Inner Speech - L2: Thinking Words in a Second Language (Educational Linguistics)

Inner Speech - L2

Educational Linguistics Volume 6

General Editor: Leo van Lier Monterey Institute of International Studies, U.S.A. Editorial Board: Marilda C. Cavalcanti Universidade Estadual de Campinas, Brazil Hilary Janks University of the Witwatersrand, South Africa Claire Kramsch University of California, Berkeley, U.S.A. Alastair Pennycook University of Technology, Sydney, Australia

The Educational Linguistics book series focuses on work that is: innovative, trans-disciplinary, contextualized and critical. In our compartmentalized world of diverse academic fields and disciplines there is a constant tendency to specialize more and more. In academic institutions, at conferences, in journals, and in publications the crossing of disciplinary boundaries is often discouraged. This series is based on the idea that there is a need for studies that break barriers. It is dedicated to innovative studies of language use and language learning in educational settings worldwide. It provides a forum for work that crosses traditional boundaries between theory and practice, between micro and macro, and between native, second and foreign language education. The series also promotes critical work that aims to challenge current practices and offers practical, substantive improvements.

The titles published in this series are listed at the end of this volume.

María C.M. de Guerrero

Inner Speech – L2

Thinking Words in a Second Language

13

María C.M.de Guerrero, Inter American University of Puerto Rico, Metropolitan Campus, Puerto Rico, USA

Library of Congress Cataloging-in-Publication Data Guerrero, María C.M. de Inner speech—L2 : thinking words in a second language / María C.M. de Guerrero. p. cm. (Educational linguistics ; v.6) Includes bibliographical references and index. ISBN-10: 0-387-24577-4 ISBN-13: 9780387245775 e-ISBN-10: 0-387-24578-2 ISBN-13: 9780387245782 1. Second language acquisition. 2. Thought and thinking. I. Title Printed on acid-free paper P118.2.G84 2005 418—dc22

2005043227

¤ 2005 Springer Science+Business Media, Inc. All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, Inc., 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now know or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks and similar terms, even if the are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed in the United States of America. 9 8 7 6 5 4 3 2 1 springeronline.com

SPIN

11381921

To Giovanna and Marcela

CONTENTS

PREFACE ACKNOWLEDGMENTS CHAPTER 1. UNDERSTANDING INNER SPEECH. HISTORICAL AND THEORETICAL FOUNDATIONS A historical overview of the study of inner speech Fundamental principles ofsociocultural theory Social origin of higher mental functions Mediation of higher psychological processes The genetic approach in the analysis of higher mental functions Activity theory Defining and delimiting inner speech Terms associated with the concept of inner speech Verbal thought Thinking in (a) language Language o/thought/language/or thought Intrapersonal communication Self-talk Covert linguistic behavior Mental rehearsal Private speech Conclusion CHAPTER 2. THINKING WORDS IN ONE'S FIRST LANGUAGE. INNER SPEECH: THE LI PERSPECTIVE The sociocultural approach to inner speech The thought-speech connection: Vygotsky's view of inner speech Inner speech as activity: Luria's neuropsychological view Inner speech in the planning of speech production: A. A. Leontiev's view The psychophysiology of inner speech: Sokolov's view Inner speech as intrapersonal communication: Vocate's view Overview of the sociocultural approach to inner speech Cognitive approaches to inner speech: A miscellany A sociocomputational approach to inner speech

xi xvii

1 1 9 10 12 13 13 14 17 17 19 20 21 21 22 22 24 25

27 27 28 38 42 43 47 49 51 51

viii

CONTENTS

Clark's supracommunicative view of inner speech A modularist view of inner speech Information-processing perspectives on inner speech Overview of cognitive approaches to inner speech Brain imaging: Technology in search of inner speech Reviewing the LI literature: Implications for L2 inner speech Conclusion CHAPTER 3. THINKING WORDS IN A SECOND LANGUAGE. INNER SPEECH: THE L2 PERSPECTIVE Inner speech as verbal thought in the L2 Thinking in a second or foreign language Nature of verbal thought in the L2 Inner speech as internalization of the L2 Inner speech in L2 reading and writing Inner speech and mental rehearsal of the L2 Mental rehearsal as a "Din" Mental rehearsal and inner speech development Mental rehearsal as language play Neuroimaging research ofL2 inner speech activity Summary of research on inner speech andL2 learning Conclusion CHAPTER 4. METHODOLOGY OF RESEARCH ON INNER SPEECH. THE CHALLENGE OF STUDYING COVERT VERBAL ACTIVITY The genetic method The study of private speech Advantages and limitations of researching inner speech through private speech Verbal reports Pros and cons of using verbal reports Methods of verbal data collection in the study of inner speech Advantages and limitations of researching inner speech through verbal reports Laboratory tools in the study of inner speech Speech (articulatory) interference Early mechanical devices Electrophysiological techniques Advantages and limitations of mechanical, speech interference, and electrophysiological techniques and their applicability in the study of inner speech Neuroimaging

52 52 53 55 56 57 58

59 60 60 65 72 74 11 78 80 81 82 84 88

89 89 91 94 95 96 97 104 108 108 109 109

111 112

CONTENTS

Advantages and limitations of neuroimaging and its applicability in the study of inner speech The problem of method in the study of inner speech. What to do? Conclusion CHAPTER 5. L2 INNER SPEECH: WHAT LEARNERS SAY Section 1. Inner speech and mental rehearsal of the L2 Participants Data collection and analysis Results and discussion Summary of results Section 2. Early stages ofL2 inner speech development Participants and data collection Content analysis Quantitative analysis Discussion Section 3. Pros and cons of a verbal report methodology in the study ofL2 inner speech The questionnaire The interviews The diary Conclusion

ix

114 116 118 119 119 120 121 122 13 7 138 139 140 144 146 148 149 149 150 151

CHAPTER 6. AN INTEGRATED VIEW OF THE ORIGIN, NATURE, AND DEVELOPMENT OF L2 INNER SPEECH Inner speech as internalization of the L2 Private speech as a transitional phase in the internalization of L2 social speech Early inner speech manifestations in the internalization of an L2 Overview of inner speech as internalization of the L2 Externalization of thought through 12 inner speech Overview of L2 inner speech in the externalization of thought Form andfunctions ofL2 inner speech The structure of L2 inner speech The functions of L2 inner speech Inner speech as mediator of verbal tasks in the L2 L2 inner speech and proficiency Inner speech and the creation of an L2 identity Conclusion

154 161 168 168 175 175 175 177 181 183 184 188

CHAPTER 7. DEVELOPING L2 INNER SPEECH: A PEDAGOGICAL PERSPECTIVE The relationship between teaching andL2 inner speech development Understanding the role of the LI in covert L2 processes

191 191 193

153 153

x

CONTENTS

The role of inner speech in becoming literate in the L2 Learners' strategies for the development ofL2 inner speech Instructional mediation in the internalization of the L2 and development of L2 inner speech Providing opportunities for engagement and participation in L2 external activities Fostering internalization and externalization of the L2 Developing a conceptual foundation in the L2 Raising awareness about inner speech Conclusion

195 197 199 199 200 205 208 210

CHAPTER 8. SYNTHESIS AND DIRECTIONS FOR FURTHER RESEARCH

213

APPENDIX

221

Instructions for keeping the diary (English version)

221

REFERENCES

223

AUTHOR INDEX

241

SUBJECT INDEX

247

PREFACE

According to Vygotsky (1986), The decreasing vocalization of egocentric speech denotes a developing abstraction from sound, the child's new faculty to "think words" instead of pronouncing them. This is the positive meaning of the sinking coefficient of egocentric speech. The downward curve indicates development toward inner speech, (p. 230)

The purpose of this volume is to explore the faculty to "think words," not as the ability to mentally evoke words in the native (or first) language (LI) but as the faculty to conjure up in the mind words in a second language (L2).1 To think words-rather than to pronounce them-is possible through inner speech, a function that humans develop in the course of childhood as they internalize the speech of the social group among which they grow. This means internalizing and being able to conduct inner speech in a particular linguistic code, the LI. But humans, at a very early or more mature age, may also come into contact and interact verbally with speakers of other languages, in classrooms or natural settings. The possibility thus emerges of internalizing an L2 in such a way that inner speech in the L2 might evolve. In this book, it is argued that, given certain conditions of L2 learning, it is possible for learners to attain inner speech in the L2. This book examines the distinctive nature of L2 inner speech and the processes that engender it and characterize its development. Inner speech in the L1 has been substantially investigated, especially after Vygotsky made it a central theme in his book Thought and Language (1986). The same has not occurred in the second language acquisition (SLA) field, perhaps because the construct of inner speech, which is tightly related to a view of language learning as a predominantly social phenomenon, does not seem to fit the theoretical premises and goals of the mainstream SLA information-processing approach. A few attempts to explore inner speech and its connections to L2 learning have been made within the sociocultural theory perspective. These efforts, however, have remained in the form

1 A second language (L2) is understood in this book as a language other than the primary one learned from birth, this one also referred to as the native language or first language (LI). When specific references to a "foreign" language (FL) are made in the literature cited, the term FL will be respected. In addition, at certain points in the discussion, it will be necessary to be more precise about the conditions in which a new language is learned or used. In these cases, the term FL will be employed to denote languages that are learned primarily in the classroom, when contact with target language speakers is mostly limited to the instructional setting or when the learners are not immersed in a community where the target language is spoken.

xii

INNER SPEECH-L2

of isolated articles and short sections in larger volumes.2 A recognition of the crucial role inner speech plays in mediating verbal thought in any language and the absence of an extended and comprehensive treatment of the topic from an L2 perspective have provided the main impetus for writing this book. The book also addresses a need in L2 research for studies that focus on the internal, rather than external, uses of the L2. As Cook (1998) has pointed out, exclusive attention to the social and interactive uses of the L2 misses covert language functions that are vital for L2 users. The book draws mainly from sociocultural theory 3 for insights into the nature, origin, and development of inner speech in the L2. Based on the work of Vygotsky and others, sociocultural theory views the mind as a fundamentally social construct. Central to sociocultural theory is the notion that intellectual activity is rooted in the social world. One of the theory's strongest claims is that higher mental processes are mediated by signs, that is, tools of a psychological nature. Inner speech, or internalized social speech, is the most powerful tool of thought mediation. Children develop inner speech as they first engage in and then internalize the verbal practices of the community. Thus, the child's ability to "think words" has a social and cultural origin. As social, communicative speech is transformed into mental speech for oneself, it undergoes important changes in form and function. It should be clarified from the start that emphasis is given in this volume to the internal (covert) and nonaudible forms of speech for oneself, such as mental rehearsal and internal self-talk, rather than to the external (overt) manifestations of self-directed speech, such as vocalized private speech and audible language play. These phenomena are treated and receive due attention as important aspects in the development of L2 inner speech, but they do not constitute the main focus of this book. The book is intended for researchers, educators, and students in the fields of L2 and FL learning, applied linguistics, language and cognition, and psycholinguistics. Although some previous knowledge of the concept of inner speech and of the tenets of sociocultural theory might be useful to readers, the book provides extensive background information on the historical, theoretical, conceptual, methodological, and empirical bases of the study of inner speech, both from an LI and an L2 perspective, which may facilitate understanding of the main arguments and ideas presented in this book. In particular, researchers interested in the application of sociocultural theory to L2 learning will find in this volume an L2 perspective on one of sociocultural theory's most salient core concepts: how the mind gets to be mediated by an interiorized system of signs, in this case, the L2. The volume includes pedagogical implications and suggestions for the development of inner speech that might be of interest to practicing language teachers as well as teacher educators and students in language teaching programs. The book's critical review of the methods that have been and could be

2 See, for example, Ohta's (2001) treatment of the development of inner speech in L2 learning in her book Second Language Acquisition Processes in the Classroom (pp. 18-21). 3 Rather than pointing to a unified theory of mind, the term "sociocultural theory" is used in this book to refer broadly to a host of approaches inspired in the sociohistorical school of psychology associated with Vygotsky and others. For a more extended discussion of sociocultural theory see Chapter 1.

PREFACE

xiii

applied in studying inner speech might be of value to those researchers interested in pursuing further study of the phenomenon. In summarized form, the book comprises a discussion of the historical and theoretical foundations of the concept of inner speech; a review of studies related to LI and L2 inner speech and its methodology of research; an interpretive account of the origin, nature, and development of L2 inner speech from a sociocultural theory point of view; and various pedagogical implications and suggestions for further research. Chapter 1 provides an overview of the historical trajectory and theoretical foundations underlying conceptualizations of inner speech. In one of the earliest references to inner speech, Plato refers to thinking as the soundless dialogue the soul has with itself, thus establishing the essential link between thinking and speaking-thought and language-that would characterize all other later conceptions of the phenomenon. The philosophical treatment of inner speech as a unit of thought and language has continued until present times, although the main perspective has been from an implicit LI point of view. An important dimension of the phenomenon was brought about by Vygotsky's insistence that inner speech is a derivative of social speech and should thus be examined from a genetic (developmental) point of view. The chapter addresses the significant contribution of Vygotskyan sociocultural theory to the topic of inner speech as well as of other theoretical and methodological approaches. A review of the main lines of research on inner speech is offered. This includes studies in psychology, psychophysiology, neuropsychology, education, communication theory, and the philosophy of language. Chapter 1 also introduces the main principles of sociocultural theory that are most relevant to a study of inner speech and ends with an explanatory section defining inner speech and other constructs frequently associated with it. Chapter 2 focuses on research and theoretical views on inner speech in the LI. The chapter is organized in three main sections. In the first section, the work of representative sociocultural theorists on inner speech is discussed. The section starts with Vygotsky's view of inner speech as the convergence of two distinct lines of development, speech and thought. Synthesizing ideas from various sources, Vygotsky provided a unique and indelible account of the nature of inner speech-its form and functions-as well as of the psycho-social processes in which inner speech is involved: internalization, thought formulation, and externalization. Vygotsky's views are followed in the chapter by those of Luria, a colleague of Vygotsky who chose to concentrate on the neurophysiological aspects of inner speech, that is, on the brain correlates of inner speech activity. Three other sociocultural perspectives in this chapter are A. A. Leontiev's (1981) model of speech production-including an inner programming stage-, Sokolov's (1972) psychophysiological research of inner speech activity, and Vocate's (1994b) approach to inner speech as a form of intrapersonal communication. The second part of the chapter presents cognitive perspectives of inner speech, including Frawley's (1997) sociocomputational theories, Clark's (1998) supracommunicative view of language and cognition, Carruthers's (1996; Carruthers & Boucher, 1998) modularist approach to inner speech, and research within the information-processing paradigm. The third major section of Chapter 2 introduces a

