Practice in a Second Language

THE CAMBRIDGE APPLIED LINGUISTICS SERIES

Series editors: Michael H. Long and Jack C. Richards

This series presents the findings of work in applied linguistics that are of directrelevance to language teaching and learning and of particular interest to appliedlinguists, researchers, language teachers, and teacher trainers.

Recent publications in this series:

Cognition and Second Language Instruction edited by Peter RobinsonComputer Applications in Second Language Acquisition by Carol A. ChapelleContrastive Rhetoric by Ulla ConnorCorpora in Applied Linguistics by Susan HunstonCriterion-referenced Language Testing by James Dean Brown and Thom HudsonCritical Pedagogies and Language Learning by Bonny Norton and Kelleen TooheyCulture in Second Language Teaching and Learning edited by Eli HinkelExploring the Dynamics of Second Language Writing edited by Barbara KrollExploring the Second Language Mental Lexicon by David SingletonFeedback in Second Language Writing edited by Ken Hyland and Fiona HylandFocus on Form in Classroom Second Language Acquisition edited by Catherine

Doughty and Jessica WilliamsImmersion Education: International Perspectives edited by Robert Keith Johnson and

Merrill SwainInsights into Second Language Reading: A Cross-Linguistic Approach by Keiko KodaInterfaces Between Second Language Acquisition and Language Testing Research edited

by Lyle F. Bachman and Andrew D. CohenLearning Vocabulary in Another Language by I. S. P. NationNetwork-based Language Teaching edited by Mark Warschauer and Richard KernPragmatics in Language Teaching edited by Kenneth R. Rose and Gabriele KasperResearch Genres: Explorations and Applications by John SwalesResearch Perspectives on English for Academic Purposes edited by John Flowerdew and

Matthew PeacockResearching and Applying Metaphor edited by Lynne Cameron and Graham LowSecond Language Vocabulary Acquisition edited by James Coady and Thomas HuckinSociolinguistics and Language Teaching edited by Sandra Lee McKay and Nancy H.

HornbergerTeacher Cognition in Language Teaching: Beliefs, Decision-Making, and Classroom

Practice by Devon WoodsText, Role, and Context by Ann M. JohnsUnderstanding Expertise in Teaching: Case Studies of Language Teacher Development

by Amy B. M. Tsui

Practice in a SecondLanguagePerspectives from AppliedLinguistics and CognitivePsychology

Edited by

Robert M. DeKeyserUniversity of Maryland at College Park

cambridge university pressCambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo

Cambridge University Press32 Avenue of the Americas, New York, NY 10013-2473, USA

www.cambridge.orgInformation on this title: www.cambridge.org/9780521865296

C Cambridge University Press 2007

This publication is in copyright. Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place withoutthe written permission of Cambridge University Press.

First published 2007

Printed in the United States of America

A catalog record for this publication is available from the British Library.

Library of Congress Cataloging in Publication Data

Practice in a second language : perspectives from applied linguistics and cognitivepsychology / edited by Robert M. DeKeyser.

p. cm. (The Cambridge applied linguistics series)Includes bibliographical references and index.ISBN-13: 978-0-521-86529-6 (hardback)ISBN-10: 0-521-86529-8 (hardback)ISBN-13: 978-0-521-68404-0 (pbk.)ISBN-10: 0-521-68404-8 (pbk.)1. Language and languages Study and teaching. 2. Second language acquisition.I. DeKeyser, Robert. II. Title. III. Series.P51.P68 2007418.0071dc22 2006049272

ISBN 978-0-521-86529-6 hardbackISBN 978-0-521-68404-0 paperback

Cambridge University Press has no responsibility forthe persistence or accuracy of URLs for external orthird-party Internet Web sites referred to in this publicationand does not guarantee that any content on suchWeb sites is, or will remain, accurate or appropriate.

Contents

List of contributors ixSeries editors preface xi

Introduction: Situating the concept of practice 1Robert M. DeKeyser

I FOUNDATIONS 19

Chapter 1 Input in the L2 classroom: An attentional perspectiveon receptive practice 21Ronald P. Leow

Chapter 2 Output practice in the L2 classroom 51Hitoshi Muranoi

Chapter 3 Interaction as practice 85Alison Mackey

Chapter 4 Feedback in L2 learning: Responding to errors duringpractice 111Jennifer Leeman

II INSTITUTIONAL CONTEXTS 139

Chapter 5 A cognitive approach to improving immersion studentsoral language abilities: The Awareness-Practice-Feedback sequence 141Leila Ranta and Roy Lyster

Chapter 6 Second language education: Practice in perfect learningconditions? 161Kris Van den Branden

vii

viii Contents

Chapter 7 Meaningful L2 practice in foreign language classrooms:A cognitive-interactionist SLA perspective 180Lourdes Ortega

Chapter 8 Study abroad as foreign language practice 208Robert M. DeKeyser

III INDIVIDUAL DIFFERENCES 227

Chapter 9 Age-related differences and second language learningpractice 229Carmen Munoz

Chapter 10 Aptitudes, abilities, contexts, and practice 256Peter Robinson

Conclusion: The future of practice 287Robert M. DeKeyser

Glossary 305Index 313

Contributors

Robert M. DeKeyser, University of Maryland, USA

Jennifer Leeman, George Mason University, USA

Ronald P. Leow, Georgetown University, USA

Roy Lyster, McGill University, Canada

Alison Mackey, Georgetown University, USA

Carmen Munoz, University of Barcelona, Spain

Hitoshi Muranoi, Tohoku Gakuin University, Japan

Lourdes Ortega, University of Hawaii at Manoa, USA

Leila Ranta, University of Alberta, Canada

Peter Robinson, Aoyama Gakuin University, Japan

Kris Van den Branden, Katholieke Universiteit Leuven, Belgium

ix

Series editors preface

Sometimes maligned for its allegedly behaviorist connotations but criti-cal for success in many fields from music and sport to mathematics andlanguage learning, practice is undergoing something of a revival in theapplied linguistics literature, and this latest volume in the CambridgeApplied Linguistics Series will undoubtedly heighten interest. To theextent that language is a skill, it behooves us to determine optimal waysof providing opportunities for skill development as well as the optimaltiming of those opportunities.

Robert M. DeKeysers introduction to Practice in a Second Languagereminds us, among other things, of progress in this area in cognitivepsychology and that the notion of practice lies at the intersection of anumber of other issues of fundamental importance in second languagelearning and teaching. They include relationships between declarativeand procedural knowledge /skill, automatization, rule-based and item-based learning, implicit and explicit knowledge, the value and timingof feedback, and the transferability, or not, of potentially skill-specificand task-specific abilities. In other words, the practice issue and theoryand research findings on practice are important for a wide variety oftheoretical and practical issues in second and foreign language acquisitionand teaching.

The book you have before you consists entirely of original contribu-tions authored by some of the leading researchers in the field, assembledand edited by Professor DeKeyser of the University of Maryland. Pro-fessor DeKeyser has himself published groundbreaking empirical studieson practice, broadly construed. Practice in a Second Language is a majorcontribution to scholarship in SLA and applied linguistics, of value toresearchers and classroom practitioners alike, and a worthy addition tothe Cambridge Applied Linguistics Series.

Michael H. LongJack C. Richards

xi

Introduction: Situating the conceptof practice

Robert M. DeKeyser

Practice gets a raw deal in the field of applied linguistics. Most lay-people simply assume that practice is a necessary condition for languagelearning without giving the concept much further thought, but manyapplied linguists eschew the term practice. For some, the word conjuresup images of mind-numbing drills in the sweatshops of foreign languagelearning, while for others it means fun and games to appease students onFriday afternoons. Practice is by no means a dirty word in other domainsof human endeavor, however. Parents dutifully take their kids to soccerpractice, and professional athletes dutifully show up for team practice,sometimes even with recent injuries. Parents make their kids practice theirpiano skills at home, and the worlds most famous performers of classicalmusic often practice for many hours a day, even if it makes their fingershurt. If even idolized, spoiled, and highly paid celebrities are willing toput up with practice, why not language learners, teachers, or researchers?The concept of second language practice remains remarkably unexam-ined from a theoretical point of view. Misgivings and misunderstandingsabout practice abound and are often rooted in even deeper misunder-standings about what it is that language learners are supposed to learn.In this introductory chapter, I will try to provide some conceptual andterminological clarification in preparation for the rest of the book. In theconcluding chapter, I will then formulate tentative recommendations forthe praxis of practice, as they follow from these conceptual distinctionsand from the other chapters.

It should be clear from the outset, of course, that the word practice inthe title is not meant as the opposite of theory, as in foreign languageteaching policy vs. actual practice in secondary schools, practicingprofessionals, or the praxis of second language teaching. The con-tributors to this book all understand practice in a much more focusedway, as specific activities in the second language, engaged in systemati-cally, deliberately, with the goal of developing knowledge of and skillsin the second language. But within this broad definition there are stillmany different ways one can understand the concept of practice. Beforewe zero in on the meanings of practice in applied linguistics, however,

1

2 Robert M. DeKeyser

let us have a brief look at how cognitive and educational psychologistshave used the term.

The notion of practice in cognitive psychology

The study of skill acquisition is an important area within cognitive psy-chology (for a good, concise overview see Carlson, 2003). Researchersin that area have documented the acquisition of skills in a wide vari-ety of domains, from algebra, geometry, and computer programmingto learning how to drive a car or how to roll cigars. Increasingly, theyalso employ neuroimaging and other neurological data to document howdifferent skills and different stages in the acquisition of the same skillare represented in the brain (e.g., Anderson, Bothell, Byrne, Douglass,Lebiere, & Qin, 2004; Posner, DiGirolamo, & Fernandez-Duque, 1997).Researchers who study skill acquisition processes all agree that reactiontime and error rate decline gradually as a function of practice with agiven task. But how is practice defined in this literature?

