head-headlinkage The discourse-level strategy found inlanguages of the world consisting of aback-referenceattheonsetofanewDUtotheonsetoftheprecedingDU

Lexicaloverlap Verbal repetition of a constituent fromtheprecedingdiscourse

Matrixclause The superordinate clause on which anembedded predication is dependent. Anembeddedpredicationmaybedependenton (part of) a main clause or anotherembeddedpredication

Progressiveclause A clause containing entirely newinformation and moving the discourseforwardtoanewpoint

Proposition The semantic correlate of a (possibly

desententialized)clause2

Propositionaloverlap Repetition of an idea expressed in theprecedingdiscourse

Reciprocalcoupling Aneventsequence inwhich,e.g., speakerAsayssomethingtoB,tobefollowedbythestatement‘BhearsA’,inwhichcasethe actions ‘say’ and ‘hear’ arereciprocals3

Resumptiveclause Aclause thatbyrestatementor summary

r ec ap it ul at es i nf or ma ti on a lr ea dyprocessedinthediscourse

Segmentation The linguistic reflection of the speaker’splantoorganizehisdiscourseinphases

Sentencetype The class a certain clause belongs toaccordingtoitsspeechact:• adeclarativeclause• aninterrogativeclause• adirectiveclause

———2 DefinitionbyLehmann(1988:182).3 DescribedassuchbyThompson& Longacre(1985:213).



272 INDEXOFLINGUISTICTERMS

Small-scalesegmentation Minimal linguistic articulation of a

boundary between two adjacent DU ’s .Usually the only linguistic marking atsuch points i s the occurrence of theparticle d° , but other particles may beused

Summary-headlinkage The discourse-level strategy found inlanguages of the world consisting of summarizing(partof)a DUattheonsetof thenextDU

Tail-headlinkage The discourse-level strategy found inlanguages of the world consisting of repeating(partof)thelastfragmentofaDUattheonsetofthenextDU

Transition-marking m°n… d° Instance of text articulation where in aclause combination co-ordinated by m°n… d°,theparticle d°doesmarktheonsetof a new D U (cf . Complementation-marking m°n… d° )



GENERALINDEX

afterthought,248agreement,81-82,151-52,222-23,232,

245,253

anacoluthon,233,236anaphoricgrounding,157anteriority,4-5,90aspect,3-9,22-23,138,254

back-reference, 76,139, 145, 163, 168,177-78,189

behavioraldiscourse,47boundary, 9, 15-17, 22, 30, 58-59, 69,

71-72,79,82,111,139,147-48,158,159n17, 163, 165, 169, 171, 173,

178n27,180,183n31,186,191,194,196,198-200,245,247,253,255

buildupunit(BU ),139-40,174

cataphoricgrounding,189cataphorictemporalsuccession,176causalrelations,13,22,85-89,258clausechainingparticiple,32-33closure,20,31,34,57,64,71,193,235coherencelink,157coherencestrand,138,153-54,169,188,

191-92coincidentuse,50n25communicativemode,146complementationmarking m°n… d°,40concessive-adversative value of §pe¤,

115n23contextually prepared information, 17-

20,158,173-74continuativeclause,146n7,157,159n17,

174n25,194-95,200,255continuity/discontinuity,7-8,15,30-31,

51,61,71,82-83,138-39,153-200,224-25,230,247,252,259

desententialization, 149-53, 155, 199-200,220,222,253,256,258,260

developmentunit(DU ),137,140discourseanalysis,1,3,11,21,254discoursecomplexity,223,231,238,253

discourseconnectedness,160discoursepeak,147discourseperspective,68discoursepragmatics,77discourseturbulence,61,165,175,222distributiveactions,47n21durativeaction,161

embeddedparagraph(EP ),140,146,160emphasis,2,78,221,256entirely given information, 17, 19-20,

185entirelynewinformation,17,20,158erzählteZeit,51Erzählzeit,51event,9-12expectancychain,59,139,174expositoryparagraph,146

focalization,43,49,213,249focusfunction,47n21,233frameofreference,8,116,162,179,247,