xiv

INNER SPEECH-L2

recent and highly revealing line of investigation on inner speech: the application of brain imaging techniques, such as PET and MRI. To conclude the chapter, a series of questions with implications for L2 inner speech are drawn from the review of the literature on LI inner speech. In Chapter 3, the aim is to review research on inner speech from an L2 perspective. Although there are not many studies dealing strictly with inner speech in the L2, the literature offers a wide range of investigations on related topics. The first area of research to be reviewed is the issue of verbal thought in an L2. A frequent concern among L2 educators is the belief that learners need "think in the L2M to learn the language successfully. The ramifications of this belief as well as viewpoints on the role of the LI and gestures in L2 verbal thought are looked into. An important aspect of verbal thought is the question of conceptual change in the acquisition of an L2 and how this change may affect the learner's access to an L2 or LI in the process of making meaning. The implications of conceptual change for the creation of an L2 identity are also explored. A second area of research is the process of internalization of L2 social speech and its impact on the development of the L2 as a tool for thought. A third area of review is the research linking inner speech and L2 reading and writing. Within this area, Sokolov's (1972) psychophysiological study of inner speech during FL reading stands out. The fourth group of studies reviewed in this chapter deals with mental rehearsal of the L2 and its relationship to inner speech. Mental rehearsal, defined as the "covert practice of the L2," underlies several internal phenomena implicated in the development of L2 inner speech, such as spontaneous playback of the L2 and covert language play. Lastly, the chapter presents studies within the emerging neuroimaging L2 field, already showing how languages are organized in the bilingual brain and neural areas that are involved in L2 inner speech activity. In Chapter 4, the methodological challenge of investigating inner speech is addressed. Because of its covert and elusive nature, inner speech is an exceedingly difficult phenomenon to examine empirically. Vygotsky overcame the problems posed by the inaccessibility and fluidity of inner speech by looking at it from an experimental-developmental viewpoint, a method that allowed him to make inferences about inner speech through the observation of egocentric speech. Following Vygotsky, many have employed the "genetic" method by focusing on private speech, that is, on the ontogenetic predecessor of inner speech. Alternative methods in the study of inner speech are then analyzed. One of the most productive of these is the employment of verbal report data. Verbal report methodology, such as questionnaires, interviews, think-aloud techniques, first-person narratives, learner diaries, and thought-sampling, takes advantage of those aspects of inner speech that are available to intro- or retrospection. Another strong methodological line of research on inner speech that is discussed is the use of laboratory tools and techniques, such as speech interference, electrophysiological measurements, and neuroimaging. The chapter points out the pros and cons of the various methodologies of research and offers ideas on how to deal with the problem of method in further studies of inner speech. Chapter 5, based on the author's verbal report data, provides a classroom learners' perspective on how inner speech in the L2 is developed and experienced. A major

PREFACE

XV

section of this chapter is devoted to describe the purpose, methodology (questionnaires and interviews), and findings of two studies conducted by the author on inner speech and mental rehearsal oftheL2 (Guerrero, 1990/1991,1994; and Guerrero, 1999). The studies sampled a large number of Spanish-speaking English-as-a-second-language (ESL) learners with proficiency levels ranging from the most basic to the most advanced. These studies were instrumental in showing how inner speech in the L2 develops across time as well as in throwing light on the multifunctional nature of mental rehearsal. A second major section of this chapter details another investigation by the author (Guerrero, 2004), this one focusing on the very early stages of L2 inner speech. In this study, verbal report data in the form of learner diaries and stimulated recall feedback were inspected in order to learn about the nature of incipient L2 inner speech and the covert efforts that learners make in internalizing the language. The chapter concludes with a critical appraisal of the methodology utilized in the studies. Chapter 6 takes note of the insights afforded by preceding chapters and attempts to integrate them in a discussion of the origin, nature, and development of inner speech in the L2, coherent with principles of sociocultural theory. L2 inner speech is first viewed as the culmination of a process of internalization of L2 social speech. It is argued that private speech may be for some learners an important phase in this process; however, the interiorization of the L2 seems to occur for the most part as covert activity, including such behaviors as inward repetition of the L2, recall and delayed reprocessing L2 speech, and preparatory use of the L2. The externalization of inner speech in the L2 is also attended. L2 inner speech may be externalized in two ways: as self-regulatory private speech and as overt speech production (speaking and writing). Two complementary hypotheses are presented regarding the extent to which the L2 is implicated in the inner speech processes leading to production: (a) in early stages of L2 development and when L2 instruction is decontextualized, learners will formulate thought first through the medium of the LI and then translate their LI coded thoughts into the L2, and (b) learners who have attained a very high level of development in their L2 and have somehow reconstructed their conceptual and semantic bases through an L2 will be able to mediate their thinking through a wider linguistic foundation consisting of both the LI and the L2. Chapter 6 identifies two "macro" functions of L2 speech for oneself: the cognitive/regulatory or thinking function and the rehearsal one. The role of inner speech as a tool for the processing and performance of verbal tasks in the L2 (reading, writing, listening, and speaking) is also recognized. Chapter 6 attempts to explain the changing nature of L2 inner speech relative to proficiency as a gradual transformation that starts as the inward reproduction of social speech and culminates with its appropriation and conversion into a rich and powerful tool for thought. Finally, the chapter focuses on the way inner speech contributes to the creation of an L2 identity. This aim of Chapter 7 is to discuss the pedagogical implications of the view of L2 inner speech presented in previous chapters for the benefit of educators and other professionals interested in L2 teaching. The chapter starts with the caveat that developing the faculty of "thinking words in another language" is an exceedingly complex achievement and that there are no simple pedagogical formulas for its

xvi

INNER SPEECH-L2

attainment. To understand the role of teachers in the development of L2 inner speech, it is suggested that teachers should first acknowledge the fact that thinking "in" an L2 goes beyond the mere translation into the L2 of thoughts already coded in the LI. To properly "think" in an L2 is to engage the new language in the creative process of concretizing thought in the form of words. Another important issue in this chapter is the need to recognize the LI as a critical cognitive resource that learners will resort to whether teachers discourage it or not. The chapter also examines the role of inner speech in becoming literate in an L2. Learner strategies, such as covert repetition, which learners naturally deploy to internalize the L2, as well as forms of instructional mediation helpful in the development of L2 inner speech are presented. Chapter 8 provides a synthesis of the theoretical ideas, empirical findings, and teaching implications presented in the book and offers suggestions for further research on the topic of "thinking words" in the L2. It is hoped this book will contribute to an overall understanding of how the minds of L2 learners get to be shaped and enriched by the language they are learning and what it means to be able to think words in a language other than the LI.

ACKNOWLEDGMENTS

As I wrote this book, I was privileged to receive the help and encouragement of many individuals, friends, colleagues, and family members. I would like to express my gratitude to all of them. First and foremost, I want to thank Inter American University of Puerto Rico (IAU) for granting me the sabbatical leave that allowed me to write a major portion of the book. Likewise, I extend my appreciation to the administrative and library staff at IAU-Metro for facilitating my research endeavors related the book. I am greatly indebted to various people who generously took the time to read and make comments on various parts of this book: Linda Borer, Pete Brooks, Millie Commander, Mark Darhower, Tim Murphey, Walter Murray, Amy Ohta, Aneta Pavlenko, Elizabeth Platt, Merrill Swain, and Olga Villamil. To Jim Lantolf, my deep gratitude for sharing his expertise in the field of sociocultural theory and for making the journey into the realm of "inner speech" a genuinely mediated learning experience. Special thanks go to several scholars for sending me research that was essential to the book-Beatriz Centeno-Cortes, Fred DiCamilla, Rick Donato, Xavier Gutierrez, Antonio Jimenez, Tim Murphey, Aneta Pavlenko, Jim Lantolf, and Robert W. Schrauf-and to Dr. Sukhi Shergill and Dr. Eraldo Paulesu for their permission to reproduce fMRI and PET images, respectively. I am grateful to Leo van Lier (Educational Linguistics Series General Editor), Renee de Boo, Marie Sheldon, Mary Panarelli, and Deborah Doherty at Springer/Kluwer for their editorial support and assistance. I also thank profusely the anonymous readers of the final manuscript for their helpful suggestions for revision as well as for their positive remarks. Finally, I thank my relatives across the American continent, my extended family in Puerto Rico, and my friends and colleagues at IAU for their continued encouragement and interest in the book, and very especially my husband and two daughters for their love, understanding, and support. Maria C. M. de Guerrero

CHAPTER 1 UNDERSTANDING INNER SPEECH Historical and Theoretical Foundations

The purpose of this chapter is to provide an overview of the study of inner speech that may serve as a conceptual foundation for future chapters. The first half of the chapter attempts to set the historical and theoretical bases of the phenomenon. A brief history of the treatment of inner speech is offered in the hope that getting to know its philosophical roots and the constellation of theoretical ideas surrounding it may help understand current conceptions of the phenomenon. This section also introduces basic principles of the sociocultural theory framework adopted in the book for the interpretation of inner speech processes involved in L2 learning. The second half of the chapter attempts to define and delimit the topic of inner speech as well as clarify some of the terms that are used in relation to it. First, some expressions associating inner speech to thinking, such as "verbal thought," "thinking in (a) language," and "language of/for thought," are discussed. Next, the scope of such terms as "intrapersonal communication," "self-talk," "covert linguistic behavior," and "mental rehearsal," and the role inner speech plays in each of these phenomena are explored. Finally, a differentiation is made between the concepts of "private speech" and "inner speech." A HISTORICAL OVERVIEW OF THE STUDY OF INNER SPEECH Inner speech has a long trajectory in the history of ideas. One of the earliest documented references to inner speech is Plato's famous passage in which Theaetetus asks Socrates what he means by thinking, to which Socrates responds: I mean the conversation which the soul holds with herself in considering anything. [190] I speak of what I scarcely understand; but the soul when thinking appears to me to be just talking-asking questions of herself and answering them, affirming and denying. And when she has arrived at a decision . . . this is called her opinion.... To form an opinion is to speak, and opinion is a word spoken, -I mean, to oneself and in silence, not aloud or to another. (Plato, 1952, p. 538)

Plato's definition of thinking as an internal dialogue with one's soul and as words spoken in silence highlights not only the role of inner speech in thinking but also the

2

CHAPTER 1

dialogic nature of inner speech. The Platonic definition of thinking as dialogue and words has also been understood as an endorsement of the notion that thinking equals speaking (Sokolov, 1972, p. 34). Plato, however, has been traditionally considered an "idealist," for whom ideas as paradigms supposedly existed in pure form apart from the beings they sought to influence. Knowing, for Plato, consisted in being able to remember such paradigms. In this sense, in the act of reasoning through inner or external dialogue, some aspects of the world of ideas the soul had known before incarnating in the material world became accessible to the speaker. Thus, whereas Plato's definition of thinking suggests an equation of thinking with speaking, his view of thought is still an idealist one in that true ideas exist in pure state. Plato, then, is not conflating inner speech with thinking in the above quote but rather making inner dialogue the means to access true knowledge. Plato's musings on thinking and speaking reflect a perennial philosophical debate on the relationship between thought and speech.4 According to Sokolov (1972), this debate can be traced to the Greek philosophers' notion of logos, which blended the concepts of thinking, language, and being. Since then, the history of ideas on the relationship between thinking and speech has polarized into two extreme positions: (1) Thought and speech are identical and therefore thinking is merely speaking without sound, and (2) thought and speech are not related. This last idea implies that thought is sufficient onto itself, that it exists in pure form, and that speech is just the "expression" of thought. In the early 20th century, the notion of the total identification of thought with speech crystallized in the behaviorists' concept of thought as speech minus sound, whereas the idea of the complete separation of speech from thought was the trademark of the Wurzburg school of psychology, for which intellect became "a pure spiritual power" (Vygotsky, 1986, p. 207). These two conceptualizations of the relationship between thought and speech would have great impact on present day views of inner speech. Although the notion of inner speech can be tracked to ancient times, the origin of the term inner speech has been credited to Wilhem von Humboldt, a German linguist who wrote about an "inner form of speech" (Kozulin, 1986, p. 276). Humboldt, however, did not operationally define inner speech (Vocate, 1994b, p. 14). It would be the job of others, such as Max Muller, Alexander Potebnya, and Vygotsky himself, to further develop Humboldt's concept of inner speech. It is known that Humboldt's ideas influenced Muller, another German linguist, who, arguing in defense of the inseparability of thought and language, wrote in 1892: "There is no reason without language. There is no language without reason" (cited in Sokolov, 1972, p. 20). Potebnya, a Ukrainian linguist who was also a follower of Humboldt, however, did not

4 A thorough discussion of the relationship between thinking and speech is a monumental task much beyond the scope and goals of this book. The overview presented here may appear grossly limited in this respect; the intention, however, has been merely to delineate the position that inner speech has occupied in discourses on the relationship between thought and speech since ancient times. For a more thorough discussion of the debate on the relationship between speech and thought throughout the ages, see, for example, Sokolov (1972).

UNDERSTANDING INNER SPEECH

3

equate thought with language. Neither did he think that language is a mere vestment for thought. In a book titled Thought and Language (1926), Potebnya wrote: "Language is not a means of expressing an already formulated thought but a means of creating it" (cited in Sokolov, 1972, p. 22).5 According to Sokolov (1972), the first researchers who systematically investigated inner speech were two French authors, Victor Egger, a philosopher and psychologist, author of the 1881 La Parole Interieure, and Gilbert Ballet, a physician interested in aphasia and author of the 1886 Le Langage Interieur.6 For these scholars, inner speech was not equivalent to thought. Inner speech was simply a vehicle for thought that had no participation in the thinking processes except at the last stage. Egger and Ballet were particularly interested in whether the images in which words were represented in verbal memory were auditory, visual, or motor. Applying introspection to himself, Egger concluded that inner speech was based on auditory images: "My inner speech (maparole interieure)... is a reproduction of my voice" (cited in Sokolov, p. 42). To others, such as Strieker, inner speech consisted of motor representations (Sokolov, p. 42). The question of the nature of mental representations was taken up by psychologists of the Wurzburg school, who, also through introspection, went as far as to deny any role of mental representations in thinking and subsequently to reject the notion of inner speech as necessary for thought (Sokolov, p. 35). Vygotsky, the figure who is today recognized as the major proponent of a theory of inner speech, was prompted by historical circumstances-the 1917 Russian Revolution and ensuing attempts at redesigning scientific theories from a Marxist, historical materialism perspective-to reconceptualize the existing philosophical notions regarding the relationship between thought and speech and the related phenomenon of inner speech (Joravsky, 1989; Kozulin, 1990a). With the exception of Potebnya, whose ideas Vygotsky followed quite closely, Vygotsky found all previous conceptions of inner speech inadequate. In defining the phenomenon, Vygotsky (1986) first disagreed with its conceptualization as verbal memory: "It was in this sense that inner speech was understood by the French authors who tried to find out how words were reproduced in memory-whether as auditory, visual, motor, or synthetic images" (p. 224). To Vygotsky, verbal memory is only part of the phenomenon of inner speech but not all of it. Second, Vygotsky disapproved of the concept of inner speech that equates thinking with speaking in silence. This notion, best represented in Miiller's view of inner speech as "speech minus sound," Watson's definition of inner speech as "sub vocal speech," and the Russian reflexologists's concept of inner speech as a reflex truncated in its motor part (Vygotsky, 1986, p. 225), again was found to be incomplete. Vygotsky also opposed Goldstein's broad definition of inner speech as all the mental processes that precede the act of speaking, including thought, motives, and emotions.