Carlson, a well-known contemporary theorist of skill acquisition,defines practice simply as repeated performance of the same (or closelysimilar) routines (1997, p. 56). This definition could easily be misin-terpreted and may seem a throwback to the heydays of behaviorism; itdoes not define practice in terms of cognitive processes at all. Clearly, thiscannot be what Carlson has in mind, however, because he defines skillas the ability to routinely, reliably and fluently perform goal-directedactivities as a result of practice with those activities (1997, p. 45). Thedefinition in Newell and Rosenbloom is more precise: Practice is the sub-class of learning that deals only with improving performance on a taskthat can already be successfully performed (1981, p. 229). Not onlydoes this definition make the learning/improving component of practiceexplicit, it also states clearly that practice in the narrow sense appliesonly to a task that can already be successfully performed. But the defini-tion still remains vague because it does not say what constitutes a task.Is speaking French a task, or requesting a glass of water in French, ordoes using a conditional verb form to do this constitute a task? Any taskoutside the laboratory, whether in school or in the real world, consistsof many components. Could practice be defined as referring to repeatedengaging/improving in task components that can already be successfullyperformed? If so, how narrowly could a task component be defined?And would practice of task components separately be better or worsethan practice of the whole task? There is no general answer to the latterquestion (VanLehn, 1989), but Lee and Anderson (2001), for instance,show how the learning of at least one complex task reflects the learningof much smaller parts.

Situating the concept of practice 3

This brings us to the related question of the specificity of practice effectsin skill learning. If tasks can be defined at such a low level, then howdifferent can they be before they are different altogether? In other words,how much does practice on a task that shares certain characteristics withanother, but also differs from it in a crucial way, contribute to improvingperformance on this other task? What determines transfer of the effect ofpractice? A number of studies have shown that the practice effect is quitespecific in the sense that there is only minimal transfer between tasksthat superficially appear to be each others mirror image, such as writingversus reading a computer program (see esp. Singley & Anderson, 1989;cf. also Anderson, 1993, chap. 9; Muller, 1999).

This specificity of the practice effect is explained by the well-knowndistinction between declarative and procedural knowledge. In most formsof skill acquisition, people are first presented with information, e.g.,rules about how to write a computer program or put a French sentencetogether in explicit form (declarative knowledge). Through initialpractice they incorporate this information into behavioral routines (pro-duction rules, procedural knowledge). This procedural knowledgeconsists of very specific rules and can be used fast and with a low errorrate, but the disadvantage is its lack of generalizability.

Once established, procedural knowledge can become automatized.Automatization is a rather difficult concept because the term is usedat three levels of generality, at least. In the broadest sense, it refers to thewhole process of knowledge change from initial presentation of the rulein declarative format to the final stage of fully spontaneous, effortless,fast, and errorless use of that rule, often without being aware of it any-more. In a narrower sense, it refers to the slow process of reducing errorrate, reaction time, and interference with/from other tasks that takesplace after proceduralization. In the most specific sense, it designates amerely quantitative change in the subcomponents of procedural knowl-edge to the exclusion of any qualitative change or restructuring (i.e.,excluding changes in which small subcomponents make up procedu-ral knowledge at a given stage of skill development or how they worktogether).

Automatization in the last two meanings of the word is characterizedby the power law of practice: regardless of the domain of learning, bothreaction time and error rate decline over time according to a very specificfunction that is mathematically defined as a power function; hence theterm (see esp. Anderson, 2000; Newell & Rosenbloom, 1981). While theexact nature of the processes underlying the shape of this function (e.g.,quantitative vs. qualitative change) is still a matter of debate, and whilesome question the universality of the power law (see esp. Anderson,2000; Delaney, Reder, Staszewski, & Ritter, 1998; Haider & Frensch,2002; Logan, 1988, 1992, 2002; Palmeri, 1997, 1999; Rickard, 1997,

4 Robert M. DeKeyser

1999, 2004), all agree that reaction time and error rate (some studies alsodocument decreased interference with/from simultaneous tasks) declinegradually as a function of practice with a given task.

The more automatized procedural knowledge becomes the clearerthese effects. The transfer found between program reading/writing or,in the case of L2 learning, between production/comprehension skills,then, is explained by the fact that practice in each skill reinforces tosome extent the declarative knowledge that is applicable to both. Theprocedural knowledge, however, is too specific to be transferred fromone skill to another; therefore, the practice effect is highly skill-specific.

It should be pointed out here that automatized knowledge is not exactlythe same as implicit knowledge. While implicit knowledge or implicitmemory is always defined with reference to lack of consciousness orawareness (see, e.g., Reingold & Ray, 2003), absence of awareness isnot a requirement for automaticity. Hence, one can have knowledge thatis implicit but not automatic (because error rate is too high and speedis too low) in cases of incomplete implicit learning (the pattern may bemerely probabilistic, so the learner feels unsure, hesitates, and often getsit wrong). On the other hand, one can have knowledge that is automaticbut not implicit (because the learner has attained high speed and lowerror rate but is still conscious of rules, for instance, because he or she isa language teacher, whether of L1 or L2, or a linguist).

A further question concerns the nature of the practice effect for rel-atively complex tasks: Does it reflect speeding up small components(automatization in the narrowest sense), changing the nature of smallcomponents (restructuring), speeding up the way they work together,or changing the nature of how they work together (strategic change)?Increasingly, researchers find that automatization in the narrowest senseis probably much more limited than often assumed and that attention issubject to a far greater degree of top-down control (Pashler, Johnston, &Ruthruff, 2001, p. 648).

In the same way that it is hard to decide whether and how to breakup a task into components to be practiced separately, it is hard to decidehow often to provide feedback on performance in complex tasks. Wulf,Schmidt, and Deubel (1993) found that constant feedback was betterfor learning fine parameters of a task, whereas intermittent feedback(63 percent of the time) was better for learning the task as a whole. Asthis study involved a perceptual-motor task, it is unclear to what extentits findings would generalize to a cognitive skill such as language learning,however. Moreover, decisions need to be made about when to providefeedback. On the one hand, immediate feedback may disrupt the execu-tion of higher-order routines that are also being learned (cf. Schooler &Anderson 1990, quoted in Anderson, 2000), but on the other hand, feed-back should not be delayed too much because it may be most efficient

Situating the concept of practice 5

when it is provided while the procedural knowledge that led to theerror is still active in memory. Most importantly, perhaps, a substantialamount of evidence suggests that what is best for improving performancein the short run can be worst in the long run, especially for transfer. Lessfrequent feedback leads to less immediate improvement but to better per-formance in the long run, at least for a variety of perceptual and motortasks (Schmidt & Bjork, 1992). Needless to say, eventual performanceand transfer are more important in real life and even to some extent inschool contexts than short-term performance.

The notion of practice in educational psychology

Educators and educational psychologists do not doubt the importance ofpractice. Even during the heydays of the cognitive revolution, Ausubel,Novak, and Hanesian wrote:

Although much significant meaningful learning obviously occurs during initialpresentation of the instructional material, both overlearning and most long-term retention presuppose multiple presentations of trials (practice). Bothlearning process and outcome customarily encompass various qualitative andquantitative changes that take place during these several trials. Learning andretention, therefore, ordinarily imply practice. Such practice, furthermore, istypically specific (restricted to the learning task) and deliberate (intentional).(1978, p. 311)

Much more recently, Bransford, Brown, and Cocking, for instance,stated that in deliberate practice, a student works under a tutor (humanor computer based) to rehearse appropriate practices that enhance perfor-mance (1999, p. 166) and pointed out that deliberate practice can lead toan enormous reduction in the time it takes individuals to reach real-worldperformance criteria. Ericsson and associates (Ericsson, Krampe, &Tesch-Romer, 1993; Ericsson & Charness, 1994; Starkes & Ericsson,2003; see also Ericsson, 1996; Ericsson & Lehmann, 1996) have docu-mented in astounding detail the enormous amount of deliberate practiceit takes to become a truly expert performer such as a world-class musi-cian, chess player, or athlete. Others (e.g., Maguire, Valentine, Wilding, &Kapur, 2003; Wilding & Valentine, 1997; cf. also Ericsson, 2003) havedocumented large effects of deliberate practice in performance of highlyspecific memory tasks.

As is often the case, however, terminology varies from author to author.Legge, for instance, makes a distinction between practice, which maysimply involve using the skills that have been acquired, sometimesimperfectly (1986, p. 228), and training, which involves a deliber-ate scheme to assist learning. On the other hand, Haskell (2001) gives

6 Robert M. DeKeyser

training a more narrow, negative meaning of drilling or teaching recipes,which leads to the well-known lack of transfer. It should be clearthat we will use the term practice in a sense that is both narrowerthan Legges (for we are dealing with activities planned to assist initiallearning of new elements of a language) and much broader than meredrilling (for we include a variety of loosely structured communicativeactivities).