258

genitiveabsolutevs.conjunctparticiple,221-53

giveninformation,17

head-headlinkage,165hierarchy,32-33,35-37,52,84,162,257historical present,6, 105, 168, 182n30,

186

infinitivalclause,152informationflow,22,139,201,231-32,259

information processing, 13, 17-21, 36,84,201-53,256-57



GENERALINDEX 274

informationstatus,19-20,37,47n21,57,

144,158,159n17,183n31,189

leveloftextarticulation,11,15,32,47,58,72,144,154,194-95,198,256

leveloftheRealWorldconstruction,11,14-16, 58, 67, 72, 154, 177, 194,196,198,256

lexicaloverlap,139,167,189,198,225,230

maininformation,17,20

modeofdiscourse,146,160motivatingclause,205-06,209-11,220,

256,259

narrative time, 11-12, 68, 93, 97, 99,105, 107, 125, 146, 160, 171, 177,179,191,198,225,230,247,257

near-minimalpair,21,24-84,256-57neuterpluralfillingsubjectslot,151-52newinformation,17

paragraph,139,146-47,186participanttracking,41,61,81,133n40,147,223

pointofview,41,43,49,107,119,121,123-25,176,195,214,220,258

posteriority,4pos tposed e mbedde d pre di ca ti on s

receiving a ‘causal’ interpretation,201-21

presentationofRealWorldrelations,13-15,17,42,84-136,256-57

progressive clause, 158, 163, 174n25,182,183n31,191,195-96propositionaloverlap,32,139,186processingload,49,223,253punctualaction,143-44,186

question-and-answerparagraph,183

reality-level,11recapitulation,58,168-69reciprocal coupling, 69, 139-40, 145,

153,189,191representation,257restatement,20,37,145,162,167,179,

198

resumptive clause, 19, 31, 37, 162-63,

166, 168, 182n30, 198, 200, 247,255,257

rulesvs.regularities,21

script-predictableinformation,189segmentation, 7, 9, 16, 40, 52, 68, 71,

137-200,230,259semantically non-specific relations, 4-5,

13-14, 17, 22, 85, 108-29, 134-36,176,199,255,258,260

semanticallyspecificrelations,12-14,16,

21,85-112,129-36,255,258sentencetype,122,207,258sequenceparagraph(SP ),140, 163, 168,

186n35simultaneity,4,8,90slippage,236n20,238solitary d°,139,148spatialsetting,60-61,71,129specific parts of the day or year filling

subjectslotofasubclause,93-96storyofMania,140-48

storytellingmode,146subjectcontinuity,80subsidiaryinformation,17,20,42summary,20,37,145summary-headlinkage,139

tail-headlinkage,139,179,186temporal relations, 11-14, 89-112, 254,

258temporalsubclauses,3-9terminativeaction,198

text articulation, 13, 15-17, 19-20, 22,24-25,32,46,71,84,137-200,256-58

text comprehension, 17, 19, 135, 220,255,260

text grammar, 1, 3, 22, 200,222, 231,233-53,255,259-60

texttype,46,50,52,84,257texture,37,160thematicunit,16,19,32,65,68-69,137,

139,160,163-78,182,194

thematiccoherence,159-63topicmanagement,62,78-79,81-82,84,223,231,236,238,247,257,260

topicswitch,31,59,80,82,145,189,191,231

transitionmarking m°n… d°,57,71



INDEXLOCORUM

Xenophon Agesilaus1.6 46n16,n17

1.7-10 1121.9 47n181.10 46n161.19-25 431.25 94n5,n61.38 46n16

165n201.38-2.2 532.2 1502.9-11 962.13-14 122

2.15 25,282.23 62.25 114n203.1 46n16

Anabasis1.1.1-3 1641.1.4 2141.1.7 33n71.2.1 124n311.2.21 212

1.3.9 2051.3.20 152n141.4.2 1011.4.11-13 2491.5.9 2151.6.5-7 1191.6.7-8 1831.6.10-11 1181.7.18-19 1091.8.1 106,180n291.8.8 93n4,106