5 Potebnya seemed to have greatly influenced Vygotsky, who not only read Potebnya's book (Kozulin, 1986, p. xv) but also borrowed from the latter the title of his own book Thought and Language (Kozulin, 1986, p. 269). 6 For a historical account of French approaches to inner speech at the end of the 19th century, see Puech (2000).

4

CHAPTER 1

In short, Vygotsky rejected both reductionist and all-encompassing interpretations of inner speech in defense of his own thesis of inner speech as "an autonomous speech function" (p. 248) enabling verbal thought, but being neither identical with thought nor its mere clothing. For Vygotsky, echoing Potebnya, "thought does not express itself in words, but rather realizes itself in them" (p. 251). According to van der Veer and Valsiner (1991), Vygotsky was not truly original in his theory of inner speech, as he extrapolated ideas from various sources and elaborated on them. For his concept of inner speech, Vygotsky drew not only from Humboldt and Potebnya but also from the Russian linguist Jakubinsky, from whom he quoted extensively (and not always acknowledged; see van der Veer & Valsiner, pp. 367-368), in particular when comparing inner to dialogic speech, and from the French psychologist Paulhan, from whom he borrowed the distinction between sense and meaning in discussing the semantic aspect of inner speech. To van der Veer and Valsiner, where Vygotsky was genuinely original was in connecting inner speech to egocentric speech (p. 370), a phenomenon that Piaget had noted but dismissed as inconsequential to the child's mental development. Fundamentally novel, too, in Vygotsky's treatment of inner speech, was his attempt at explaining the problem from the perspective of historical materialism. For Vygotsky (1986), the issue of inner speech was central to an understanding of verbal thought, a capacity that is not innate but rather a historical-cultural product: "Once we acknowledge the historical character of verbal thought, we must consider it subject to all the premises of historical materialism" (pp. 94-95). Van der Veer and Valsiner's comments notwithstanding, then, it is fair to say that Vygotsky's synthesis of ideas was no minor feat, as he managed to provide the first extended, theoretically integrated, and to this day most enduring treatment of inner speech. Vygotsky's theories on language and cognition laid the grounds for a sociohistorical approach to inner speech that would characterize a long line of Soviet scholars, among these Blonskii, Anan'ev, Luria, Galperin, Zhinkin, Baev, Sokolov, A. A. Leontiev, and Ushakova (as discussed in Sokolov, 1972, and Ushakova, 1994).7 Crucial to an understanding of inner speech as a socio-cultural-historical phenomenon would also be the theory of activity, whose best known spokesperson was A. N. Leontiev, a colleague of Vygotsky. Conceptualizing inner speech from an activity theory point of view reinforces the notion of inner speech as a process originating in human, social, practical, and communicative activity rather than as an inherent faculty of the mind. Activity theory would provide the theoretical basis for Luria's neuropsychological studies on inner speech and A. A. Leontiev's psycholinguistic research on the inner programming stages of speech production, both to be discussed at length in Chapter 2.

7 The grouping together of these Soviet scholars under the general umbrella of sociohistorical theory is not meant to suggest that they were in agreement in all respects. Actually, serious divergences and rifts took place among Vygotsky's followers after his death (Kozulin, 1986, 1990a; van der Veer & Valsiner, 1991). Yet, despite disagreements, it is safe to say that all the mentioned scholars shared a belief in inner speech as a phenomenon with social origins and intimately related to higher mental development.

UNDERSTANDING INNER SPEECH

5

While the sociohistorical approach to the problem of speech and thought from the standpoint of dialectical materialism may have succeeded in the former Soviet Union in revealing the failure of mechanistic and idealist conceptions of inner speech, as Sokolov (1972, p. 2) contended, the controversy around the issue appears to be far from settled in other parts of the world. Present day deliberations on the relationship between language and thought and the role played by inner speech in that relationship are still very much alive-if one is to judge from volumes such as Carruthers and Boucher's Language and Thought (1998). To a large extent, what has brought back the debate to center stage has been a renewed interest in consciousness and how language contributes to create the experience of consciousness (Baars, 1997; Carruthers, 1996; Dennett, 1991; Jackendoff, 1987;McCrone, 1999).8 One important issue in this respect is what role inner speech plays in conscious thought and whether consciousness is possible at all without it. Recently, Carruthers and Boucher (1998) made a distinction around the role of language in cognition and thought, with implications for the function of inner speech in this relationship. The question is basically the same old one: Is language independent of thought, or are language and thought inextricably related? According to Carruthers and Boucher (1998), answers to this question fall into two camps: the communicative conception of language and the cognitive conception of language. The communicative perspective, best represented in the ideas of philosophers of mind and language such as Locke, Russell, Grice, Searle, Fodor, Chomsky,9 and Pinker, entails a belief in language as primarily a means of communication, rather than as an essential catalyst of thought. Modern versions of this position uphold a modularist view of mind and language in which language is seen as a separate module acting as an input-output device for central cognition. Thinkers of the communicative tradition tend to see inner speech as a mere encoding/decoding process for pure thoughts. The fundamental thinking operations are carried out in some form of abstract representational or computational language of thought. For Fodor (1975), for example, the "language of thought" is not a natural language. It is Mentalese, an innate representational metalanguage in which the computations, or cognitive processes, are carried out. A variant of this position is the supracommunicative conception of language, which holds thought to be independent from language but regards language as an enhancer or facilitator of some forms of human thought (Carruthers & Boucher, 1998, pp. 8,1415). This is the view held by Clark (1998) and Jackendoff (1996). In this account, language in the form of inner speech helps thinking, rather than constitutes thinking, by stabilizing thought and making it the focus of conscious attention, aiding memory through mental verbal rehearsal, and contributing to self-reflection. The cognitive conception, on the other hand, holds language to be intrinsically 8 That "consciousness is back" is a point well made by Frawley (1997, p. 121). See his summary of recent work on consciousness in various fields of knowledge. 9 When it comes to the function of "thinking in words," Chomsky believed language is a mere conduit for thought expression: "Insofar as we are using language for 'self-communication,' we are simply expressing our thoughts" (1975, p. 57).

6

CHAPTER 1

related to thinking; therein the notion of thinking in natural languages. Carruthers and Boucher (1998) found instances of this position in an assorted group of philosophers such as Leibnitz, Wittgenstein, Davidson, Dummett, Dennett, and Carruthers, as well as psychologists such as Vygotsky, Luria, and Sokolov. There are differences, however, among these thinkers in exactly how language is implicated in thought. Carruthers and Boucher contended there are two main parties: those who think that thought requires language and those who see language as constitutive of thought. This is an important distinction: "It is one thing to say that language is requiredfor, or is a necessary condition o/thought, or certain kinds of thought... and it is quite another to claim that language itself is constitutively involved in those thoughts, or is the medium of those thoughts" (Carruthers & Boucher, 1998, p. 1). If one followed Carruthers and Boucher's reasoning, it would appear that Vygotsky would fit the requirement category rather than the constitutive one. Actually, as will be seen in depth in Chapter 2, Vygotsky believed that language, a sociocultural product, is essential for thinking, only, however, as it pertains to verbal thinking. For Vygotsky (1986), thought is not identical to language, and nonverbal forms of thinking do exist. Thus, in his view, inner speech is "a distinct plane of verbal thought" (italics added, p. 148), with unique structural and functional characteristics that differentiate it from external speech and pure thought. Carruthers (1996, 1998a, 1998b), on the other hand, would represent the other category of the cognitive conception of language. Carruthers believed language to be constitutively involved in thinking, at least in conscious propositional thinking. For him (1996), in propositional thought, inner speech is the thinking. His position is that certain types of thought cannot be entertained at all unless by deployment of natural language sentences, via inner speech. Furthermore, unlike Vygotsky, Carruthers did not regard language and mind as cultural constructs; in fact, Carruthers endorsed a cognitive (but not sociocultural) view of language and mind that is basically modularist and nativist. Beyond the mainly philosophical-and quite polemic-treatment of inner speech, as illustrated in the thinkers mentioned above, the study of inner speech has branched out into multiple disciplinary directions in present times. Inner speech is discussed in practically every field of knowledge where thought and language cross each other. Thus, there has been extensive treatment of it in the cognitive sciences, in areas such as psycholinguistics, neurolinguistics, and cognitive psychophysiology, and in applied disciplines such as speech communication, speech pathology, and language education. Much of the research involving inner speech in the cognitive sciences adopts the mainstream information-processing psychological perspective, with its metaphor of mind as computer and its multiple store model of memory. Within this framework, inner speech is typically taken to be a process consisting of articulatory, subvocal rehearsal and auditory images preserved in memory. Some attempts, however, have been made in the cognitive sciences to integrate principles of sociocultural theory, as for example in the work of Frawley (1997), for whom the socially derived language for thought (which subsumes private and inner speech, p. 183) frames the biologically given language of thought. Inner speech is a critical issue in neurolinguistics, psychiatry, speech pathology, and clinical psychology where speech-thinking disorders

UNDERSTANDING INNER SPEECH

7

are investigated; thus, it is common to find studies of inner speech in relation to aphasia, stuttering, deafness, schizophrenia, and mental conditions characterized by disturbances in inner speech as a form of self-talk, such as anxiety, depression, gambling, and agoraphobia (Morin, 1993). A major line of research has been the cognitive psychophysiology of covert verbal responses, where inner speech is best known as covert linguistic or verbal behavior. The study of covert verbal behavior within cognitive psychology constitutes an attempt to explicate the role of covert physiological activity at the level of speech musculature and body organs, including the brain, in the processing of higher mental functions. Two theoretical frameworks have informed work in this area: the so-called reflex theories of thought, best illustrated in the work of Bechterev, Sechenov, and Pavlov, and behaviorism, as first espoused by Watson and later reformulated by Skinner and Lashley.10 (For thorough discussions of this research, see McGuigan, 1978, and Sokolov, 1972). According to McGuigan (1978), "the early behaviorists were probably the primary force advocating theories of thought in which muscle responding was critical" (p. 9). This perspective was to influence two major representatives of the physiological approach to inner speech in the Soviet Union, Sokolov and Luria. In Luria's case, what seems to have attracted him to the work of reflexologists and behaviorists was "their analysis of the neurophysiological mechanisms involved in psychological processes" (Wertsch, 1981, p. 11). Both Sokolov and Luria managed to integrate psychophysiology into the overarching theoretical framework of Vygotskyan sociocultural theory. Crucial for the development of research in the area of cognitive psychophysiology have been advances in technology that have provided refined tools for the detection and measurement of observable covert verbal behavior (McGuigan, 1978). A major contribution of this line of research on inner speech has been in solidly establishing the material (bodily) correlates of verbal thinking and hence in dissolving the notion that thinking and speech are two disconnected realms. As Korba (1989) put it, a study of inner speech from the psychophysiological point of view can overcome "the traditional mind/body dichotomy through an examination of the 'interfunctional relationship' between the higher mental processes of the cortical speech areas (the mind), and the proprioceptive speech musculature and speech specific neural structures and pathways (the body)" (p. 218). Inner speech is also an important construct within the speech communication field, more precisely within the field of intrapersonal communication. In communication theory, the notion of intrapersonal communication was differentiated from interpersonal communication in 1953 by Ruesch, Block, and Bennett (cited in Dance & Larson, 1976, p. 30). Discussion of inner speech as a form of intrapersonal

10 Lashley was interested in investigating the physiological correlates of higher mental processes, as the following quote indicates: "I once invented a thought-reading machine; a system of levers which magnified and recorded movements of the tongue. I first had my subjects speak a word, then think it silently. Movements of the tongue were minute but otherwise identical with those of speech" (cited in McGuigan, 1978, p. 52).