A large part of the educational literature on practice concerns the issueof transfer, whether it be from one classroom task to another or from theclassroom to performance on the job. A central concept here is that oftransfer-appropriate processing: transfer is likely to occur to the extentthat the cognitive operations involved in the new context, task, or testrecapitulate or overlap with those engaged in during initial learning (see,e.g., Whittlesea & Dorken, 1993). Knowledge that is overly contex-tualized can reduce transfer; abstract representations of knowledge canhelp promote transfer (Bransford et al., 1999, p. 41). One particularform that this problem can take is that of trying to teach proceduralknowledge without an adequate declarative base. Singley and Anderson(1989) point out not only that transfer between related skills such asreading and writing computer programs is limited but that where it doeshappen it appears to occur via declarative knowledge of the underlyingrules. If a student only grasps a problem through a limited number ofexamples, learning may be quick but transfer is doubtful. If the principleor rule underlying the examples is thoroughly understood, transfer willbe much easier, but examples are still necessary for establishing usableknowledge (see esp. Anderson, Fincham, & Douglass, 1997). With-out an adequate level of initial learning, transfer cannot be expected.The point seems obvious, but it is often overlooked (Bransford et al.,1999, p. 41).

On the other hand, narrow procedural knowledge, while it is less gen-eralizable, is not only the most efficient in those contexts where it isapplicable but it is also more durable (Healy, King, Clawson, Sinclair,Rickard, Crutcher et al., 1995; Healy, Barshi, Crutcher, Tao, Rickard,Marmie et al., 1998).

The concept of practice in applied linguistics

Few applied linguists have attempted to define what exactly consti-tutes practice. Ellis has been among the most explicit. He states thatpractice . . . involves an attempt to supply the learner with plentifulopportunities for producing targeted structures in controlled and free lan-guage use in order to develop fully proceduralized implicit knowledge(1993, p. 109). This may seem uncontroversial, but what Ellis says in

Situating the concept of practice 7

the next few sentences makes it clear that the concept is far from obvi-ous. Because of the emphasis he puts on procedural knowledge, he takesthe point of view that (production) practice is important for teaching pro-nunciation and formulaic knowledge but not for the teaching of gram-mar rules: What is being challenged here is the traditional role it hasplayed in the teaching of grammatical items (p. 109). This point of view,of course, reflects the Chomskyan distinction between competence andperformance: practice is to improve performance, not to teach compe-tence, the most prototypical form of which is our intuitive knowledge ofgrammar rules. This competence has always been seen as either acquiredor not, but not something gradual; furthermore, once something has beenacquired, Chomskyan theory sees it as available for use in performance,except for certain constraints on the latter which are considered to bebeyond the scope of linguistics proper (see e.g., Chomsky, 1965, p. 3;1980; 1986, chaps. 1 and 2).

The Chomsky/Ellis point of view, however, is at odds with both thecognitive psychology of skill acquisition and much current theorizing inapplied linguistics, most notably with VanPattens theory of input pro-cessing. Cognitive psychologists stress the role of practice in transform-ing declarative/explicit knowledge into procedural/implicit knowledge.Clearly, it is implicit knowledge that corresponds to the Chomskyannotion of linguistic competence, not explicit, and clearly practice isneeded to achieve it, unless one believes it is also acquired completelyimplicitly and Ellis does not seem to believe that: Perhaps we do nothave to bother with trying to teach implicit knowledge directly (2002,p. 234).

VanPattens theory of input processing (e.g., 1996, 2003), on the otherhand, clearly aims at building the procedural knowledge needed for theuse of grammar rules in comprehension after the declarative knowledgeof these rules has been taught explicitly. (VanPattens thinking is clearlyin line with skill acquisition theory on this point; whether this proceduralknowledge transfers easily to production skills is a different matter, aswill be discussed.) Ellis himself appears to have shifted recently toward aless radical competence /performance, rule /item, declarative /procedural,implicit /explicit view: Production, then, may constitute the mechanismthat connects the learners dual systems, enabling movement to occurfrom the memory-based to the rule-based system and vice versa. If thisinterpretation is correct, learners may not be so reliant on input as hasbeen generally assumed in SLA. They may be able to utilize their owninternal resources, via using them in production, to both construct andcomplexify their interlanguages (2003, p. 115). This also seems to bethe position of Diane Larsen-Freeman: Output practice, then, does notsimply serve to increase access to previously acquired knowledge. Doingand learning are synchronous (2003, p. 114).

8 Robert M. DeKeyser

As previously stated, in this book we define practice as specific activ-ities in the second language engaged in systematically, deliberately, withthe goal of developing knowledge of and skills in the second language.Contributors to this volume put different emphases on the importanceof practicing what one already knows in principle (see Legges definitionabove, and the chapters by Muranoi and DeKeyser in this volume) versusdeliberately engaging in tasks that are supposed to draw attention to newphenomena or engender new insights (see esp. the chapters by Leow andMackey in this volume). As Robinson (this volume) points out, the twoprocesses typically go together anyhow. While accessing existing knowl-edge, the learner becomes aware of gaps or inconsistencies in it, whichmay lead to restructuring or expansion of this knowledge, potentially byincorporation of input from a native-speaking interlocutor.

Issues surrounding practice in applied linguistics

While much of the literature quoted in the first two sections of this chap-ter is couched in language that may be unfamiliar to second languageacquisition researchers, parallels abound between the questions asked incognitive and educational psychology and those that bedevil our ownfield. How skill-specific and how task-specific is the effect of practice;in other words, how much transfer can be expected? How much feed-back should be given, how, and at what time to maximize the effect ofpractice? Can explicit knowledge be automatized through practice tothe point of becoming equivalent to implicit acquired knowledge? Howdoes automaticity develop in the course of practice? These issues will beoutlined here and discussed in more depth in the final chapter of thisbook.

Complete answers to these questions are not available in large partbecause empirical research on practice has been quite limited in recentdecades. Between the bad memories of audiolingual mechanical drillsand the subsequent emphases in the 1970s, 1980s, and 1990s on authen-tic communication, focus on meaning, and task-based learning, fewresearchers in the post-audiolingual period have addressed the issue ofpractice head-on (cf. also Larsen-Freeman, 2003, pp. 102 and 106).

The skill-specificity issue is probably the one that has drawn themost attention in applied linguistics lately (see esp. DeKeyser, Salaberry,Robinson, & Harrington, 2002; Izumi, 2002, 2003; Muranoi, thisvolume; VanPatten, 2002a, 2002b, 2003, 2004). Cognitive psychologyhas much to say about the specificity of skills (see esp. Anderson, 1993;Anderson & Fincham, 1994; Anderson, Fincham, & Douglass, 1997;Muller, 1999; Singley & Anderson, 1989), but of course its findings can-not be transferred blindly to issues of second language acquisition. Both

Situating the concept of practice 9

Ellis (1992, 1993, but see the 2003 quote on page 7) and VanPatten (seeesp. VanPatten & Cadierno, 1993; VanPatten & Oikkenon, 1996) takethe view that while input practice leads to acquisition, output practicemerely serves to improve fluency. On the other hand, studies such as thoseby DeKeyser (1997), DeKeyser and Sokalski (1996), and Izumi (2002)clearly show a lack of transfer between receptive and productive skills atthe level of both proceduralized and automatized knowledge.

The problem of transfer, discussed so often in cognitive psychology andeven more in educational psychology, applies also at the broader level oftransfer from declarative knowledge to procedural skill and of knowledgeand skill from one context to another, in particular from the classroom tothe native-speaking environment. Transfer from declarative knowledgeto procedural skill has been discussed very widely (see e.g., Carroll, 2001;DeKeyser, 1997, 2003; Doughty, 2003; Ellis, 1992, 1993, 2003; Hulstijn,2002; Krashen, 1982, 1999; McLaughlin, 1987; Skehan, 1998), and isoften referred to as the interface issue. A typical case of the other kindof transfer, from the classroom context to the real world, is, of course,the semester abroad context. Research on the latter topic illustrates boththat this transfer is far from obvious and that study abroad is not asobviously ideal for practicing foreign language skills as is often assumed(see esp. Brecht & Robinson, 1993; Brecht, Davidson, & Ginsberg, 1995;DeKeyser, this volume). DeKeyser argues that these two transfer issues,from one kind of knowledge and skill to another and from one contextto another, are intertwined in study abroad programs.

Another prominent issue already mentioned is the separation of a com-plex skill into separate components in terms of teaching, practicing, andproviding feedback. Most L2 teaching methodologies of the last 30 years,such as communicative language teaching, the natural approach, andtask-based learning, are much less inclined to take language apart intosmall components than was the case for older methods such as grammar-translation, audiolingualism, or cognitive code. But what exactly the idealpoint is on the analytic/synthetic dimension of curriculum design, andwhat this implies for practice activities, is still far from resolved, espe-cially in the foreign language context (see esp. Ortega, this volume).

The usefulness of feedback in general, and of specific techniques suchas explicit error correction, negotiation of meaning, or recasts has beenthe subject of much debate in applied linguistics but has only recentlybecome the subject of a considerable number of empirical studies (see esp.Iwashita, 2003; Leeman, 2003; Nicholas, Lightbown, & Spada, 2001;Pica & Washburn, 2002). It appears from this literature that feedbacktends to have a substantial positive effect (see esp. Leeman, this volume,and the meta-analysis in Russell & Spada, 2006), but the amount ofempirical evidence gathered so far is insufficient to answer more spe-cific questions about when and how to give feedback with any degree

10 Robert M. DeKeyser

of certainty. Clearly, we need to know more about these questions, andabout others, which have hardly been addressed at all. How useful isfeedback for different elements of language, not just for pronunciationversus grammar versus vocabulary, but even, say, for rules versus itemsversus prototypes, simple rules versus complex rules, frequent items ver-sus infrequent items? And, perhaps most important, how should the fre-quency and the nature of feedback be adapted to the stage of learning orskill acquisition? Much work remains to be done in this area.