1.8.17 1061.8.26-27 1971.8.28-29 2021.9.24 2041.10.9-10 99

1.10.15-16 1032.1.1 292.2.13 95n8

2.2.14-15 862.3.9 124n312.4.6 250n332.4.24 95n92.5.15-24 742.6.2-4 2452.6.20 913.1.2 250n333.1.5-7 1933.1.33-34 1033.3.1 139n3,179

3.3.11-13 116,2093.4.1-2 2253.4.34 93n4,94n73.4.49 1123.5.2 93n43.5.3-7 1754.1.13 152n144.2.7 95n84.3.2 264.3.13-16 1884.5.14 31n5

4.5.18 131n394.5.29 274.6.1 95n104.6.23 95n84.7.27 95n115.1.14-17 1045.2.21-25 2465.3.4-7 1695.5.13-18 2075.6.9 2155.6.18 108

5.6.35-36 2065.8.11 207n46.2.13-15 386.3.11-15(18) 72n45,746.4.9 95n10



INDEXLOCORUM 276

6.4.10 27

6.5.5-6 1617.3.2 95n87.3.23-26 627.3.41 95n8

Cyropaedia1.3.1 214n91.4.1-2 2441.4.19-20 721.4.20 244n291.5.4-6 241

1.6.13-15 2392.2.17 58n292.4.30 293.1.43 273.2.2 283.3.28 273.3.33 284.1.7 274.1.9 22n284.5.21 186n345.3.51 28

5.3.51-57 30n25.3.57 22n28,30n25.4.19 265.4.43-50 726.1.36-38 2486.1.37 239n226.3.37 287.2.1 277.2.2 22n287.3.1 267.5.40 26

8.4.12-13 628.4.13 65n388.4.21-23 69

DeEquitandiratione1.1 7n12

Hellenica1.1.1-2 2031.1.9 2251.1.25-26 179n28

1.3.1 2181.3.7 2171.4.1-3 2241.4.6-7 2101.4.12-13 197

1.5.1 151

1.6.1-2 901.7.30 152n142.1.6 1002.1.10-12 1902.1.27 87,95n102.2.10 2182.3.23-24 1772.3.35-36 1032.3.54 152n132.4.10-11 382.4.11 40n10

2.4.11-12 42n152.4.12 983.1.7 1843.1.10-16 140-483.1.16 168n213.2.19 152n143.2.21 14n223.2.25-26 74,2313.3.1-4 883.3.4 1853.4.2-5 171

3.4.6 133n413.4.9-10 723.4.12-15 1813.4.15-16 433.4.16 94n5,n63.4.25 53.4.26 2073.4.29-3.5.1 1563.5.21-23 1253.5.25-4.1.3 1564.1.15–18 228

4.1.24 152n134.2.4-8 534.2.8 1004.2.9-23 54n284.3.1-4 534.3.3 1504.3.16-17 964.3.20-21 25,264.4.5 2194.5.1-3 130n374.5.3 130

4.6.12 934.7.7 1314.8.1 12n154.8.3 1014.8.7–9 237



INDEXLOCORUM 277

5.1.13 211

5.1.21 94n85.2.39 174n265.3.25 1925.4.9 94n85.4.47 94n5,n65.4.53-54 1235.4.56-58 2265.4.59 94n5,n66.1.1-3 81n55,1666.1.19 12n156.2.31 194

6.3.1 212n76.5.10-21 6n107.1.18 1287.1.32 186n347.1.34 1017.2.7 1507.2.12-13 1237.2.19 4n77.3.4-7.4.1 12n157.4.1 178n277.4.2-4 234

7.4.3-4 183n317.4.23-24 87.4.38 95n87.4.39 132

Memorabilia4.8.4-6 251

Symposium4.49-50 70

OtherauthorsHerodotus1.186 105n172.121g 58n303.19 105n173.74 105n173.136 105n174.95 105n174.124 105n176.97 105n176.129 69n42,72n44

Pindar N.7.31-36 13n20 P.4,292 186n34 fr.124c(94 ) 186n34

PlatoGorgias448b11-c3 115n22473e3-4 115n22474b6-7 115n22

Sophocles Philoctetes637-638 186n34

Thucydides2.13.9 102n162.21.1 102n16

michel buijs - clause combining in ancient greek narrative disourse the distribution of subclauses...

Documents