8

CHAPTER 1

communication, however, appeared rather late, in the 80s and 90s, making its way into various studies, some of which have been collected in Roberts and Watson's (1989), Vocate's (1994a), and Aitken and Shedletsky's (1997) edited volumes. One of the most thorough discussions of inner speech as a theoretical construct within the field of intrapersonal communication was forwarded by Vocate (1994b), who drew support for her notions of inner speech as an intrapersonal phenomenon from the sociocultural theories of Vygotsky and Luria and social scientist G. H. Mead. Vocate distinguished nuances of meaning among terms frequently associated with inner speech: self-talk, intrapersonal communication, and inner speech itself. Most other treatments of intrapersonal communication, however, fall under two broad categories: the traditional information processing view of communication and the "healthy self school of communication, which focuses on the role of inner speech in the formation and maintenance of a healthy emotional self (Hikins, 1989). Not surprisingly, given the importance of inner speech in language learning and cognition, several studies on inner speech have been conducted in the area of educational psychology. Researchers have investigated children's development of inner speech and its role in cognitive performance at school. Specifically, research has explored the role of inner speech as a mediator in reading (Beggs & Howarth, 1985; Hardyck & Petrinovich, 1970; Yaden, 1984), writing (Moffet, 1982, 1985; Trimbur, 1987) and mathematical problem solving (Rohrkemper, 1986). Development of awareness of inner speech and the possibility of teaching children to use inner speech have also been the focus of research (Liva, Fijalkow, & Fijalkow, 1994; Otte, 2001). Historically, the study of inner speech has been conducted primarily from an LI perspective. Before the 1990s, most of the efforts in studying inner speech from an L2 stance had taken place in the former Soviet Union. Notable among these are Sokolov' s (1972) psychophysiological research on inner speech involving the reading of FL language texts and Zachesova' s and Ushakova' s (see Ushakova, 1994) experiments on the inner speech mechanisms of artificial language acquisition. Outside the Soviet Union, only a few sporadic, non-empirical linkages between inner speech and L2/FL learning (Hellmich & Esser, 1975; Rohrer, 1987) had been made in the literature before Vygotskyan sociocultural theory started to make an impact on L2 research. Two early sociocultural theory L2 studies (Frawley & Lantolf, 1985; Lantolf & Frawley, 1984) called attention to features of inner speech externalized in the private speech of L2 learners. In the 1990s two major empirical studies drawing support from Vygotsky's sociocultural theory were conducted on mental rehearsal as phenomenon related to L2 inner speech (Guerrero, 1990/1991, 1994, 1999). These studies, based on retrospective questionnaires and interviews, were among the first to provide evidence that L2 learners do experience inner speech in the L2. Several studies produced in the late 20th and early 21 st centuries have also yielded important insights into L2 inner speech processes. Worth mentioning among these studies is the research in the areas of verbal thought among bilinguals (John-Steiner, 1985b), preferences for a language of thought among L2 learners (Cohen, 1998), private speech (see, for example, Lantolf & Yanez, 2003; McCafferty, 1994a, 1994b; Ohta, 2001), and language play (Lantolf, 1997). One of the most important issues

UNDERSTANDING INNER SPEECH

9

within the L2 inner speech research is the role played by the LI in inner speech processes. Evidence from research in the areas of L2 vocabulary acquisition (Ushakova, 1994), symbolic gestures (McCafferty, 1998; McCafferty & Ahmed, 2000), reading (Upton & Lee-Thompson, 2001), and writing (Huh, 2002), suggests that inner speech is strongly influenced by LI semantics. However, the possibility that very advanced L2 learners may build a solid conceptual base in the L2 and even shift from an LI to an L2 inner voice was suggested by research among late bilinguals becoming members of an L2 community (Larsen, Schrauf, Fromholt, & Rubin, 2002; Pavlenko & Lantolf, 2000). The empirical study of inner speech has forever constituted a methodological challenge. As a covert, intangible, elusive, and highly dynamic phenomenon, inner speech has always escaped direct observation, and its investigation has therefore remained at a theoretical, mostly philosophical level. Major breaks into the empirical study of inner speech in the 20th have been, however, the Vygotskyan genetic method, which pursues inner speech through observation of its developmental precursor-private speech, and the psychophysiological study of inner speech, which focuses on the observable physiological manifestations of covert verbal behavior. Another important methodological contribution to the study of inner speech was made by the 1960s cognitive revolution in psychology, which brought with it a legitimized use of introspective methods of data collection. Finally, towards the end of the 20th century, outstanding technological advances made possible the observation of "live" inner speech through brain scanning and imaging techniques such as PET (positron emission tomography) and MRI (magnetic resonance imaging). It is likely that new technological breakthroughs will bring further exploration of inner speech from a neuropsychological perspective, particularly with respect to its anatomical features and the on-line neural processes associated with its activity. Especially interesting will be to find out, through the new imaging techniques, developmental patterns of inner speech and brain correlates of L2 inner speech, areas which remain practically uncharted territory as of today. At the same time, as sociocultural theory continues to solidify its foothold in psychology and language learning, research into inner speech should profit from a more encompassing view of the phenomenon as a manifestation of the uniquely human, artifact-mediated, social mind and not just as a purely psycholinguistic, memory-facilitating mechanism inherent to brain functioning. FUNDAMENTAL PRINCIPLES OF SOCIOCULTURAL THEORY The term sociocultural theory has come to stand in Western circles (see, for example, John-Steiner, Panofsky, & Smith, 1994; Kozulin, Gindis, Ageyev, & Miller, 2003; Lantolf, 2000a; Wertsch, 1991) for a variety of approaches to language and cognition all acknowledging the social, cultural, and historical roots of mental functioning. Within a sociocultural theory framework, the mind and its most essential feature-consciousness-are predominantly social constructs. Learning, within this theory, is also a fundamentally social act that cannot be dissociated from the specific cultural, institutional, and historical context where it takes place. Sociocultural theory

10

CHAPTER 1

has its origins in the cultural-historical or sociohistorical school of psychology (see Cole, 1990; van der Veer & Valsiner, 1991; Wertsch, 1991) developed by Vygotsky and a number of colleagues. During the 20th century Vygotskyan principles increasingly made way into scientific and educational thought, inspiring scholars both inside and outside the former Soviet Union (Kozulin et al., 2003). In the 80s, researchers in the L2 field began to adopt sociocultural theory concepts, leading to what is now known as the "sociocultural theory approach" to L2 learning.11 This book draws heavily from sociocultural theory, as a non-reductionist framework that most profitably helps understand and explain the main object of analysis-inner speech in the L2-in its complex social and individual dimensions. The following (by force, reduced) account introduces some major sociocultural theory tenets that are relevant to a discussion of inner speech in future chapters: (a) the social origin of higher mental functions and the related concept of internalization, (b) tool and sign mediation, (c) the genetic analysis of higher mental functions, and (d) basic principles of activity theory. Social Origin of Higher Mental Functions Vygotsky argued that higher psychological processes, such as voluntary attention, logical memory, and rational thought, originate in the social sphere, in the interpersonal relationships that people hold with one another and in the interaction with the cultural, physical and symbolic tools available in the social world. The development of higher psychological processes starts early in life as children internalize interpersonal processes and transform them into intrapersonal ones. This principle is captured in Vygotsky's famous "law of general development": Every function in the child's cultural development appears twice: first, on the social level, and later, on the individual level; first between people (interpsychological), and then inside the child (intrapsycho logical). This applies equally to voluntary attention, to logical memory, and to the formation of concepts. All the higher functions originate as actual relations between human individuals. (1978, p. 57)

Internalization is an essential aspect of the development of higher psychological functions. Vygotsky (1978) conceived internalization as the process by which external sign-mediated activity is reconstructed on the internal plane and begins to operate as higher mental activity. The reconstruction process, Vygotsky insisted, means the transformation, not just the transferal, of interpersonal activity into intrapersonal activity and the creation of a psychological plane that did not exist before. To clarify this point, A. N. Leontiev (1981) expressed: "The process of internalization is not the transferal of an external activity to a preexisting, internal 'plane of consciousness': it is the process in which the internal plane informed" (p. 57). The transformation of external into internal activity by necessity implies a series of developmental changes in the function and structure of social forms of behavior. To understand this process, 11 The following sources provide excellent reviews and original research on sociocultural theory as applied to L2 learning: Johnson, 2004; Lantolf, 2000a; Lantolf & Appel, 1994; and Lantolf & Pavlenko, 1995.

UNDERSTANDING INNER SPEECH

11

Vygotsky (1978, p. 57) offered the example of speech development. In the Vygotskyan view, the first speech of the child is communicative rather than intellectual; speech for the young child is an external form of social interaction and control. Gradually, speech begins to acquire a second function; it becomes a means for cognitive self-regulation, a vehicle for thinking and not just for communication. Social speech turns into egocentric speech, a transitional phase that marks the beginning of speech internalization and is characterized by less than intelligible and frequently abbreviated language. Egocentric speech serves foremost a private intellectual function and is self-directed, but it is still spoken out loud. Eventually, egocentric speech sheds its last social feature, vocalization, and turns inwards as inner speech. The transition from social speech to inner speech thus transforms not just the function of speech, from predominantly communicative to predominantly intellectual, but its very structure. Because inner speech is speech for the self, it is vocally imperceptible, syntactically reduced, and semantically condensed. Galperin (1967), who pursued the notion of internalization in depth, postulated that two different forms or stages of speech for oneself are involved in the formation of higher mental processes or actions "in the mind": external speech for oneself and inner speech. "The first form of 'action in the mind' is ordinary speech but without the volume, i.e., 'external speech to oneself" (p. 30). However, this is just a transitory step while the psychological action is being formed. When the action is mastered and automatized, external speech for oneself is no longer viable. It is too "protracted and slow" (p. 31). Speech for oneself becomes condensed into verbal meanings and the action transfers to the plane of inner speech. According to Galperin, however, the "essence" of former planes of sign mediated-action is never lost (p. 32). At moments of cognitive difficulty, it is possible to revert to more external and unfolded modes of action. Frawley and Lantolf (1985) referred to this aspect of mental activity as the principle of continuous access (see also Lantolf & Appel, 1994). Adults, for example, sometimes externalize their inner speech in the form of private (audible) speech when performing a challenging task. Some scholars have proposed the notion of appropriation as a way of conceiving the construct of internalization (Leontiev, A. N., 1981; Newman, Griffin, & Cole, 1989; Wertsch, 1995, 1998). Wertsch (1998) argues that the notion of internalization may mistakenly suggest an inadequate opposition between internal and external processes reminiscent of old Cartesian mind-body dualism. The construct of internalization may also lead to the erroneous belief that all forms of mediated action occur at an internal plane, a point that Vygotsky (1978) clearly established as false: "For many functions, the stage of external signs lasts forever, that is, it is their final stage of development" (p. 57). Rather than substituting the term internalization with other terms, Wertsch (1998) proposed to envision the construct in terms of mastery and appropriation. The notion of mastery allows for the possibility that some forms of mediated action may never turn inwards and disappear out of sight. For many people, certain mathematical or statistical processes, for example, are never fully internalized although there may be complete mastery over them through external mediation of a calculator. In turn, the notion of appropriation, which Wertsch used in the Bakhtinian

12

CHAPTER 1

sense of making one's own something that belongs to others, implies that some processes are not easily reduced to the individual plane. In the case of language appropriation, for example, words are always "half someone else's" (Bakthin, 1981, p. 293). The process of making someone else's language one's own implies that words will always reflect to some extent the semantic intentions of others. Appropriation also frequently entails resistance: resistance from words because they do not easily yield their own established intentions and resistance from speakers who will impose on words their own volitions. This process of mutual resistance and tension between words and speakers is very frequent in L2 learning. Lantolf (2000b) described L2 learning as a process of appropriation "through which the individual takes in particular features of the language through privately practising and experimenting with these features" (p. 88). Specifically, for Lantolf, private speech is a form of mediation in the appropriation of an L2. An important concept within the issue of internalization is that of the "zone of proximal development" (ZPD). Vygotsky envisioned the ZPD as a psychological site, created through dialogic interaction between the child and more mature or expert partners, where the child's functions that are in the process of maturing have the potential to grow and develop. It is in the ZPD where children begin to internalize or carry out internally, through inner rather than external speech, those cognitive operations that were first carried out externally in conjunction with others and socially created artifacts. The notion of the ZPD has been interpreted in the literature as applying not just to child-adult, face-to-face interactions but also any type of noviceexpert, peer-peer, adult-adult, or individual-cultural artifact relationship where potential cognitive development and internalization of social modes of thinking may occur. Mediation of Higher Psychological Processes Sociocultural theory posits that the human mind is culturally mediated. In other words, higher forms of thinking (voluntary attention, logical reasoning, remembering, planning, problem solving) do not develop naturally as part of predetermined development but evolve from lower forms (elementary perception, involuntary attention, natural memory) as the individual comes into contact with and internalizes social forms of activity, which serve as mediators of mental activity. Vygotsky (1978) argued that the mind does not act upon reality or apprehend it directly but does so indirectly through signs. Signs, in contrast to physical tools that mediate labor and manipulation of nature, are psychological tools of mediation. Sign-mediated activity includes primarily the use of language but also of other semiotic tools, such as gestures, mnemonic techniques, mathematical symbols, and diagrams. These signs vary not only from culture to culture but also in the course of history, hence the historico-cultural embeddedness of signs and the consciousness they mediate. A major proposition, then, of sociocultural theory is that the adult human mind is fundamentally the result of sociocultural mediation and not a product of natural development. This does not mean that sociocultural theory denies the existence of

UNDERSTANDING INNER SPEECH

13

innate, biological requisites and constraints on psychological processes. Actually, one of the most important developments of the theory was demonstrating the neurological and psychophysiological bases of higher mental functions, an endeavor that fell upon researchers such as Luria(1973,1981) and Sokolov(1972). What sociocultural theory argues is that nature provides the necessary starting point, that is, the capacity to operate at the level of lower psychological functions. Culture, in turn, through social activity and tool mediation, provides for the transformation of the biological givens into higher mental functions. In this respect, as John-Steiner et al. (1994) stated, sociocultural theory "stresses the unification of nature and culture..., the interweaving of the biological and the social" (p. 4). The Genetic Approach in the Analysis of Higher Mental Functions A cornerstone of sociocultural theory is its approach to the study of mind and consciousness from a genetic or developmental point of view. Vygotsky (1978) argued that to truly understand the essence of higher psychological functions, they need to be studied in their formation, as processes that change over time and not as stable properties of the mind. This implies both ontogenetic and microgenetic observation (Wertsch, 1985b). Ontogenesis reveals the development of an organism or mental function from inception to full development, allowing for the observation of multiple causal forces (biological, cultural, and historical). Microgenesis, on the other hand, focuses on processes as they occur in real time. It usually involves observing changes that occur "right before one's eyes" (Vygotsky 1978, p. 61) either spontaneously or, in some cases, as a result of experimental manipulation. Both ontogenetic and microgenetic analyses imply explanation and not just description. In other words, what is aimed at is the reconstruction of the origins and stages of development of mental phenomena. This does not mean, according to Vygotsky, that description of the external, objective aspects of psychological events should be discarded; however, such description should be subordinated to a study of their underlying "causal-dynamic relations"(p. 63). Activity Theory An important complement to sociocultural theory is the theory of activity, an outgrowth of Vygotsky's original proposals, whose major exponent was A. N. Leontiev (1981). Activity theory attempts to explain what people do in particular cultural, institutional, and historical settings and the motivations-biological or social-for people to do what they do. In other words, activity theory provides an account of the why, where, when, and how of people's social and mental behavior (Lantolf & Pavlenko, 2001; Wertsch, 1998). A. N. Leontiev (1981) established three levels on which to examine human

14

CHAPTER 1

activity.12 The first is the level of activity in general, which is determined by people's motives. Activities may respond to purely biological needs or may be driven by social and cultural forces. The second level corresponds to individual actions and the goals to which actions are directed. The third level focuses on the concrete operations that are carried out and the particular tools of mediation employed under specific environmental conditions.13 Any activity, including L2 learning, cannot be fully understood if these components (motives, goals, conditions, and mediational means) are not taken into consideration. Activity theory is extremely relevant to L2 research because it provides explanation for many of its most pressing concerns: what makes learners learn an L2, what accounts for differential results in acquisition, and how learners approach the task of learning, to name a few. One of the most important contributions of activity theory to L2 learning-or any type of learning, for that matter-is that it assigns learners with agency, that is, the capacity to establish personal goals, set up conditions, and choose the means that best suit their motives or needs in learning (Lantolf & Pavlenko, 2001, p. 145). Another important contribution is in explaining differential outcomes in L2 acquisition. Two students, for example, who take the same class, are offered the same opportunities for learning, and are even engaged in the same learning tasks, may end up obtaining very different results because their motives, goals, and strategies in approaching the learning experience may have been very different (Coughlan & Duff, 1994; Gillette, 1994; A. A. Leontiev, 1981). By the same token, similar outcomes in learning a language may be the result of very different educational experiences. In this respect, activity theory is particularly effective in reminding us that students' actions are always embedded in particular learning contexts and are influenced by particular learning histories. DEFINING AND DELIMITING INNER SPEECH Even though inner speech has been characterized in many ways, interestingly, from ancient to recent times the notion that it is a silent manifestation of speech directed to the self\\&$ remained constant. Notice the similitude between Plato's description of inner speech as "a word spoken . . . to oneself and in silence" (1952, p. 538) and Frawley's (1997) contemporary portrayal of inner speech as "silent speech for oneself (p. 95). Three elements are essential in this characterization: First, inner speech is

12

Vygotsky and A. N. Leontiev differed on what constitutes the chief means of mediation in the development of higher psychological functions. Whereas for Vygotsky it was semiotic mediation, for A. N. Leontiev, it was social activity. Because going into the relative merits of each proposal would take us too far afield, it might suffice to say that an interesting middle ground-tool-mediated action-has been advanced by Wertsch (1985b) following Zinchenko. 13 Several activity theory notions, such as motive, goal, and mediated activity, had already been present in Vygotsky's account of human intellectual functions but became much more explicitly addressed in A. N. Leontiev's theory after Vygotsky's death. Many of the explanations of inner speech that followed Vygotsky are expressed in activity theory terms (Galperin, 1967; A. A. Leontiev, 1981; Luria, 1981).