Automatization, on the other hand, is an issue that has not drawn muchfocused attention yet, let alone provided the accumulation of evidencethat is needed to guide practice. As argued on page 3, automatization inthe broad sense has many faces. In applied linguistics these are illustrated,for instance, in the work of Healy et al. (1998), who show that strategyshifts from rules to items as well as from items to rules can both occuras a result of ample practice with linguistic structures, and in that ofDeKeyser (1997), who shows that automatization in a more narrow sensecan take place for second language grammar rules following the samepower-function learning curve documented in the acquisition of skillsin other domains. Many questions remain, however, especially aboutthe integration of such findings from a skill acquisition perspective withfindings from the second language acquisition literature.

While the acquisition of complex skills, skill specificity, skill transfer,feedback on performance, and automatization of skills are issues forwhich applied linguists can certainly find much inspiration in the cogni-tive and educational literature, we also face a number of difficult choicesthat are characteristic of second language teaching, such as the relation-ship between form and meaning and the difference between teachingform and teaching forms. Clearly, form-meaning connections are theessence of language, and taking them apart more than necessary forpractice activities would be unwise, but there are areas of language suchas phonetics, phonology, and morphological paradigms where narrowlyfocused, repeated practice activities with forms can be useful (DeKeyser,1998). Such practice activities have traditionally been called drills.They have been alternately advocated, demonized, derided, and resusci-tated, often even without making the distinction between different kindsof drills.

As (talking about) drilling has been so out of fashion for a numberof years, it may be good to remind some younger readers of this vol-ume of the three-way distinction made by several authors after the con-cern for communicative language teaching became well established butbefore it evolved, at least at the level of academic debate, into an almostexclusive focus on meaning. Paulston (1970, 1972; cf. also Paulston &Bruder, 1976) made a three-way distinction between mechanical, mean-ingful, and communicative drills (MMC). Mechanical drills were defined

Situating the concept of practice 11

as drills where there is complete control of the response and only onecorrect way of responding (1976, p. 3). Such drills can be carried outwithout any knowledge of the L2 or even any understanding of the rulebeing practiced by mere superficial analogy; replacing all lexemes withnonsense words would leave the drill intact. Meaningful drills are verydifferent in the sense that the student cannot complete these drills with-out fully understanding structurally and semantically what is being said(p. 7). In other words, they require that the student access the form-meaning links characteristic of the L2 in order to understand what isasked and to respond with an answer that is correct not only in termsof form but also in terms of intended meaning. In communicative drillsthere is still control over the structures used by the student, but the imme-diate goal from the students point of view becomes actual exchange ofinformation: the student is telling the teacher or other students some-thing they did not know. The main difference between a meaningfuldrill and a communicative drill is that in the latter the speaker adds newinformation about the real world (p. 9).

Byrne (1986) also made a three-way distinction between presentation,practice, and production (PPP). It is clear from Byrnes discussion (esp.p. 5), however, that the practice stage roughly corresponds to Andersonsproceduralization stage, and the production stage to automatization. Inother words, PPP is completely different from MMC: presentation pre-cedes MMC, practice combines the mechanical and meaningful, and pro-duction includes but goes beyond the communicative in MMC (becauseit goes beyond drills).

As I have argued previously (DeKeyser, 1998), mechanical drills canonly serve a very limited purpose, because they do not make the learnerengage in what is the essence of language processing, i.e., establishingform-meaning connections. Far too often, however, all drills are equatedwith mechanical drills. The debate between Wong and VanPatten (2003,2004) and Leaver, Rifkin, and Shekhtman (2004) clearly illustrates thispoint: where Wong and VanPatten saved their sharpest criticisms formechanical drills, other kinds of drills appeared to be guilty by associa-tion, which led to the reaction by Leaver, Rifkin, and Shekhtman.

While we did not include the idea of multiple repetition as impliedby drills as a necessary component of our definition of practice (seeabove) any more than the idea of mechanical practice, we definitely donot exclude it either. It is an element of the definition used by cogni-tive psychologists such as Carlson (see above), it continues to be animportant element of the praxis of practice all over the world, and itis gradually gaining respectability again in empirical research on secondlanguage learning (e.g., Bygate, 2001; Gass, Mackey, Alvarez-Torres, &Fernandez-Garca, 1999; Lynch & Maclean, 2000, 2001; Van den Bran-den, 1997; see also Van den Branden, this volume).

12 Robert M. DeKeyser

The real question is not so much whether repeated performance of anarrowly construed task has a role to play in second language learning,but how strongly form-focused activities such as drills should be inte-grated into the curriculum without reverting to a structural syllabus thatmerely teaches the structure of the day or becoming obsessed witha mere focus on forms instead of focus on form (see esp. Doughty &Williams, 1998; Long 1988; Long & Robinson, 1998). Discussions oftask-based learning (Ellis, 2003; Long & Crookes, 1992; Long & Norris,2000; Skehan, 1996; Wesche & Skehan, 2002), in particular, continueto deal with this issue.

Finally, as the table of contents for this book suggests, the need forand usefulness of different kinds of practice varies considerably depend-ing on the institutional context and the characteristics of the individuallearner. Second language teaching, foreign language teaching, bilingualimmersion programs, and study abroad are all characterized by their ownopportunities for and constraints on practice, and students of differentages, with quantitative and qualitative differences in aptitude, can benefitdifferentially from different forms of practice.

Far from being an outdated concept, then, practice stands at the cross-roads where many different questions intersect: questions about the rela-tionship between competence and performance, implicit and explicitlearning, production and comprehension, analytic versus synthetic syl-labi, accuracy versus fluency, and about individualization of instruction.

The structure of this volume

The first section in the book discusses the basic forms of practice and feed-back that are applicable to most circumstances, whether they emphasizeinput or output. The second section deals with variations in institutionalcontext and what impact they have on conditions for different kindsof practice. The third section discusses how practice activities could beadapted to differences in age and aptitude between learners. Short intro-ductions to the issues dealt with in the individual chapters can be foundat the beginning of each of these sections. The final chapter of this bookdraws together the findings from the previous sections against the back-ground of the conceptual distinctions made in the introductory chapterand provides a differentiated view of what constitutes good L2 skill prac-tice. The book ends with a glossary, which provides definitions of termsused in more than one chapter in this book along with some core conceptsfrom individual chapters.

While the concept of practice certainly applies to the four skills oflistening, speaking, reading, and writing, many chapters in this bookfocus on oral skills, largely because that is where the issue of practice has

Situating the concept of practice 13

been perceived as most problematic and has generated the most empiri-cal research and theoretical debate. Even there, however, much remainsto be done. While much SLA research has dealt with oral input, thereis little empirical research on systematic listening practice learning tolisten and listening to learn have been seen as very different issues (cf.Vandergrift, 2004). At present a [second language] course based pri-marily on psychological principles has not been realized (Healy et al.,1998). It is the aim of this book to contribute to such a goal.

References

Anderson, J. R. (1993). Rules of the mind. Hillsdale, NJ: Lawrence Erlbaum.Anderson, J. R. (2000). Learning and memory. An integrated approach (2nd

ed.). New York: John Wiley.Anderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., & Qin,

Y. (2004). An integrated theory of the mind. Psychological Review, 111(4),103660.

Anderson, J. R., & Fincham, J. M. (1994). Acquisition of procedural skillsfrom examples. Journal of Experimental Psychology: Learning, Memoryand Cognition, 20(6), 132240.

Anderson, J. R., Fincham, J. M., & Douglass, S. (1997). The role of examplesand rules in the acquisition of a cognitive skill. Journal of ExperimentalPsychology: Learning, Memory and Cognition, 23(4), 93245.

Ausubel, D. P., Novak, J. D., & Hanesian, H. (1978). Educational psychology:A cognitive view. New York: Holt, Rinehart, and Winston.

Bransford, J. D., Brown, A. L., & Cocking, R. R. (1999). How people learn:Brain, mind, experience, and school. Washington, DC: National AcademyPress.

Brecht, R. D., Davidson, D. E., & Ginsberg, R. B. (1995). Predictors of for-eign language gain during study abroad. In B. Freed (Ed.), Second lan-guage acquisition in a study abroad context (pp. 3766). Amsterdam /Philadelphia: Benjamins.

Brecht, R. D., & Robinson, J. L. (1993). Qualitative analysis of second languageacquisition in study abroad: The ACTR / NFLC project. Washington, DC:National Foreign Language Center.

Bygate, M. (2001). Effects of task repetition on the structure and control oforal language. In M. Bygate, P. Skehan, & M. Swain (Eds.), Researchingpedagogic tasks: Second language learning, teaching, and testing (pp. 2348). New York: Longman.

Byrne, D. (Ed.). (1986). Teaching oral English (2nd ed.). Harlow, UK: Longman.Carlson, R. A. (1997). Experienced cognition. Mahwah, NJ: Lawrence Erlbaum.Carlson, R. A. (2003). Skill learning. In L. Nadel (Ed.), Encyclopedia of cognitive

science (Vol. 4, pp. 3642). London: Macmillan.Carroll, S. E. (2001). Input and evidence. Amsterdam: John Benjamins.Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT

Press.Chomsky, N. (1980). Rules and representations. New York: Columbia Univer-

sity Press.

14 Robert M. DeKeyser

Chomsky, N. (1986). Knowledge of language: Its nature, origin, and use. NewYork: Praeger.

DeKeyser, R. M. (1997). Beyond explicit rule learning: Automatizing secondlanguage morphosyntax. Studies in Second Language Acquisition, 19(2),195221.