UNDERSTANDING INNER SPEECH

15

spoken language,14 that is, it is language in action rather than language as an abstraction; second, it is silent, that is, it cannot be heard by people who are in the presence of the person experiencing it; and third, it has an orientation to the self, that is, it serves private rather than public purposes. Whereas these fundamental entailments remain at the basis of most, if not all, definitions of inner speech, it does not seem to be an easy concept to reduce to a few words. Vygotsky (1986) himself referred to inner speech in a multiplicity of ways: as "an entirely separate speech function" (p. 235), a "mental draft," (p. 243), "inner dialogue" (p. 243), "practically wordless 'communication'" (p. 243), "speech almost without words" (p. 244), "a distinct plane of verbal thought" (p. 248), "thought connected with words," (p. 249), and "thinking in pure meanings" (p. 249). Sokolov (1972) defined it as "soundless, mental speech, arising at the instant we think about something, plan or solve problems in our mind, recall books read or conversations heard, read and write silently" (p.l). He also called it "concealed verbalization" and "the speech mechanism of thinking" (p.l). Korba (1989) referred to it as "covert, m/rapersonal language behavior" (p. 219) whereas Morin (1993) equated it with "selftalk" or "internal dialogue" (p. 223). In other instances, inner speech has been interpreted as a "voice in the head" (Beggs & Howarth, 1985, p. 396) and as a rehearsal mechanism supporting interaction between the "inner ear" (auditory imagery) and the "inner voice" (subvocalization) (Smith, Reisberg, & Wilson, 1992). When defining inner speech, it might be useful not just to include modality in its definition but to characterize it functionally as well as developmentally.15 In other words, inner speech is not simply a silent form of self-directed speech; it is, furthermore, an instrument for thought resulting from the internalization of social speech. Functionally, two main aspects of inner speech emerge from the different treatments and approaches to inner speech: its cognitive, thinking function and its communicative, talking-to-self role. Whereas inner speech always implies some form of cognitive activity, it may not always be characterized as self-talk. As Vocate (1994b) maintains, for internal self-talk to occur, inner speech must be sufficiently organized in linguistic form, that is, semantically and syntactically elaborated, in order to sustain a conscious dialogue with the self. This notion of self-talk as a distinct mode of inner speech implies a view of inner speech as a phenomenon that occurs in stages or levels of processing, that is, as a progression or movement from thought to external speech and vice versa, involving various levels of semantic and syntactic coding. At some point in this progression, inner speech may be closer to thought than 14 Characterizing inner speech as spoken does not necessarily make it audible. The "spoken" element means that inner speech, though silent to an outsider, is always linked to speech kinesthesis even when articulation is maximally inhibited. Sokolov's (1972) research demonstrated that inner speech is accompanied by motor speech reactions that become reduced as mental tasks get automatized but never fully disappear (p. 263; see also, pp. 121, 153). Lately, neuroimaging research has confirmed the activation of motor speech areas of the brain during the performance of silent language tasks (see Chapter 2). 15 The author thanks an anonymous reviewer for suggesting revising the definition of inner speech along these lines.

16

CHAPTER 1

speech; this might be the stage characterized by Vygotsky (1986) as "thinking in pure meanings" (p. 249) or by Sokolov (1972) as "thinking in allusions to words" (p. 122). At another point, inner speech might be more expanded and specific in syntax and meaning; this might be the more propitious stage for "inner talking" or self-talk, as Sokolov (1972) and Vocate (1994b) respectively claim. Developmentally, inner speech is the result of the process of interiorization of social speech, a process that starts early in life in the close interaction with speaking human beings. The Vygotskyan belief, endorsed in this book, is that the first speech of the child is social and communicative; it is loud speech for others whose main purpose is communication. Gradually, social speech begins to acquire a second function, a thinking function, as the child learns to apply external language for intramental purposes. (This social speech for self-regulatory purposes was known in Vygotskyan times as egocentric speech, but is now more commonly referred to in the literature as private speech.) At some point in their development, children drop audible vocalization from their private speech, thus beginning the transition into inner speech. What marks the capacity for inner speech, however, is not so much the absence of vocalization but the development of the ability to think in reduced syntactic and semantic complexes. From an ontogenetic point of view, then, the development of inner speech is a process of condensation of social into inner speech. Not everything is ontogenetic development about inner speech, however. Inner speech is also on-line speech activity. And whereas it is true that inner speech must be syntactically reduced and semantically condensed for the mind to operate with a maximum of efficiency during automatized verbally-mediated tasks, it is nonetheless a fact that inner speech may at times require a more elaborate and unfolded structure, with high levels of sematicization, lexicalization, and grammaticization, for example, when leisurely self-talking (as discussed above) or when thinking over what to write. It is important to stress the notion that inner speech is activity, on-going inner speakz>2g or language using, rather than a set of language structures or an inner language system. Inner speech as activity is the conceptualization that characterizes the Vygotskyan sociocultural approach to inner speech. As Wertsch (1979a) pointed out, when discussing private or inner speech, Vygotsky "was specifically interested in speech rather than language" (p. 3); he "was mainly concerned with emphasizing the social activity of speech or speaking rather than the structure of the language system" (p. 4).16 Understanding inner speech as on-going speech activity was also crucial in Luria's neuropsychological study of inner speech (see, for example, 1981) and in A. A. Leontiev's (1981) psycholinguistic approach to it. This point is important to keep 16 The fact that Vygotsky was interested in speech as activity rather than in language as a system is reflected in the original Russian title of his book Myshlenie i Rech', where recW means speech as opposed to yazyk, meaning language. Similarly, the word myshlenie means thinking rather than thought, for which there is another word, mysi. A more accurate English translation for the title of Vygotsky's book would have been Thinking and Speech and not Thought and Language (Kozulin, 1986, p. lvii; Wertsch, 1979a, p. 3). Vygotsky's view of inner speech as activity and process rather than as a language foundation was snared by Luria, who, as Vocate (1987) pointed out, "was very explicit in talking about inner speech as a process rather than an entity having any morphological characteristics" ( p. 157).

UNDERSTANDING INNER SPEECH

17

in mind because discussion of the structural aspects of inner speech (its abbreviated syntax and condensed semantics, for example) may erroneously lead to a conceptualization of inner speech as some type of internal language entity, a body of language knowledge, a collection of word forms, or a network of meanings, existing somewhere in the recesses of the mind. The activity of inner speech recruits internalized (or external) linguistic resources-and is in itself internalized (social speech) activity-but does not constitute those resources itself. From an activity point of view, therefore, inner speech might best be conceived as utilizing language for thinking or verbal thinking in action. In future chapters, the possibility of developing what might be called "inner speech in the L2," that is, the capacity to draw from linguistic resources in the L2 in the ongoing and dynamic activity of thinking through words, will be explored. What follows is some necessary deliberation on terms frequently associated with the concept of inner speech but sufficiently distinct to merit separate attention. TERMS ASSOCIATED WITH THE CONCEPT OF INNER SPEECH Verbal Thought As has been seen in the historical overview at the beginning of this chapter, there are multiple interpretations of the relationship between speech and thought and the role of inner speech in this relationship. The position to be adopted in this book is in agreement with the Vygotskyan view of speech and thought as distinct, separate entities intersecting in verbal thought. In verbal thought, speech becomes a tool for thinking and thinking adopts a symbolic structure. The mechanism that makes possible verbal thought is inner speech. Within the framework of sociocultural theory, verbal thought is ontogenetically a product of social interaction, of contact and participation with other speaking beings. An individual is capable of verbal thought when he/she internalizes language-a social, communicative, symbolic tool-and thought becomes mediated by language. Vygotsky (1986) described this as the capacity to "think words" (p. 230).17 The egocentric speech of a child already denotes verbal thinking; it is thinking mediated by words that has not yet lost its vocal, social character. As the child fully internalizes language and social discursive forms of reasoning, verbal thinking begins to operate at an internal plane, via inner speech. At times of cognitively demanding situations, verbal thinking may become overt again and materialize itself as private speech. A crucial element in the internalization of social speech in the child is the development of word meanings, a process that leads to the formation of the essential, conceptual and symbolic basis of human thought (Luria, 1973, p. 326). It is this process that marks the transition from "lower," natural forms of thinking to "higher," cultural verbal thought. For Vygotsky, all the natural mental processes, such as 17

Kozulin (1990a; 1990b) has pointed out that Vygotsky's frequent use of the term "word," as in the expression "to think words," seems to stand as a synecdoche for verbal discourse in general.

18

CHAPTER 1

memory, attention, perception, thinking, volition, and emotion remain as "lower" psychological functions until they are transformed into "higher" mental functions by the mediating effect of semiotic tools or artifacts, such as language, gestures, and mnemonic techniques (Wertsch, 1985b, pp. 24-27). Just as physical tools effectively mediate practical activity, psychological tools-language and other sign systems-intercede in "higher" mental activity. Verbal thought is symbolic thinking mediated by inner speech. As a speech mechanism that relies predominantly on condensed semantics and reduced syntax, inner speech is particularly effective in reducing dim, amorphous thought to a concrete and refined level of word meanings and syntactic logic. Vygotsky (1986) argued that the transition from nonverbal to verbal thought is primarily enabled by word meanings. Word meanings help establish the semantic base of verbal thought, first in the form of subjective, idiosyncratic senses, and then in more objective and precise cultural meanings. The condensation of thought into word meanings is an inner speech activity, described by Vygotsky as "to a large extent thinking in pure meanings" (p. 249) and by Sokolov (1972) as "thinking in allusions to words" (p. 121). In the externalization of thought, word meanings is the first stage of verbal thought (in speech comprehension, it is the last). From then on, verbal thought progresses through what Vocate (1994b) called "inner speech coding" and Sokolov (1972) called "inner talking" (p. 121), a process of semantic synthesis and syntactic expansion that culminates in fully elaborated syntactic, semantic, and phonological expression. Language, or rather, languages are then responsible for imposing a particular structure on thought. It is probably in this sense that Vygotsky (1997) stated that "we always think in some language" (p. 171). The main interest in this book is to explore the process by which the L2 may also provide structure for verbal thought and to what extent inner speech, through word meanings or inner talk, is responsible for the capacity to think verbally in the L2. Of course, not all thought is verbal or mediated by inner speech. Some forms of human behavior do not require the presence of verbal thought: Object manipulation in practical activity and unplanned behavior usually do not involve verbal mediation. According to Vocate (1994b), however, all symbolic-"higher"-forms of thinking are verbal to the extent that they deploy word meanings in their initial stage, regardless of the ultimate mode in which they are rendered, whether visual, musical, kinesthetic, or other. In other words, the fact that word meaning is the essential component of verbal thought does not mean that verbal thought is always coded in "words" or linguistic structure (Vocate, 1994b, p. 18). The semantic basis of verbal thought may be realized through various symbolic modes-graphic, numerical, linguistic, gestural-in the "coding" stage of inner speech. Research conducted on abstract gestures (those representing ideas) tends to give weight to Vocate's claim. As McNeill (1987) and McCafferty and Ahmed (2000) observed, speech and gesture seem to be strongly interconnected at the level of inner speech, combining imagery, as provided by gestures, and syntactic form, as supplied by speech. The fact that gestures frequently accompany speech to maximize expression and are sometimes even more effective in fulfilling the semantic potential of thought than speech itself suggests that verbal

UNDERSTANDING INNER SPEECH

19

thinking is symbolic rather than purely linguistic. Despite the logic behind the argument positing verbal thinking as the essence of all symbolic thinking, it should be pointed out that the customary conceptualization of verbal thought is restricted to the linguistic mode, with all other modes-imagistic, musical, mathematical, or other-viewed as nonverbal and therefore not mediated by inner speech (John-Steiner, 1985a; Keller & Keller, 1996). Thinking in (a) Language When people say that they think "in language" or "in a particular language," they are usually referring to the idea that their thoughts appear in their minds as words. It is a purely introspective belief based on the subjective impression that thoughts in their minds have the characteristic ring, form, and dialogicity of speech. The term "to think in words" or "to think in language" is an expression loosely used by scholars and common people alike to refer in general to the notion of verbal thought. Sokolov (1972), who explored deeply the psychophysiological aspects of inner speech, stated: Inner speech . . . is of great importance to psychology as a whole, chiefly because of its close connection to thought. In thinking over silently some question, in comparing and generalizing the data of the problem being solved, we often notice that we utter to ourselves separate words and occasionally fragmentary phrases. At times, moreover, especially in solving difficult problems, we enter into a kind of discussion with ourselves: we formulate mentally a number of propositions, criticize them from various points of view, and finally select one of them, rejecting the rest. We then say that we "think in words. " (Italics added, p. 34)