DeKeyser, R. M. (1998). Beyond focus on form: Cognitive perspectives on learn-ing and practicing second language grammar. In C. Doughty & J. Williams(Eds.), Focus on form in classroom second language acquisition (pp. 4263).New York: Cambridge University Press.

DeKeyser, R. M. (2003). Implicit and explicit learning. In C. Doughty & M. Long(Eds.), Handbook of second language acquisition (pp. 31348). Oxford:Blackwell.

DeKeyser, R. M., Salaberry, R., Robinson, P., & Harrington, M. (2002). Whatgets processed in processing instruction? A commentary on Bill VanPattensProcessing instruction: An update. Language Learning, 52(4), 80523.

DeKeyser, R. M., & Sokalski, K. J. (1996). The differential role of comprehen-sion and production practice. Language Learning, 46(4), 61342.

Delaney, P. F., Reder, L. M., Staszewski, J. J., & Ritter, F. E. (1998). The strategy-specific nature of improvement: The power law applies by strategy withintask. Psychological Science, 9(1), 17.

Doughty, C. (2003). Instructed SLA: Constraints, compensation, and enhance-ment. In C. Doughty & M. H. Long (Eds.), Handbook of second languageacquisition (pp. 256310). Oxford: Blackwell.

Doughty, C., & Williams, J. (1998). Pedagogical choices in focus on form. InC. Doughty & J. Williams (Eds.), Focus on form in classroom second lan-guage acquisition (pp. 197261). New York: Cambridge University Press.

Ellis, R. (1992). Second language acquisition and language pedagogy. Clevedon,UK: Multilingual Matters.

Ellis, R. (1993). The structural syllabus and second language acquisition. TESOLQuarterly, 27 (1), 91113.

Ellis, R. (1994). The study of second language acquisition. Oxford: OxfordUniversity Press.

Ellis, R. (2002). Does form-focused instruction affect the acquisition of implicitknowledge? Studies in Second Language Acquisition, 24(2), 22336.

Ellis, R. (2003). Task-based language learning and teaching. Oxford: OxfordUniversity Press.

Ericsson, K. A. (Ed.). (1996). The road to excellence: The acquisition of expertperformance in the arts and sciences, sports, and games. Mahwah, NJ:Lawrence Erlbaum.

Ericsson, K. A. (2003). Exceptional memorizers: made, not born. Trends in Cog-nitive Sciences, 7(6), 2335.

Ericsson, K. A., & Charness, N. (1994). Expert performance. American Psychol-ogist, 49(8), 72547.

Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C. (1993). The role of deliberatepractice in the acquisition of expert performance. Psychological Review,100(3), 363406.

Ericsson, K. A., & Lehmann, A. C. (1996). Expert and exceptional performance:Evidence of maximal adaptation to task constraints. Annual Review ofPsychology, 47, 273305.

Situating the concept of practice 15

Gass, S., Mackey, A., Alvarez-Torres, M. J., & Fernandez-Garca, M. (1999).The effects of task repetition on linguistic output. Language Learning, 49(4),54981.

Haider, H., & Frensch, P. A. (2002). Why aggregated learning follows thepower law of practice when individual learning does not: Comment onRickard (1997, 1999), Delaney et al. (1998), and Palmeri (1999). Jour-nal of Experimental Psychology: Learning, memory and cognition, 28(2),392406.

Haskell, R. E. (2001). Transfer of Learning: Cognition, instruction, and reason-ing. New York: Academic Press.

Healy, A. F., King, C. L., Clawson, D. M., Sinclair, G. P., Rickard, T. C., Crutcher,R. J. et al. (1995). Optimizing the long-term retention of skills. In A. F. Healy& L. E. Bourne, Jr. (Eds.), Learning and memory of knowledge and skills.Durability and specificity (pp. 129). Thousand Oaks, CA: Sage.

Healy, A. F., Barshi, I., Crutcher, R. J., Tao, L., Rickard, T. C., Marmie, W. R.et al. (1998). Toward the improvement of training in foreign languages. InA. F. Healy & L. E. J. Bourne (Eds.), Foreign language learning. Psycholin-guistic studies on training and retention (pp. 353). Mahwah, NJ: LawrenceErlbaum.

Hulstijn, J. (2002). Towards a unified account of the representation, acquisi-tion, and automatization of second-language knowledge. Second LanguageResearch, 18(3), 193223.

Iwashita, N. (2003). Negative feedback and positive evidence in task-basedinteraction: Differential effects on L2 developments. Studies in Second Lan-guage Acquisition, 25(1), 136.

Izumi, S. (2002). Output, input enhancement, and the noticing hypothesis:An experimental study on ESL relativization. Studies in Second LanguageAcquisition, 24(4), 54177.

Izumi, S. (2003). Comprehension and production processes in second languagelearning: In search of the psycholinguistic rationale of the output hypothesis.Applied Linguistics, 24(2), 16896.

Krashen, S. D. (1982). Principles and practice in second language acquisition.Englewood Cliffs, NJ: Prentice-Hall.

Krashen, S. D. (1999). Seeking a role for grammar: A review of some recentstudies. Foreign Language Annals, 32(2), 24557.

Larsen-Freeman, D. (2003). Teaching language: From grammar to grammaring.Boston, MA: Heinle.

Leaver, B. L., Rifkin, B., & Shekhtman, B. (2004). Apples and oranges are bothfruit, but they dont taste the same: A response to Wynne Wong and BillVanPatten. Foreign Language Annals, 37(1), 12532.

Lee, F. J., & Anderson, J. R. (2001). Does learning a complex task have to becomplex? A study in learning decomposition. Cognitive Psychology, 42,267316.

Leeman, J. (2003). Recasts and second language development: Beyond negativeevidence. Studies in Second Language Acquisition, 25(1), 3763.

Legge, D. (1986). Skills. In R. Harre & R. Lamb (Eds.), The dictionary of devel-opmental and educational psychology (pp. 2258). Oxford: Blackwell.

Logan, G. D. (1988). Toward an instance theory of automatization. Psycholog-ical Review, 95(4), 492527.

16 Robert M. DeKeyser

Logan, G. D. (1992). Shapes of reaction-time distributions and shapes of learningcurves: A test of the instance theory of automaticity. Journal of Experimen-tal Psychology: Learning, Memory and Cognition, 18(5), 883914.

Logan, G. D. (2002). An instance theory of attention and memory. PsychologicalReview, 109(2), 376400.

Long, M. H. (1988). Instructed interlanguage development. In L. Beebe (Ed.),Issues in second language acquisition: Multiple perspectives (pp. 11541).New York: Harper & Row.

Long, M. H., & Crookes, G. (1992). Three approaches to task-based syllabusdesign. TESOL Quarterly, 26(1), 2756.

Long, M. H., & Norris, J. (2000). Task-based teaching and assessment. InM. Byram (Ed.), Encyclopedia of language teaching (pp. 597603). London:Routledge.

Long, M. H., & Robinson, P. (1998). Focus on form: Theory, research, andpractice. In C. Doughty & J. Williams (Eds.), Focus on form in classroomsecond language acquisition (pp. 1541). New York: Cambridge UniversityPress.

Lynch, T., & Maclean, J. (2000). Exploring the benefits of task repetition andrecycling for classroom language learning. Language Teaching Research,4(3), 22150.

Lynch, T., & Maclean, J. (2001). A case of exercising: Effects of immediate taskrepetition on learners performance. In M. Bygate, P. Skehan & M. Swain(Eds.), Researching pedagogic tasks. Second language learning, teaching,and testing (pp. 14162). New York: Longman.

Maguire, E. A., Valentine, E. R., Wilding, J. M., & Kapur, N. (2003). Routesto remembering: the brains behind superior memory. Nature Neuroscience,6(1), 905.

McLaughlin, B. (1987). Theories of second-language learning. London: EdwardArnold.

Muller, B. (1999). Use specificity of cognitive skills: Evidence for productionrules? Journal of Experimental Psychology, 25(1), 191207.

Newell, A., & Rosenbloom, P. S. (1981). Mechanisms of skill acquisition and thelaw of practice. In J. R. Anderson (Ed.), Cognitive skills and their acquisition(pp. 155). Hillsdale, NJ: Lawrence Erlbaum.

Nicholas, H., Lightbown, P. M., & Spada, N. (2001). Recasts as feedback tolanguage learners. Language Learning, 51(4), 71958.

Palmeri, T. J. (1997). Exemplar similarity and the development of automaticity.Journal of Experimental Psychology: Learning, Memory and Cognition,23(2), 32454.

Palmeri, T. J. (1999). Theories of automaticity and the power law of practice.Journal of Experimental Psychology: Learning, Memory, and Cognition,25(2), 54351.

Pashler, H., Johnston, J. C., & Ruthruff, E. (2001). Attention and performance.Annual Review of Psychology, 52, 62951.

Paulston, C. B. (1970). Structural pattern drills: A classification. Foreign Lan-guage Annals, 4(2), 18793.

Paulston, C. B. (1972). The sequencing of structural pattern drills. TESOL Quar-terly, 6, 197208.

Paulston, C. B., & Bruder, M. N. (1976). Teaching English as a second language:Techniques and procedures. Cambridge, MA: Winthrop.

Situating the concept of practice 17

Pica, T., & Washburn, G. (2002). Negative evidence in language classroomactivities: A study of its availability and accessibility to language learners.Working Papers in Educational Linguistics, 18(1), 128.

Posner, M. I., DiGirolamo, G. J., & Fernandez-Duque, D. (1997). Brain mech-anisms of cognitive skills. Consciousness and Cognition, 6, 26790.