As the above passage suggests, "thinking in words" is the everyday way of referring to inner speech. As soon as one begins to ponder on the expression, however, the ambivalence of such expression reveals itself. "To think in words" may be a way of saying that thinking and speaking are one and the same and that thought is identical with speech. This was the behaviorists' notion of inner speech as "speech minus sound" that Vygotsky rejected, as well as the more recent cognitive conception of inner speech as constitutive of thought, exemplified in Carruthers's (1996) pronouncement that "inner thinking is mostly done in inner speech" [italics added] and, more specifically, in "natural language sentences" (p. 50). As noted before, Vygotsky (1997) himself expressed: "We always think in some language" (p. 171). Within the context of his ideas, however, Vygotsky' notion of "thinking in some language," most probably refers to the organizing power of language in general, and in particular of specific languages, to shape and give structure to thought, without necessarily subscribing to any isomorphism of language and thought. The idea of "thinking in a language" is frequently found in discussions of verbal thought in the context of L2 learning. It is quite common, in fact, to hear L2 students say "I think in my native language when writing" or "I try to think in [my L2] when I am alone" just as it is frequent to hear teachers advising students to think in the L2. As early as 1957, for example, Eugene Nida, stated: "thinking in a foreign language is absolutely essential. One cannot expect to speak with facility while going through

20

CHAPTER 1

the process of translating ideas from English into another language" (p. 24). Stern (1980) identified "learning to think in [an L2]" (p. 67) as one of the ten strategies of a "good" language learner. As A. A. Leontiev (1981) pointed out (and will be discussed at length in Chapter 3), the notion of thinking in an FL entails an equation of thought with communication (see pp. 103-109). That is, it reduces thinking to only its final stage before externalization, the stage when thought in the form of word meanings is translated into an objective linguistic code. To Leontiev, this is a limited approach to the notion of "thinking in a foreign language," which should be substituted by the recognition that speech is, from the moment of thought inception, creatively involved in human reasoning and not just a process of attaching verbal labels to thoughts (p. 109). Like Leontiev, John-Steiner (1985b) believed the common understanding of the expression "to think in a foreign language" is a simplistic one: "When a native speaker of Spanish says in great frustration, 'I cannot think in English,' he refers to those planning processes that immediately precede an utterance and that are frequently subvocal in nature" (p. 364). Language of Thought/Language for Thought The phrase "language of thought" is Fodor's (1975) well known nomenclature for the abstract representational system in which cognitive processes or mental "computations" are carried out. The language of thought is an innate mentalese that antecedes the learning of any natural language. The term "language of thought" in the Fodorian sense, therefore, should not be taken to represent the linguistic code of any recognizable language-say French, Dutch, or Mandarin-used as a vehicle for communication intra- and interpersonally. Though Fodor (1975) employed terms such as "internal language" and "private language" to alternatively refer to the language of thought, these terms are not equivalent to either "inner speech" or "private speech," in the sociocultural theory sense. Frawley (1997) made a distinction between the terms "language of thought" and "language for thought." Whereas the language of thought is the logical, representational system, biologically determined to take charge of mental computations, the language for thought is public or socially-derived language that frames the language of thought in culturally-specific ways. "The language for thought is a vehicle for thinking, not a higher-order language of thought" (Frawley, 1997, p. 182). The language for thought operates through private or inner speech at the level of metaconscious awareness, that is, at a level in which awareness is directed at itself. The language for thought is particularly effective for self-conscious thinking, or metaconscious awareness, because of its fundamental property of self-referentiality or reflexivity, which allows thinking to focus upon itself. As with the expression "to think in a language," the phrase "language of thought" is sometimes used in connection to L2 learning. Cohen (1998), for example, used the expression "language of thought" to refer to the linguistic code, LI or L2, in which or through which learners conduct their thinking. Thus, one can speak of Spanish or Polish as the language of learners' thoughts. In this sense, the phrase is not at all

UNDERSTANDING INNER SPEECH

21

related to the Fodorian notion of language of thought as a kind of mentalese. Intrapersonal Communication Intrapersonal communication is best understood as a level of human communication in which the communicator is at the same time the speaker (or sign-generator) and the addressee. All human beings, normally equipped of language and intellect, engage in intrapersonal communication. As Korba (1989) put it, we are all "intralocutors" (p. 237). Intrapersonal communication does not necessarily have to be silent or internal. Overt, private speech also qualifies as intrapersonal communication. Vocate (1994b) explained the roles of self-talk and inner speech in intrapersonal communication. Self-talk is a phenomenon of intrapersonal communication that can take place internally or externally. In both forms, whether silently or out loud, self-talk is generated by a communicator whose audience is himself or herself. Inner speech, by definition, is an internal intrapersonal process. Because inner speech is a process intimately related to language production and reception, inner speech supports all levels of human communication, the intrapersonal, the interpersonal, and the public (Vocate, 1994b). (See also definitions of intrapersonal communication in Cunningham, 1989; Dance, 1994; Johnson, 1984.) Self-talk The belief that humans talk or hold dialogues with themselves in the privacy of their minds is at least as ancient as Plato. This generalized and perpetuated perception that words can be spoken in silence has made talking to oneself, or self-talk, the concept most frequently and readily associated with inner speech. Self-talk, however, is not the same as inner speech. In the first place, self-talk can be both silent and audible; it can be both internal and external (Vocate, 1994b). (In this book, the main interest is in the internal, silent mode of self-talk.) In the second place, talking to oneself is only one way in which inner speech is deployed. As has been seen, inner speech is a more encompassing phenomenon than self-talk. Internal self-talk is a silent, dialogic mode of inner speech, a mode that is characterized by self-awareness and self-interaction (Vocate, 1994b). It is important to emphasize that self-talk and inner speech are not coterminous. Inner speech is a cognitive process of transforming thoughts into words, or vice versa, which involves far more than just talking to oneself. In self-talk, the syntax and semantics of inner speech are quite elaborate and expanded and the pace of thoughts is rather slow, sufficiently so to allow for a "conversation" of sorts to be conducted. Some stages of inner speech, however, are characterized by extreme condensation in meaning and structure and great swiftness, all factors that defy the possibility of reflection or self-awareness, a necessary condition for self-talk to take place.

22

CHAPTER 1

Covert Linguistic Behavior In the psychophysiological literature, the terms inner speech and covert linguistic (or verbal) behavior are frequently used to refer to the same event, inner speech highlighting the psychological aspect of the phenomenon and covert linguistic behavior the physiological (Korba, 1989). Covert linguistic behavior involves subvocal activity, that is, concealed activation of the speech mechanisms, both at the muscular and the neural levels, but it may also implicate other physiological responses associated with language-mediated cognitive activity, such as eye movement and heart rate (McGuigan, 1978). As discussed in the historical overview in this chapter, the study of covert linguistic behavior is an offspring of early reflexological and behavioristic approaches to psychology, which not only reduced all psychological processes to behavior but also restricted the proper study of psychology to that of observable behavior (see McGuigan, 1978, for a summary of these theories); therein the importance of studying covert linguistic behavior through the observation of its measurable physiological responses-lips and tongue movements, finger and arm activation, electrical activity of speech areas in the brain, and the like. Chapter 4 delves into some of the most common methods of measuring the physiological aspects of covert verbal behavior, including the recent brain scanning and imaging techniques. Although the physiological concomitants of inner speech that characterize it as covert linguistic behavior are recognized in this book, in no way is inner speech reduced to mere subvocal speech. As discussed earlier, inner speech is more than just "speech minus sound;" it is internalized, condensed social speech enabling and giving structure to verbal thought. Mental Rehearsal One of the most frequent inner speech activities humans engage in is mental rehearsal. Mental rehearsal may be defined as the covert, spontaneous or deliberate, repetition of words, letters, or other verbal material. Mental rehearsal is an important selfregulatory mechanism implicated in a wide variety of higher, symbolically-mediated, cognitive processes, such as planning, remembering, reviewing past experiences, and learning in general. Rehearsal does not simply imply rote repetition. Psychologists recognize two kinds of mental rehearsal. One is maintenance rehearsal, which relies on simple repetition and is effective in keeping the information in memory for a short period of time, as for example when repeating a name over and over while looking it up in the phone directory. The other kind is elaborative rehearsal, which works with the material to be remembered at various levels of semantic analysis18 and therefore contributes to

18

Semantic analysis is said to vary in terms of "depth" (Craik & Lockhart, 1972). The more meaning is extracted from the material to be remembered the deeper the analysis that results.

UNDERSTANDING INNER SPEECH

23

longer-lasting retention (Bransford, 1979; Craik & Lockhart, 1972; Houston, 1986).19 Elaborative rehearsal is a complex cognitive process that may involve going repeatedly over some verbal material in preparation for speech production or as recapitulation of some linguistic event (Honeycutt, Zagacki, & Edwards, 1989; Smith, 1983). Such forms of rehearsal are said to perform preparatory and review functions, respectively. Rehearsal of verbal material may be deliberate, that is, the person may engage in it with the intent to remember what is rehearsed, or it may be spontaneous, that is, a word, a phrase, a song, may suddenly appear and ring insistently in a person's mind. Information-processing models of memory assign rehearsal an important role in learning. According to the "working memory" model (Baddeley, 1986; Gathercole & Baddeley, 1993), one of the components of short-term working memory is the "phonological loop," a system that operates by holding information in phonological form and maintaining it afresh by means of an articulatory or subvocal rehearsal process. In this model, the auditory imagery and subvocal articulation that accompany the handling of verbal (or verbally coded) material in working memory constitute inner speech processes sometimes referred to as the "inner ear" and the "inner voice," respectively (Baddeley & Lewis, 1981; Campbell, 1992; Smith etal, 1992). Research has highlighted the importance of auditory imagery and articulatory rehearsal in language learning. The exact role of these covert processes in various language related operations, such as reading or learning new vocabulary, is, however, still not clearly understood and continues to be the focus of much research and speculation in the literature. In the cognitive sciences, mental rehearsal has been acknowledged as a crucial process in consciousness, by which contents in the form of words are recursively called forth for further scrutiny and processing: "We build elaborate systems of mnemonic association We refine our resources by incessant rehearsal and tinkering, turning our brains . . . into a huge structured network of competencies The principle [sic] components of this technology for brain-manipulation are words" (Dennett, 1998, p. 292). Clark (1998) stressed the importance of mental rehearsal in making external symbolic resources available for cognitive exploitation. He refers to "inner rehearsals" as "models of linguistic productions" (p. 169). "We follow someone's vocal instructions as we learn to windsurf. Or we mentally rehearse such instruction as we practice on our own. Such phenomena reveal linguistic formulations as somehow helping us to focus, monitor, and control behaviour" (p. 173). Ultimately, mental rehearsal allows thinking about thinking, or metacognition. Citing Jackendoff, Clark stated, "the mental rehearsal of sentences may be the primary means by which our own thoughts are able to become objects of further attention and reflection" (p. 177). It is known that L2 or FL students mentally rehearse the language they are learning. Mental rehearsal has been recognized as a widespread language learning strategy, 19 Another distinction is between passive and active rehearsal (Bugelski, 1979). In passive rehearsal, the person simply mouths or reads the material over and over, without paying much attention to the contents of the material. In active rehearsal, the person tries to actively draw meaning from the material while repeating it.

24

CHAPTER 1

which has been broadly defined as "the covert practice of the L2" (Guerrero, 1994, p. 84). L2 internal rehearsal has been associated with a wide range of covert verbal behaviors, such as spontaneous recall, repetition, imitation, self-correction, making up sentences, practicing imaginary conversations, and planning language production. Two well-known L2 phenomena that have been related to mental rehearsal are the "din in the head" (Krashen, 1983) and "language play" (Lantolf, 1997). Because mental rehearsal is internal speech activity, it provides an important link to investigating the development of inner speech in the L2. But it should be clear that not all mental rehearsal of the L2 is tantamount to fully developed L2 inner speech. In the case of low proficiency learners, who are in the process of internalizing L2 social speech, mental rehearsal represents an incipient form of inner speech characterized by covert attempts aimed at the appropriation and eventual utilization of the L2 as a tool for thought and self-communication. Private Speech Private speech may be succinctly defined as "audible speech to oneself (Frawley, 1997, p. 95). In contrast to inner speech, which is covert speech for the self, private speech is overtly articulated and audible. Private speech shares features of form and function with inner speech: They are both abbreviated and condensed;20 they are also both self-regulatory in function. The main distinction between private and inner speech is that the former is spoken out loud and the latter is silent (although to the "inner speaker," inner speech may be sonorous "in the head."). Developmentally, private speech antecedes inner speech; it is external speech that has not yet lost its vocal character in its way to internalization. The term private speech is related to egocentric speech, the name Piaget and Vygotsky used to refer to the type of loud self-talk that children before school age engage in. Piaget's hypothesis was that egocentric speech was a transitory, inconsequential, phenomenon among children in their process of socialization. Egocentric speech, for Piaget, was speech about the child and for the child and therefore indicative of the autistic thought of the child. In Piaget's view, egocentric speech was likely to disappear as the child moved to a socialized frame of mind. In response to Piaget, Vygotsky postulated the reverse: Egocentric speech was an intermediate phase between social speech-its precursor-and inner speech. For Vygotsky, egocentric speech did not disappear but went underground as inner speech. In the 1960s, Flavell (cited in Girbau, 1996) adopted the terms private speech and social speech to refer respectively to speech addressed to the self and speech addressed to others. Subsequently, the name private speech has been preferred over egocentric speech in the literature, a move that suggests a desire to break away from Piaget's original conceptualization of the phenomenon (Girbau, 1996; Kohlberg, Yaeger, & Hjertholm, 1968; Zivin, 1976). 20 Empirical studies have found that private speech is not always as abbreviated and incomprehensible as Vygotsky claimed (Goudena, 1992; Wertsch, 1979a).