Reingold, E. M., & Ray, C. A. (2003). Implicit cognition. In L. Nadel(Ed.), Encyclopedia of cognitive science (Vol. 2, pp. 4815). London:Macmillan.

Rickard, T. C. (1997). Bending the power law: A CMPL theory of strategyshifts and the automatization of cognitive skills. Journal of ExperimentalPsychology: General, 126(3), 288311.

Rickard, T. C. (1999). A CMPL alternative account of practice effects innumerosity judgments. Journal of Experimental Psychology: Learning,Memory, and Cognition, 25(2), 53242.

Rickard, T. C. (2004). Strategy execution in cognitive skill learning: An item-leveltest of candidate models. Journal of Experimental Psychology: Learning,Memory, and Cognition, 30(1), 6582.

Russell, J., & Spada, N. (2006). The effectiveness of corrective feedbackfor the acquisition of L2 grammar: A meta-analysis of the research.In J. M. Norris & L. Ortega (Eds.), Synthesizing research on lan-guage learning and teaching (pp. 13364). Philadelphia / Amsterdam: JohnBenjamins.

Schooler, L. J., & Anderson, J. R. (1990). The disruptive potential of immediatefeedback. In Proceedings of the 12th annual conference of the CognitiveScience Society (pp. 7028). Cambridge, MA.

Schmidt, R. A., & Bjork, R. A. (1992). New conceptualizations of practice:Common principles in three paradigms suggest new concepts for training.Psychological Science, 3(4), 20717.

Singley, M. K., & Anderson, J. R. (1989). The transfer of cognitive skill. Cam-bridge, MA: Harvard University Press.

Skehan, P. (1996). A framework for the implementation of task-based instruc-tion. Applied Linguistics, 17, 3862.

Skehan, P. (1998). A cognitive approach to language learning. Oxford: OxfordUniversity Press.

Starkes, J. L., & Ericsson, K. A. (Eds.). (2003). Expert performance in sports:Advances in research on sports expertise. Champaign, IL: Human Kinetics.

Van den Branden, K. (1997). Effects of negotiation on language learners output.Language Learning, 47(4), 589636.

Vandergrift, L. (2004). Listening to learn or learning to listen? Annual Reviewof Applied Linguistics, 24, 325.

VanLehn, K. (1989). Problem solving and cognitive skill acquisition. In M. I.Posner (Ed.), Foundations of cognitive science (pp. 52779). Cambridge,MA: MIT Press.

VanPatten, B. (1996). Input processing and grammar instruction: Theory andresearch. Norwood, NJ: Ablex.

VanPatten, B. (2002a). Processing instruction: An update. Language Learning,52(4), 755803.

VanPatten, B. (2002b). Processing the content of input-processing and process-ing instruction research: A response to DeKeyser, Salaberry, Robinson, andHarrington. Language Learning, 52(4), 82531.

18 Robert M. DeKeyser

VanPatten, B. (2003). From input to output: A teachers guide to second languageacquisition. New York: McGraw-Hill.

VanPatten, B. (Ed.). (2004). Processing instruction: Theory, research, and com-mentary. Mahwah, NJ: Lawrence Erlbaum.

VanPatten, B., & Cadierno, T. (1993). Input processing and second languageacquisition: a role for instruction. The Modern Language Journal, 77(1),4557.

VanPatten, B., & Oikkenon, S. (1996). Explanation versus structured input inprocessing instruction. Studies in Second Language Acquisition, 18(4), 495510.

Wesche, M. B., & Skehan, P. (2002). Communicative, task-based, and content-based language instruction. In R. B. Kaplan (Ed.), The Oxford Handbookof Applied Linguistics (pp. 20728). New York: Oxford University Press.

Whittlesea, B. W. A., & Dorken, M. D. (1993). Incidentally, things in general areparticularly determined: An episodic-processing account of implicit learn-ing. Journal of Experimental Psychology: General, 122, 22748.

Wilding, J. M., & Valentine, E. R. (1997). Superior memory. Hove, UK: Psy-chology Press.

Wong, W., & VanPatten, B. (2003). The evidence is IN: Drills are OUT. ForeignLanguage Annals, 36(3), 40323.

Wong, W., & VanPatten, B. (2004). Beyond experience and belief (or, waitingfor the evidence): A reply to Leaver et al.s Apples and Oranges. ForeignLanguage Annals, 37(1), 13342.

Wulf, G., Schmidt, R. A., & Deubel, H. (1993). Reduced feedback frequencyenhances generalized motor program learning but no parameterizationlearning. Journal of Experimental Psychology: Learning, Memory and Cog-nition, 19, 113450.

Yule, G. (1997). Referential communication tasks. Mahwah, NJ: LawrenceErlbaum.

PART I:FOUNDATIONS

The chapters in this section form the foundations of the book in the sensethat they deal with aspects of practice that should hold true for all learn-ers in all contexts. Part II will deal with variations due to institutionalcontexts and Part III with individual differences.

Input and output have been discussed many times in the applied lin-guistics and second language acquisition literature, but only in recentyears has attention been focused on what exactly the role of input andoutput practice is. For centuries, language teaching, whether grammar-translation, audiolingual, cognitive-code, or communicative, had putemphasis on output activities but with very divergent underlying philoso-phies. Krashens monitor model and Krashen and Terrells naturalapproach were radical breaks with that tradition in the sense that theysaw only a minimal role for output practice, seeing output as largelyunproblematic, provided the relevant competence had been acquired.Acquisition of competence, in turn, was viewed as a matter of enoughmeaning-focused processing of the right (comprehensible) kind ofinput, not a matter of systematic practice in the sense of this book (i.e.,specific activities in the second language, engaged in systematically, delib-erately, with the goal of developing knowledge of and skills in the secondlanguage; see Chapter 1), even though some of the activities advocatedby Krashen and Terrell as a means of providing meaningful and compre-hensible input would certainly qualify as practice from our point of view.Since then the profession has seen a return to focus on form, but howmuch of it is necessary, at what stage of the learning process, in particularwhether form should be learned through practice with focus on form, orwhether form should be taught and then practiced, how much practiceis needed in comprehension and how much in production and for whatpurposes continue to be debated.

The chapters in this section illustrate the various answers to thesequestions. Ron Leow describes three different ways of looking at inputpractice: (1) what he calls the attention strand, i.e., the point of viewthat learners should engage in systematic activities that will focus theirattention on form to start the acquisition process; (2) what VanPatten has

19

20 Foundations

called processing instruction, i.e., activities aimed at the acquisition ofform-meaning links through repeated processing of these links in a for-mat that makes them essential to the task, following explicit instruction;and (3) the skill acquisition approach as represented by cognitive psychol-ogists such as Anderson and Newell, which holds that skills in a varietyof domains are acquired by proceduralizing and automatizing declar-ative knowledge, the latter most often being acquired through explicitinstruction of rules accompanied by examples.

Hitoshi Muranoi argues strongly that there is a need for output prac-tice. Drawing on Swains output hypothesis, Levelts speech productionmodel, and Andersons model of skill acquisition, he describes a varietyof functions of output practice: noticing, hypothesis testing, consciousreflection, and automatization. He advocates output practice throughvarious forms of text reconstruction.

Ideally, of course, input and output practice are integrated in the formof interaction between native and non-native speakers. Not only do learn-ers need to practice skills like holding the floor or getting the floor, whichcan only be learned in an interactional context; but as Alison Mackeyschapter shows, learners can also benefit from the specific kinds of com-prehensible input and pushed output that are brought about through thenegotiation of meaning in NS-NNS interaction.

An especially problematic aspect of practice, in particular of interac-tive practice, is the role of feedback. With its renewed focus on form,the profession is less suspicious of feedback than a generation ago, butexactly how and when to provide what kind of feedback continues to bea question of great concern. Jennifer Leeman guides us through the mazeof terminology on this point and argues that, while feedback is clearlybeneficial, it can be hard to determine whether it is enhanced salience,metalinguistic information, or pushed output that plays the biggest role,in particular as this may depend on the elements of language in question.

1 Input in the L2 classroom:An attentional perspectiveon receptive practice

Ronald P. Leow

Introduction

The role of input is undoubtedly crucial in the process of second/foreignlanguage (L2) learning. Input may be defined as the L2 data (form-basedand/or meaning-based) that learners receive either in the formal class-room or in a naturalistic setting. Indeed, how L2 input is presented to L2learners in the classroom and its effects on the processes learners employto interact with the input (input processing) have been the focus of severalstrands of second language acquisition (SLA) studies conducted withina psycholinguistic framework. The theoretical underpinnings of most ofthese psycholinguistic studies appear to include some role for attention(and possibly awareness) in the processing of L2 grammatical or lin-guistic data in adult learners L2 development (e.g., Robinson, 1995;Schmidt, 1990, 1993, 1995, 2001; Tomlin & Villa, 1994; VanPatten,2004).

The term practice has several connotations (see DeKeyser, this vol-ume for an elaborated discussion of this term in several fields of inquiry)in both the applied linguistics and cognitive psychology literatures. Inapplied linguistics, the notion of pedagogical practice in the typical class-room assumes some form of performance by learners in response to L2grammatical input they receive in this setting, which may be providedprior to or during practice. In addition, the L2 input is usually manipu-lated in some form by the teacher. For example, learners may be exposedto L2 that has been carefully selected and manipulated by the teacher tohighlight some linguistic data and may be requested to interact with itin several ways, such as selecting options related to the linguistic data inthe input, performing a task, and so forth.

From an attentional perspective, then, receptive practice will bebroadly defined in this chapter as follows: any exposure1 to manipu-lated L2 input that provides not only various exemplars of targeted L2forms or structures upon which learners attention to (and/or awarenessof) is directly or indirectly premised but also some form of opportu-nity to perform a limited productive or nonproductive task or activity

21

22 Ronald P. Leow

(e.g., selecting one out of two options, completing a problem-solvingtask, translating) during the exposure. The general assumption is that,via receptive practice, learners pay attention to the targeted linguisticdata in the input while processing the grammatical information sub-stantially enough to be capable of recognizing, interpreting, and/or pro-ducing such forms or structures after exposure. By definition, studiesin which learners were required to simply process the L2 input (e.g.,textual or input enhancement, input flooding, some skill acquisitionstudies) or to produce the targeted forms or structures during exposure(see, for example, Muranoi for output, Leeman for oral feedback, andMackey for oral interaction in this volume) have been excluded in thischapter. The ultimate goal of receptive practice is to promote robustinput processing leading to subsequent internalization of the linguis-tic data.

It is important to note that this chapter does not presume that receptivepractice is the only pedagogical avenue for successful L2 development inthe classroom setting given that this type of practice addresses only oneaspect of the acquisitional process, usually referred to as an early stageof processing the incoming L2 data. A wealth of studies have addressed,for example, the role of output in L2 development (e.g., de la Fuente,2002; Izumi & Bigelow, 2000; Izumi, Bigelow, Fujiwara, & Fearnow,1999; Izumi, 2002; Muranoi, this volume; Swain, 1995; Swain & Lapkin,1995), that appears to provide abundant empirical support for such a rolein SLA.

Theoretical underpinnings

Cognitive psychology and cognitive science have provided an explana-tion of or theoretical account for the role cognitive processes play inSLA (e.g., Anderson, 1983, 1990; Bialystok, 1978, 1990, 1994; Carr &Curran, 1994; Ellis, 1993; Gass, 1988; Hulstijn, 1989; Hulstijn &Schmidt, 1994; Robinson, 1995; Schmidt 1990 and elsewhere; SharwoodSmith, 1981, 1986, 1991; Tomlin & Villa, 1994; VanPatten, 1996). Onepremise shared by the fields of cognitive psychology, cognitive science,and SLA is that learning does not take place without attention. Therole of attention is deemed crucial for further long-term memory storageof L2 information to take place (for comprehensive reviews of studiesthat explicate such a role of attention in language learning, see Carr &Curran, 1994; Schmidt, 1990 and elsewhere; Robinson, 1995; Tomlin& Villa, 1994).

Given this role for attention in L2 development, the next section willreport on current SLA research that has employed receptive practice inan effort to promote L2 development in the L2 classroom and situate

An attentional perspective on receptive practice 23

these studies with respect to their theoretical underpinnings and the roleof attention in receptive practice.

Receptive practice in SLA classroom-based research

The 1990s witnessed several strands of research (attentional /attention-focusing tasks, processing instruction [PI], and skill acquisition) thatshared the same purpose of providing receptive practice to learners topromote their intake and subsequent processing of targeted linguisticdata contained in the L2 input. However, the studies seem to differ some-what with respect to the role attention plays in receptive practice. Inthe attentional strand, attention appears to be assigned a primary role(e.g., Rosa & Leow, 2004a, 2004b; Rosa & ONeill, 1999); in the pro-cessing instruction (PI) strand, it has a secondary role (e.g., VanPatten,2004, Wong, 2004a), and in the skill acquisition strand, it is subsumedwithin the theoretical framework (e.g., de Graaff, 1997; DeKeyser, 1997;DeKeyser & Sokalski, 1996). There are also methodological differencesin the timing of the provision or lack thereof of grammatical informationon the L2 linguistic data embedded in the input. For example, some stud-ies, as part of their research designs, provided explicit grammatical infor-mation prior to practice (e.g., Benati, 2004; Cadierno 1995; Cheng, 2002;DeKeyser & Sokalski, 1996; Farley, 2001a, 2001b, 2004; Morgan-Short& Bowden, 2006; Rosa & Leow, 2004a, 2004b; VanPatten & Cadierno,1993; VanPatten & Oikkenon, 1996; VanPatten & Sanz, 1995; VanPat-ten & Wong, 2004), during practice in the form of feedback (e.g., deGraff, 1997; DeKeyser, 1997; Morgan-Short & Bowden, 2006; Rosa &Leow, 2004a, 2004b; Sanz, 2004; Sanz & Morgan-Short, 2004), or noneat all (e.g., Benati, 2004; Farley, 2004; Rosa & ONeill, 1999; Sanz, 2004;Sanz & Morgan-Short, 2004; VanPatten & Wong, 2004b). In addition,whether any research methodology was employed to operationalize andmeasure the construct of attention, that is, whether learners did indeedpay attention to the targeted forms or structures during practice, can beviewed from an offline (e.g., Benati, 2004; Farley, 2004; de Graaff, 1997;DeKeyser, 1997; Sanz, 2004; Sanz & Morgan-Short, 2004; VanPatten &Cadierno, 1993; VanPatten & Oikkenon, 1996; VanPatten & Sanz,1995; VanPatten & Wong, 2004) versus an online (Rosa & Leow, 2004a,2004b; Rosa & ONeill, 1999) data elicitation perspective. Online dataelicitation measures (for example, think-aloud protocols) are employedto gather information on learners internal processes during exposureto and/or interaction with the L2 data. Offline measures are conductedafter exposure and, consequently, can only make inferences as to whetherlearners either paid attention to or became aware of targeted forms orstructures in the input.2

24 Ronald P. Leow

Studies conducted within the attentional strand

The theoretical underpinning of studies conducted within the attentionalstrand that provided receptive practice to adult L2 learners (e.g., Rosa &Leow, 2004a, 2004b; Rosa & ONeill, 1999) is derived from Schmidts(1990 and elsewhere) noticing hypothesis.3 According to this hypothesis,learners need to consciously notice or demonstrate a conscious appre-hension and awareness of some particular form in the input before thatform can be processed further. In Schmidts view, since focal attentionis isomorphic with awareness, there cannot be any dissociation betweenawareness and learning. Schmidt also distinguishes two levels of aware-ness to account for the distinction between intake /item learning andrestructuring /system learning, namely, awareness at the level of noticingand at the level of understanding. At the level of understanding, learnersare able to analyze, compare, test hypotheses, and verbalize the underly-ing rules of the language.4

Studies conducted within an attentional framework have employedconcurrent data elicitation procedures (e.g., think-aloud protocols) toestablish that attention was indeed paid to the targeted forms or struc-tures before its effects were statistically analyzed. In addition, the designof the experimental tasks that provided receptive practice was premisedon Loschky and Bley-Vromans (1993) notion of task-essentialnessthat required learners to pay attention to the targeted grammatical formsin the input in order to successfully complete the task. According toLoschky and Bley-Vroman, the grammatical point itself is the essenceof what is to be attended to (p. 139). Due to the nature of the experimen-tal tasks, implicit feedback was inherently provided during performance(Rosa & ONeill, 1999), while type of feedback (whether explicit orimplicit) was also addressed in Rosa and Leow (2004a, 2004b).

Rosa and ONeill (1999) investigated how intake of Spanish contrary-to-fact conditional sentences was affected both by awareness and by thecondition under which a problem-solving task was performed. Partici-pants were 67 adult L2 learners of Spanish, with no prior knowledge ofthe targeted structure, who were randomly divided into four conditionswith different degrees of explicitness and one control group. Two factorswere varied to create the four experimental conditions: explicit formalinstruction (FI) on Spanish contrary-to-fact conditional sentences (pro-vided in a handout prior to practice) and directions to search for rules(RS) while solving each sentence of the puzzle, yielding the followinggroups: [+FI, +RS], [+FI, RS], [FI, +RS], [FI, RS]. Participantsin the [RS] conditions were requested to memorize the content infor-mation as they solved each sentence of the puzzle. The control groupwas not provided with any instruction prior to completing the puzzle.

An attentional perspective on receptive practice 25

Concurrent data on learners awareness were gathered through the use ofthink-aloud protocols performed while participants were completing theproblem-solving tasks. Intake was measured through a multiple-choicerecognition test administered immediately after practice.

Receptive practice was provided via a problem-solving task (a multiple-choice jigsaw puzzle) divided into two pasted sections on a page: (1) apiece of the puzzle depicting an event, a person, or the result of an eventand (2) another piece of the puzzle with the main clause of a condi-tional sentence of either one of two experimental targeted structures.Each page also had three other pieces of the puzzle each with a subordi-nate clause written on it. Participants were required to select one of thethree unpasted pieces that would correctly fit between the picture and themain subordinate clause. Implicit feedback was inherent in the practice,given that a correct match indicated accuracy. Rosa and ONeill foundthat all the experimental groups, including the control, increased signifi-cantly their intake from the pre-test to the post-test. In other words, theabsence of explicit grammatical presentation of L2 linguistic data did notimpede learners ability to recognize the targeted structure immediatelyafter exposure.

Rosa and ONeill, like Leow (1997, 2001a), reported that awarenessat the level of noticing and at the level of understanding translated intoa significant improvement in intake scores from the pre-test to the post-test. Likewise, results indicated that learners who demonstrated under-standing of the targeted structure performed significantly better on intakeposttests than learners who evidenced noticing only. Finally, Rosa andONeill reported significant relationships between learners in the [+FI]and the [+RS] conditions and awareness at the level of understanding.They suggested that providing learners with formal instruction couldbe an effective way of directing their attention to form, thus promotingthe emergence of high levels of awareness (p. 544) and orienting thelearners to the formal aspects of the input by simply reminding themto search for rules was almost as powerful as presenting them with theactual rules (p. 545).