UNDERSTANDING INNER SPEECH

25

Most of the literature on private speech associates the term with audible self-talk. It should be pointed out, however, that in some instances the term private speech has been used to encompass both audible and silent (inner) speech for the self (Lantolf, 2000b), possibly to highlight the similar orientation to self of both types of speech. Private speech has also been held synonymous to intrapersonal communication (Lantolf, 2003; Lantolf & Yanez, 2003). In this book, the distinction between the audible (private speech) and silent (inner speech) forms of self-addressed speech will be kept. The term intrapersonal communication, in turn, will be used to include both private (audible) and inner (silent) speech. CONCLUSION This chapter started with a brief account of the historical trajectory of the concept of inner speech since ancient times. From its beginnings, it is clear the concept has consistently linked thinking and speech; the nature of the association, however, has shifted with the theoretical stance from which the concept has been viewed. Vygotsky, the most important theoretician to date on inner speech, proposed a dialectic relationship in which thinking is neither identical with speech nor independent from it but attains shape and expression through inner speech. Vygotsky's perspective on inner speech represented the sociocultural theory approach to inner speech, which would inspire research among numerous researchers in the Soviet Union. Philosophical debates around the role of inner speech in the relationship between thinking and speaking continue, however, among Western scholars. On the research front, inner speech has been approached primarily as a psychophysiological phenomenon, as a psycholinguistic event associated with memory, or as an instance of intrapersonal communication. The sociocultural theory approach to inner speech has been concerned mostly with the study of its developmental precursor, private speech, and has provided the theoretical basis for the neuropsychological research of Luria and others. Because of its crucial role in shaping cognition and in literacy development, inner speech has also been recognized as an important object of study in language education. In all fronts, inner speech has been approached from an LI perspective, but studies on inner speech in the L2 have begun to emerge. Sociocultural theory, a theoretical approach that posits the human mind as socioculturally constructed, provides conceptual support to the analysis of inner speech in this book. Particular emphasis has been given to the sociocultural theory concepts of internalization, mediation, and genetic development as well as to basic principles of the related theory of activity. In this discussion, inner speech has been found to originate in the communicative, interpersonal sphere of social discourse and to constitute an internalized version of public speech aiding in the execution of higher psychological processes. Inner speech, resting primarily on language but also on other semiotic tools such as symbols and gestures, is thus conceived as an effective mechanism of sign-mediated psychological activity. Sociocultural theory approaches the study of mind and consciousness from a genetic, or developmental, point of view

26

CHAPTER 1

that stresses the social origin and nature of mental phenomena, including inner speech, and their development over time. Activity theory complements sociocultural theory with a solid framework for the study of human behavior, both social and mental, placing it in specific cultural, institutional, and historical settings. In this chapter, a broad definition of inner speech as silent speech directed to the se/fhas been forwarded. However, the difficulty of pinning down inner speech to a few words compels one to characterize inner speech as a complex thought-language mechanism essential to verbal thinking and covert self-communication. This characterization implies a view of inner speech as on-going mental activity-as inner speaking-mther than as some type of internal language structure. To further delimit the scope of inner speech, some associated concepts have been discussed. Verbal thought, for example, has been defined as symbolic thinking mediated by inner speech. In verbal thought, speech and thought come together through the mechanism of inner speech. The expression thinking in language or thinking in a certain language has been explained as either referring to the popular perception that thoughts appear in the mind as words (or as words of a particular language) or to the more sophisticated notion that inner speech constitutes thought rather than facilitates it. The difference between the terms language of thought and languagefor thought has also been pointed out. The label language ofthought, usually associated with Fodor's (1975) conception of an innate mentalese in which mental computations are carried out, is not to be confused with the expression language for thought, used by Frawley (1997) to denote language as the mediational means of thinking, that is, as inner or private speech. Other terms that are sometimes linked to inner speech have been addressed. Intrapersonal communication, for example, is a broader term than inner speech, as it encompasses all types of human communication, silent or aloud, in which the speaker and the addressee are the same. Self talk cannot be taken as coterminous with inner speech insofar as talking to oneself can be done not only silently but also loudly. Selftalk is furthermore a dialogic mode of inner speech, possible only at those stages of inner speech in which syntax and semantics are sufficiently elaborated. Covert linguistic behavior is a term used to highlight the fact that inner speech involves concealed physiological activity of the brain, speech muscles, and other organs, which can be directly observed with appropriate technology. Mental rehearsal, in turn, is the name assigned to the silent, spontaneous or deliberate repetition of language material for purposes of remembering, planning, and learning. When mental rehearsal is applied for the specific purpose of learning the L2, it represents development towards condensation and internalization as L2 inner speech. Finally, the terms private speech and inner speech are differentiated not only because developmentally private speech antecedes inner speech but also because inner speech lacks overt vocalization. Both are self-directed forms of speech, but whereas private speech is audible, inner speech is inaudible to an outsider and only sonorous to the person who experiences it.

CHAPTER 2 THINKING WORDS IN ONE'S FIRST LANGUAGE Inner Speech: The LI Perspective

One of the most distinguishing features of human nature is the capacity to speak to oneself without actually pronouncing words. When people think words, these words and their relationships have the attributes of a particular linguistic system. Among monolinguals, this system is that of the LI. Chapter 2 focuses on such phenomenon: thinking words in the LI or, put differently, inner speech in the LI. Although the greatest bulk of research on inner speech has been conducted from an LI perspective, this point of view has usually been taken for granted, and thus the majority of the literature is simply on inner speech, without reference to the language of its realization. Various theoretical and methodological approaches to inner speech, as well as empirical research, are reviewed in this chapter. The purpose is to provide the necessary background for the analysis of L2 inner speech issues in subsequent chapters. The review of the literature has been organized into three main areas: the sociocultural perspective, cognitive approaches, and brain-imaging research. First, the core premises of the sociocultural theory perspective to inner speech are thoroughly examined. Within this framework, some of the most representative proposals are presented: Vygotsky's seminal ideas on inner speech and verbal thought, Luria's neuropsychological view, A. A. Leontiev's theory of inner programming, Sokolov's psychophysiological research, and Vocate' s intrapersonal communication perspective. Second, cognitive approaches are considered, including Frawley's sociocomputational interpretation of inner speech, Clark's supracommunicative view, Carruthers's modularist perspective, and research within the information-processing paradigm. Third, studies of inner speech employing some of the latest brain imaging technologies are discussed. The chapter ends with implications from the LI literature for the investigation of inner speech in the L2. THE SOCIOCULTURAL APPROACH TO INNER SPEECH Because of its critical role in shaping and organizing the adult mind, inner speech is at the core of sociocultural theory. The key principle of sociocultural theory that mind

28

CHAPTER 2

is a social construct could not be adequately explained without recourse to the notion of inner speech, the externally derived, culture-ridden mechanism that is the pillar of human thinking. The role ascribed to inner speech within sociocultural theory is bold and far-reaching: from shaper of thought, to mediator of higher psychological processes, to catalyst of self-consciousness, to regulator of external actions. In the early decades of the 20th century, Vygotsky paved the way for the investigation of inner speech. By the 21 st century, his work has multiplied in the efforts of a host of followers and adherents. The following section is an attempt to give a comprehensive view of the sociocultural theory approach to inner speech, as reflected in the work of L. Vygotsky, A. R. Luria, A. A. Leontiev, A. N. Sokolov, and D. Vocate. The Thought-Speech Connection: Vygotsky's View of Inner Speech Inner speech is without doubt a topic of enormous magnitude within the work of Lev Vygotsky. His most extensive treatment of inner speech appears in Thought and Language (1986), a book dedicated to explore the relationship between thought and word. For Vygotsky, a clear grasp of the concept of inner speech was necessary in order to fully comprehend how thought and language are related. In the process of clarifying inner speech, Vygotsky managed to explain two distinct but convergent processes: thought development and speech development. Inner Speech in Relation to Thought and Speech In Vygotsky's reasoning, to explain inner speech is to trace the development of both speech and thought. Partly in response to Piaget, who saw mental development as gradual progress from autism to socialization, and partly in accordance with the sociohistorical approach to psychology he wanted to advance, Vygotsky (1986) postulated that the general line of "development of thinking is not from the individual to the social, but from the social to the individual" (p. 36). This basic premise underlies Vygotsky's proposed trajectory of verbal thought and inner speech.21 Analysis of animal and child studies led Vygotsky to believe that human thought and speech have different genetic roots and take different developmental routes. At a certain point, however, Vygotsky maintained, thought and speech come together. Vygotsky identified four phases in this dual development. At first, the child's thought is nonverbal and the child's speech is nonintellectual. This is the prelinguistic and preintellectual phase of development. At about the age of two, thought and language begin to interconnect, but only at a superficial level. The child begins to make correct use of grammar without really understanding the logic behind it. As Vygotsky (1986) put it, the child "masters the syntax of speech before the syntax of thought" (p. 87). At this point, the speech of the child is essentially social. Gradually, the child proceeds

21 The fact that Vygotsky stressed the social origin of thought and speech does not mean he denied their biological underpinnings. Rather, he argued that the development of verbal thought was a social event, determined by the historical and cultural circumstances surrounding the individual (1986, pp. 94-95).

THINKING WORDS IN ONE'S FIRST LANGUAGE

29

to the third phase, the stage of external operations and external speech. Mental operations are conducted externally; for example, counting is done with the aid of fingers. Language is used externally not only as a means of communication but also as a tool for thinking. This is the stage of egocentric speech, of thinking aloud, when speech is still external but directed to the self. Thought and speech coincide here, but on an external plane. The fourth stage of development occurs when external operations become internalized and language turns inward. The child can count in her head. Thought becomes verbal thought, and speech becomes inner speech, a tool for thinking silently. Thus, Vygotsky explained the origin and development of verbal thought and inner speech. It is important to differentiate, however, between verbal thought and inner speech. According to Vygotsky, thought and speech are from the start independent phenomena, and they continue to be so despite their union as verbal thought. Vygotsky (1986) compared thought and speech to "two intersecting circles. In their overlapping parts, thought and speech coincide to produce what is called verbal thought. Verbal thought, however, does not by any means include all forms of thought or all forms of speech" (p. 88). Thought may be nonverbal, as in the practical thinking of pre-linguistic children, and speech may not involve thought, as in mindlessly reciting a poem. "Thought has its own structure" (p. 250). It only becomes verbal when mediated by inner speech. Perhaps it might be helpful to think of inner speech and verbal thought as two sides of the same coin, the verbal and the thinking aspects, respectively, of the fusion of speech and thought that occurs at an intrapersonal level. But what Vygotsky called verbal thought is still thought and what he called inner speech is still speech (p. 249). These distinctions are important to keep in mind when trying to understand the nature of inner speech, that is, its form and function(s). The Nature of Inner Speech Vygotsky saw inner speech as a complex phenomenon, and he abstained from pinning it down in very precise terms. For example, he did not think appropriate to equate inner speech with verbal memory, although he did recognize that word memory was "one of the constituents of inner speech" (1986, p. 234). Likewise, he rejected the idea of inner speech as truncated speech, subvocal speech, or speech without sound. To him, all these terms, which denote inhibition of motor activity during silent speech, are "subordinate" aspects of inner speech, but they cannot describe the phenomenon as a whole (p. 234).22 When he tried to define inner speech, the result was an impressionistic account as elusive, evocative, and multilayered as the phenomenon itself: 22 Vygotsky did not deny the fact that inner speech involves suppression of motor speech reactions. In his book Educational Psychology (1997), he referred to those "suppressed, undisclosed movements of the motor speech mechanism that consist in complex elements of respiratory, muscular, and vocal reactions" that are at the basis of every thought as the "system of internal speech or silent speech" (p. 162). In fact, he considered this restraint speech mechanism as an essential sign of cultural evolution in humankind (p. 163).

30

CHAPTER 2

We can now return to the definition of inner speech . . . . Inner speech is not the interior aspect of external speech-it is a function in itself. It still remains speech, i.e., thought connected with words. But while in external speech thought is embodied in words, in inner speech words die as they bring forth thought. Inner speech is to a large extent thinking in pure meanings. It is a dynamic, shifting, unstable thing, fluttering between word and thought, the two more or less stable, more or less firmly delineated components of verbal thought, (p. 249)

The Functions of Inner Speech According to Vygotsky, speech plays two major functions: as a means for social communication and activity and as a tool for thought (see 1997, p. 171; 1986, pp. 6-7). Vygotsky explained the use of speech as a tool of thought: "We always think in some language, that is, we speak to ourselves and organize our behavior within ourselves just as we organize our behavior as a function of the behavior of other people" (1997, p. 171). This use of speech as a tool for thinking is what constitutes inner speech. But to understand this claim it is necessary to review Vygotsky's position on the role of egocentric speech, the developmental antecedent of inner speech. Unlike Piaget, who thought that egocentric speech was a passing phase that had no useful function and was a mere reflection of the child's egocentrism, Vygotsky believed that egocentric speech was instrumental in helping the child plan his actions, guide activity, and solve problems. A series of experiments conducted by Vygotsky and associates convinced him that egocentric speech was no trivial event. Through these experiments, it was observed that pre-school children resorted to egocentric speech when confronted with obstacles or difficulties in their problem solving. The children's loud vocalization of their mental operations as they were solving a problem helped them grasp the situation and find a solution. The experiments also showed that when older children encountered problems, rather than using egocentric speech, they thought silently for a while before coming up with a solution. When asked to verbalize their thoughts, the older children produced accounts of their mental operations very similar to those produced by the younger children in their egocentric speech. This led Vygotsky (1986) to postulate that "the same mental operations that the preschooler carries out through voiced egocentric speech are already relegated to soundless inner speech in schoolchildren" (p. 30). For Vygotsky, egocentric speech was an "instrument of thought in the proper sense" (p. 31), which does not atrophy once it stops being vocalized, as Piaget believed, but goes underground as inner speech. The following quote neatly summarizes Vygotsky's view: Our experimental studies indicate that the function of egocentric speech is similar to that of inner speech: It does not merely accompany the child's activity; it serves mental orientation, conscious understanding; it helps in overcoming difficulties, it is speech for oneself, intimately and usefully connected with the child's thinking. Egocentric speech develops along a rising, not a declining curve; it goes through an evolution, not an involution. In the end, it becomes inner speech, (p. 228)

This view implies that, once the child is able to "think words" (1986, p. 230) without pronouncing them, he has transferred to the intrapsychological sphere the capacity to carry out all those mental operations that he could earlier perform