Rosa and Leow (2004a, 2004b) examined (1) whether exposure to L2input under different computerized task conditions premised on degreeof explicitness had a differential impact on learners ability to recognizeand produce both old and new exemplars of the Spanish past conditionalimmediately after exposure to the input and over time; (2) whether expo-sure to L2 input under different conditions had a differential impact onlearners awareness; and (3) whether different levels of awareness influ-enced learners ability to recognize and produce the targeted structureimmediately after exposure to the input and over time. Degree of explic-itness was manipulated by combining three features: (a) a pretask provid-ing explicit grammatical information, (b) feedback concurrent to input

26 Ronald P. Leow

processing, and (c) in those cases in which feedback was provided, itsnature (i.e., implicit or explicit). In the explicit feedback (EFE) condition,if the answer was correct, the participants received a prompt reinforc-ing the reason their choice was the right one. If the answer was wrong,the participants received a prompt saying that the choice was incorrecttogether with an explanation of the reason. In this case, the participantswere instructed to try again so that they made the correct selection oftime frame for each particular sentence before going on to the next card.For the implicit condition, the participants received only feedback thatindicated whether the answer was right or wrong.

Exposure to the L2 input took place in a computer-based setting bymeans of a series of problem-solving tasks involving manipulation of theinput. The participants in the control condition were exposed to similarinput (a sentence-reading activity) in a computer setting, although not inthe form of a problem-solving task.

One hundred adult learners of Spanish with no prior knowledge of thetargeted structure were randomly exposed to past conditional sentencesunder one of six conditions premised on different degrees of explicitness:EPEFE (explicit pretask + explicit feedback), EPIFE (explicit pretask +implicit feedback), EFE (explicit feedback), EP (explicit pretask), IFE(implicit feedback), and control. Receptive practice was provided viaa jigsaw puzzle, comprising a series of 28 interactive Libra cards5 andrelatively similar in design to Rosa and ONeills puzzle. Duration ofpractice was 28 minutes with an additional 8 minutes allotted for thepretask phase. L2 development was assessed through recognition andcontrolled-production tests containing old and new exemplars of thetargeted structure.

Results indicated that explicit grammatical explanation prior to prac-tice significantly contributed to learners superior ability to recognizeand produce in writing old and new exemplars of the targeted structurewhen compared to the control group, an ability that was maintained for aperiod of three weeks. Among their findings, Rosa and Leow found that,regarding type of task condition, (1) higher degrees of explicitness had amore drastic impact; (2) processing L2 input through a problem-solvingtask proved to be an efficient way of helping learners internalize thetargeted structure; (3) the advantages of processing input under explicitconditions were more evident in production of old and new items and inrecognition of new items; (4) in the case of recognition and production ofnew items, higher levels of accuracy were reached by learners who pro-cessed input through a problem-solving task containing (a) two sourcesof information on the targeted structure (e.g., a pretask and explicit orimplicit feedback) or (b) only one source of information, in the form ofconcurrent explicit feedback; and (5) explicit feedback provided duringinput processing had stronger effects on accuracy than a pretask alone.

An attentional perspective on receptive practice 27

Regarding the role of awareness, Rosa and Leow (2004b) also foundthat (1) higher levels of awareness (i.e., understanding) were associatedwith learning conditions providing an explicit pretask as well as implicitor explicit feedback; (2) higher levels of awareness (i.e., understand-ing) were substantially more effective than lower levels (i.e., noticing)in helping learners recognize and produce novel exemplars of the tar-geted structure; and (3) higher levels of awareness were associated withsophisticated input-processing strategies such as hypothesis formationand testing, as well as with verbal formulation of rules accounting forthe form and function of the targeted structure, corroborating the find-ings of previous studies on the role of awareness in L2 behavior andlearning (e.g., Leow, 1997; Rosa & ONeill, 1999).

Studies conducted within the processinginstruction (PI) strand

The theoretical foundation of processing instruction is based on VanPat-tens (1996) model of input processing (IP)6 (see VanPatten, 2004 for anupdated version of his model), which, in turn, appears to draw from themetaphor of a limited capacity channel or processor (e.g., Broadbent,1958; Kahneman, 1973; Norman, 1968; Treisman, 1964). Capacity the-ories postulate that there is competition for attentional resources to bepaid to incoming information and that what is paid attention to maydepend on the amount of mental effort required to process the incom-ing information. For example, Kahnemans (1973) capacity model ofattention, dependent on the participants state of arousal, postulated theallocation of attentional resources from a pool of cognitive resources toincoming information. Whereas earlier filter theories (e.g., Broadbent,1958; Treisman, 1964; Norman, 1968) viewed an inevitable competi-tion for the allocation of attentional resources for incoming informa-tion, Kahnemans (1973) capacity model allows the possibility of dividingthe allocation of resources to different aspects of incoming information.According to Kahneman, performance may not be negatively affectedonce the state of arousal is adequate and if the task demands are notoverwhelming.

In line with the notion that learners have a limited capacity to processL2 information, PI advocates in its guidelines to present one thing at atime. Providing only one rule at a time can avoid overtaxing learnersprocessing resources and maximize the potential to pay more focusedattention to the targeted form or structure needed for intake (Lee &VanPatten, 1995). According to VanPatten (2004, p. 7), processingimplies that perception and noticing have occurred, but the latter twodo not necessarily imply that a form has been processed (linked with

28 Ronald P. Leow

meaning and/or form). As implied in the previous statement, the role ofattention (noticing) in PI appears to be secondary to the primary goal ofPI, which is to help L2 learners derive richer intake from input by havingthem engage in structured input activities that push them away from thestrategies they normally use to make form-meaning connections (Wong,2004a: 33). Indeed, Wong (2004) appears to view the role of attentionduring structured input activities as an incidental byproduct in thatlearners attention is drawn to the relevant form-meaning connectiononly because the learners are engaging in the structured input activities.This view, however, appears to contradict Sanz (2004) direct referenceto the important role of attention (noticing) in processing instructionwhen she writes that [i]ts input-focused practice is crucially structuredso that learners need to attend to the target grammatical form/structureto understand the meaning and complete the activity (p. 241). Sanzrelates this practice to the notion of Loschky and Bley-Vromans (1993)task-essentialness, which in itself is premised on the role of attentioncrucial for successful completion of a task.

The three main characteristics of PI are (a) grammatical explanationabout the targeted form or structure prior to practice, (b) explicit infor-mation about processing strategies (in which learners attention is explic-itly oriented to what to pay attention to and why), and (c) participationin structured input activities to promote further processing of the inputdata. Structured input activities are so called because the input has beenmanipulated to make the targeted forms or structures more salient (e.g.,sentence-initial) and discourage learners previous incorrect use of a par-ticular processing strategy (e.g., a subject-verb-object [SVO] word orderarguably employed by English-speaking students learning Spanish). Twotypes of structured input activities are used in PI: referential and affective.Referential activities require learners to pay attention to form in orderto get meaning. Learners have a correct or an incorrect answer that pro-vides feedback to the teacher concerning the learners ability to makethe correct form-meaning connection. During these activities, learnersare provided with implicit feedback about whether their answer was cor-rect. Affective activities require learners to provide an affective responsewhile processing information about the real world. There is no correct orincorrect response. One key feature of structured input activities, then,is to have learners respond to the input in some form or fashion to pro-mote active processing of the targeted form or structure while keeping thefocus on meaningful activities. Given the focus on the creation of intakefrom the input, no production of the targeted structure is promoted dur-ing the exposure. Instead, learners are pushed to make form-meaningconnections by requiring them to rely on form or sentence structure tointerpret meaning (Wong, 2004a, p. 37).

An attentional perspective on receptive practice 29

The effects of this instructional treatment have been typically com-pared with traditional instruction, defined as grammatical explana-tion and output practice of a grammatical point, and meaning-basedinstruction.7 Participants were generally from second semester and abovewith some prior knowledge of the targeted forms or structures (Spanishpreverbal direct object pronouns, ser and estar, preterit, present subjunc-tive, conditional; Italian future tense). Two exceptions were the Frenchcausative (VanPatten & Wong, 2004b) and Spanish preverbal directobject pronouns (Morgan-Short & Bowden, 2006). Assessment taskswere an interpretation task (typically a choice of A or B) and a writtenfill-in-the-blank or change-the-verb production task (e.g., Cheng, 2002;Farley, 2001a, 2001b, 2004; Morgan-Short & Bowden, 2006; VanPatten& Cadierno, 1993; VanPatten & Oikkenon, 1996) while VanPatten andSanz (1995) also employed a written video-retelling. The duration ofinstructional exposure was typically two to four days for approximatelytwo hours.

Overall results indicated significant improvement in PI learners per-formances on both the interpretation and written production tasks whencompared to a control group without exposure to the linguistic data.When compared to the traditional output group, the results revealed thatwhile the PI group typically outperformed the traditional group on theinterpretation assessment task, there was usually no significant differencein performance on the production assessment task.

Subsequent PI studies sought to isolate the variable of explicit gram-matical explanation before practice to address its potential effect dur-ing practice (Benati, 2004; Farley, 2004; VanPatten & Oikkenon, 1996;Wong, 2004a, 2004b). VanPatten and Oikkenon replicated VanPattenand Cadierno (1993) at a high school level with the groups exposed toone of three conditions: (a) explicit grammatical explanation only (about1015 minutes per session followed by activities not related to the tar-geted structure), (b) structured input activities only (SI), or