THINKING WORDS IN ONE'S FIRST LANGUAGE

31

interpsychologically in communicative discourse with others and vocally with himself during egocentric speech. With the internalization of social speech, the child is capable of applying social forms of mediation to self-mediation. This is the case, for example, of argumentation, a typical social behavior among children, whose structure is internalized to give rise to inner, logical reflection (p. 35). In fact, the internalization of social speech has a crucial role in the development of higher psychological functions-such as voluntary attention, logical memory, and concept formation. As Vygotsky explained in his book Mind in Society (1978), with the internalization of speech there is a transformation in the nature of sign operated mental functions. In this process, such basic mental operations as memory and attention, first carried out interpersonally with the aid of external signs or tools, are transformed into higher mental functions through mediation of an internal system of psychological signs (p. 57). Thus emerges the pivotal role of inner speech as mediator of higher psychological processes. In addition to the thinking function of inner speech, Vygotsky acknowledged, but did not fully develop, its intra-communicative function. Although he recognized that language can be a tool of "innercommunication between man and himself (1997, p. 172), the notion that inner speech could also serve for purposes of self-communication was left relatively unexplored. Vygotsky regarded inner speech as wordless "communication" (1986, p. 243) but was somewhat ambivalent in his representation of inner speech as either monologue or dialogue (see pp. 240-244). At one point, he stated that inner speech represents the monologue rather than the dialogue (p. 240) because inner speech lacks the intonational and gestural information that can be conveyed in conversation. Dialogue is, furthermore, spontaneous and unpremeditated, he argued, whereas monologue is leisurely and elaborated (p. 242). Yet, at another point, Vygotsky compared inner speech to the condensed-yet rich in hints and allusions-conversation between two lovers (pp. 238, 244). Dialogue of this kind is thus very much like inner speech because both tend to be abbreviated and predicative (preserving the psychological predicate). Vygotsky, however, did not probe into the dialogic function of inner speech. It would take other scholars to emphasize this aspect of inner speech. (See further down for Bakhtin's and others' contribution to the dialogic aspect of inner speech.) A related communicative function that was hinted by Vygotsky and would be later much more developed by followers such as Luria and Sokolov was the rehearsal or speech planning function of inner speech. Vygotsky (1986) believed that inner speech "serves as preparation for external speech-for instance, in thinking over a lecture to be given" (p. 88). He considered that planning through inner speech has a particularly important function when writing. Inner speech works as a "mental draft" when "we say to ourselves what we are going to write" (243). The Formal Features of Inner Speech Considerations about the functions of inner speech in Vygotsky's work naturally lead to examination of its form. In a way, form and function are indissoluble in inner

32

CHAPTER 2

speech. Inner speech is what it is because of what it does. As Vygotsky (1986) put it, "inner speech is speech for oneself; external speech is for others. It would be surprising indeed if such a basic difference in function did not affect the structure of the two kinds of speech" (pp. 225-226). Three main formal features of inner speech were discussed by Vygotsky: soundlessness, abbreviated syntax, and condensed semantics. The first characteristic Vygotsky noticed about inner speech was its lack of sound. Because speech is for the self, it does not have to audible to others. It would be wrong, however, according to him, to consider inner speech as simply "speech minus sound" (p. 235). It is not its silent character that gives inner speech its peculiar form; it is its abbreviated syntax. In inner speech, words can be omitted because "we know what we are thinking about; i.e., we always know the subject and the situation" (p. 243). Because the sentence subject is known, it can be dropped in inner speech. Thus, Vygotsky hypothesized that the typical syntactic structure of inner speech is predicative.23 The predicativeness of inner speech is sometimes so extreme that "inner speech is speech almost without words" (p. 244). Inner speech owes the abbreviatedness of its syntax to its semantics, the third formal trait discussed by Vygotsky. The grammar of sentences is reduced because full expanded meaning is not necessary in inner speech. There are three main semantic features of inner speech. One is reliance on sense, rather than meaning. Here, Vygotsky (1986) adopted F. Paulhan's distinction between the sense of a word, that is, "the sum of all the psychological events aroused in our consciousness by the word" (p. 244) and the meaning of a word, that is, its most stable and precise sense. The sense of a word is born through personal experience in context; the meaning of a word is an agreed-upon public sense, the one usually fixed in dictionaries. Inner speech, thus, evokes senses rather than meanings. Another semantic feature of inner speech is agglutination, the tendency for words to merge with one another. This phenomenon is similar to that which occurs in egocentric speech or in certain languages, such as German, when word compounding is used to designate complex ideas and grammatical relations. Finally, the semantics of inner speech is also characterized by the influx of sense, or the way in which the senses of words influence one another. Vygotsky compared this semantic trait to the relationship the title of a book has to its contents. The title condenses the contents and at the same time influences our understanding of them. The combined effect of all these semantic traits is that words in inner speech acquire a very special meaning. In this, they are like "idioms" that are difficult to explain (p. 248). "In inner speech, one word stands for a number of thoughts and 23 Actually, Vygotsky (1986) recognized that it is not always the predicate that is omitted in inner speech and that "any part of the sentence may become a psychological predicate" (pp. 220-221). As Wertsch (1979b, 1985b) pointed out, in his use of the term psychological, Vygotsky was reflecting a functional, rather than a strictly grammatical, conception of predicativity. This view would allow for either the sentence subject or predicate to be omitted in inner speech. A better way to describe the form-function relationship of the structure of inner speech that avoids the use of traditional grammatical terms is, according to Wertsch, on the basis of the given-new distinction, whereby what is omitted in inner speech is what is already given in the speaker's consciousness and what is preserved is what is introduced as new information.

THINKING WORDS IN ONE ' s FIRST LANGUAGE

33

feelings, and sometimes substitutes for a long and profound discourse" (p. 248). The peculiar nature of inner speech is determined then by its sui generis syntax and semantics. Vygotsky (1986) pointed out that, if inner speech could be recorded, it would appear very different from external speech. It would be disconnected, incomplete, even incoherent (pp. 235-236). In inner speech, "a word is so saturated with sense that . . . to unfold it into overt speech, one would need a multitude of words" (p. 247). But inner speech is not always so reduced in form. Vygotsky admitted that "inner speech may come very close in form to external speech... or even become exactly like it when it serves as preparation for external speech" (p. 88). The same process of expansion in form and meaning occurs when inner speech functions as a "draft" in writing (p. 243). The Processes of Inner Speech The instability of form described above may be a possible source of confusion when trying to grasp Vygotsky's conception of inner speech. Certainly, it is quite perplexing to find Vygotsky (1986) claiming, on the one hand, that inner speech is "to a large extent thinking in pure meanings" (p. 249) and, on the other, that inner speech may be almost like external speech (p. 88). Yet, the key to understanding what Vygotsky may have had in mind lies in the acknowledgment that, throughout his description of inner speech, he was referring to at least three distinct processes: (1) inner speech as the internalization of social speech, (2) inner speech as the formulation of thought, (3) and inner speech in the externalization of thought.24 All these processes involve expansion or reduction of speech and movement through phases. The first, internalization, occurs over a relatively long period of time. It is an ontogenetic process involving years in a child's life. It entails the well-known transition from social, to egocentric, to inner speech.25 The formulation of thought and its externalization, on the other hand, are an on-line events, usually occurring in seconds or fractions of seconds. They are microgenetic26 processes involving the fusion of thought and word into verbal thought, and the transformation of thought, through inner speech, into external speech. The first two, internalization and

24 A fourth process involving inner speech and changes in structure is the interiorization of speech heard. Vygotsky (1986) might have been referring to it when describing inner speech as "going from outside to inside. Overt speech sublimates into thoughts" (p. 226). However, little if no reference is found in Thought and Language to the role of inner speech in listening (or reading), a topic that would be thoroughly investigated by Sokolov (1972) later. 25 According to Vygotsky (1986), the process of speech and thought individuation in a child's life occurs roughly over a span of four years ("between three and seven years" of age, p. 230.) It must be presumed then that, although Vygotsky usually portrayed inner speech as a phase following egocentric speech, he conceived of its development as a slow process that takes place alongside its vocal manifestation as egocentric speech. See Frawley's (1997, p. 184) argument for pushing the onset of private speech to an earlier date than that proposed by Vygotsky. 26 As opposed to ontogenesis, microgenesis is a term used by Wertsch (1985b) to refer to the type of short-term psychological or developmental processes that Vygotsky described. According to Wertsch (2000), the process of speech production, i.e. externalization, described by Vygotsky is microgenetic (p. 24).

34

CHAPTER 2

formulation, are inherent in the essence of inner speech, that is, inner speech is always the result of intemalization and inner speech always involves the fusion of thought and speech. Conversely, the externalization of inner speech may or may not occur: Inner speech not always becomes external-though it frequently does.27 Inner Speech as the Intemalization of Social Speech. As Vygotsky repeatedly explained, the process of turning social speech inwards goes through the phase known as egocentric speech. When compared to social speech, egocentric speech appears fragmentary and incomprehensible, yet it is still vocal. Eventually, egocentric speech becomes even more reduced and unintelligible and sheds its last social feature: vocalization. Egocentric speech turns into inner speech. With "syntax and sound reduced to a minimum, inner speech is speech almost without words" (1986, p. 244) in its most concentrated form. Thus, the process of intemalization is in essence a process of condensation, from grammatically elaborated and vocally articulated external speech to abbreviated, agglutinated, idiomatic, silent speech. When Vygotsky described this process, he was delineating the development of the child's faculty to think words. Inner Speech as the Formulation of Thought. Vygotsky (1986) explicitly addressed "the process of verbal thinking from the first dim stirring of a thought to its formulation" (p. 217) and the role played by inner speech in shaping thought. Because thought and speech have different structures, the process of verbally formulating thoughts is a complex one, involving the transformation or reshaping of thought through inner speech. To understand how thinking unfolds into words, it is important to take into account how thought originates: "Thought is not begotten by thought; it is engendered by motivation, i.e., by our desires and needs, our interests and emotions" (p. 252). It is this affective-volitional basis that determines the unique structure of thought and that renders its formulation in words a highly dynamic process. Unlike speech, thought is non-linear, does not consist of separate items, and takes place all at once (pp. 251-252). Thus, Vygotsky contended, thought is not "expressed" in words; rather, it "comes into existence" (p. 218) through them. Like intemalization, thought formulation also entails movement through phases. Thought flows "as an inner movement through a series of planes . . . before it is embodied in words" (p. 218). According to Vygotsky, the transition from thought to inner word occurs through meaning. There are no ready-made words for thoughts; thoughts must first find their meaning. Vygotsky (1986) offered a summary of the phases involved in verbal thinking: "from the motive that engenders a thought through 27 Vygotsky's inconsistency in referring to the different phases of inner speech has also been noticed by Wertsch (2000), who claims that Vygotsky had two different phenomena in mind when discussing inner speech: "inner speech as one of the phases of the microgenetic process of producing speech utterances" and inner speech "as an instrument in problem solving and other forms of rational thinking" (p. 25). According to Wertsch, these are the same phenomena that A. A. Leontiev characterized as "inner programming of an utterance" and "inner speech in the strict sense," respectively (as cited by Wertsch, 2000, p. 25).

35

THINKING WORDS IN ONE'S FIRST LANGUAGE

the shaping of the thought, first in inner speech, then in meanings of words, and finally in words" (p. 253). But the movement from thought to word is not a smooth one: "The relation of thought to word is not a thing but a process, a continual movement back and forth from thought to word and from word to thought" (p. 218). Vygotsky recognized both the intricacy and fluidness of the process, as well as his own limitations in explaining the problem, when he stated: "The development [from thought to word] may stop at any point in its complicated course: an infinite variety of movements to and fro, of ways still unknown to us, is possible" (italics added, p. 254). Inner Speech in the Externalization of Thought. It is clear from the above that inner speech may not become externalized. When it does, however, inner speech requires further elaboration and transformation (see Figure 2-1). Vygotsky (1986) explained this process: "The transition from inner speech to external speech is . . . a complex, dynamic process involving the transformation of the predicative, idiomatic structure of inner speech into syntactically articulated speech intelligible to others (p. 249). Inner speech may be externalized as oral or written speech. In the process of externalization, inner speech may become very much like external speech in form. Vygotsky gave two examples of unfolded inner speech (already noted before): when inner speech is used in preparation before delivering a lecture (p. 88) and when inner speech functions as a draft of what we are going to say or write (p. 243). Vygotsky was quite explicit when describing the process of transformation that occurs as inner speech becomes written speech. According to Vygotsky, while inner speech is reduced and condensed because it is for the self, written speech is fully expanded because it must be understood by others. The shift from "maximally compact inner speech to maximally detailed written speech" (p. 182) requires not only full expansion of the internal syntax of words but also a great degree of semantic elaboration, processes that demand from the thinker a high level of conscious awareness and deliberate control.

motive

thought

inner speech condensed, subjective semantics; predicative syntax

transition to external speech (may take place as expanded inner speech) external meanings of words, articulated syntax

external speech

Figure 2-1. Vygotksy 's conceptualization of the transition from thought (preceded by motive), through inner speech, to external speech.

Because of its complexity and depth, Vygotsky's treatment of inner speech remains, in summation, one of the most powerful and thought-provoking in the psychological literature. Vygotsky's position on the origins, role, and nature of inner speech is pivotal, so much so that, after Vygotsky, one might be able to classify views of inner

36

CHAPTER 2

speech as Vygotskyan or non-Vygotskyan. What, then, does the Vygotskyan view of inner speech so singularly entail? In a few words, a social-psychological phenomenon of outstanding proportions. Going beyond its superficial appearance as mere speech without sound, Vygotsky saw in inner speech the convergence of thinking and speaking, the fusion that makes possible that uniquely human attribute: verbal thought. Vygotsky firmly demonstrated the social basis and character of inner speech, tracing its evolution in children from external, communicative speech, to egocentric speech, to silent, self-directed speech. Vygotsky, however, never depicted inner speech as just a form of speech. For him, inner speech was a unique formation, a blend of thought and speech with its own characteristics and functions. Structurally, Vygotsky observed the soundlessness, syntactic abbreviatedness, and condensed semantic nature of inner speech. Functionally, Vygotsky regarded inner speech primarily as a tool for thought, instrumental in all forms of higher mental activity, from planning and guiding action, to organizing conscious thought, to problem solving and self-reflection. Fundamentally, Vygotsky approached inner speech as a process, not a fixed inner language code or a stable state of mind, but a fluid, dynamic, ever-changing movement between thought and word. The process-like nature of inner speech was plainly evident in Vygotsky's microgenetic account of the transition from thought to word in the formulation and externalization of thought as well as in his ontogenetic depiction of the internalization of social speech. Developing Vygotsky's Conception of Inner Speech Vygotsky laid an exceedingly sound theoretical and empirical foundation for the study of inner speech. Unfortunately, because of his interest in multiple other topics and his untimely death, some of his ideas regarding inner speech were left unexplored or underdeveloped. Numerous explications, extensions, and interpretations of Vygotsky's theories have been proposed providing much needed clarification in some obscure areas or fresh new perspectives as further research has been conducted. Akhutina (1978), for example, has done a fine job of explicating Vygotsky's (and his followers') account of the multi-level nature of inner speech and the role of inner speech in producing utterances. John-Steiner has extended the discussion of the functional aspects of inner speech (John-Steiner et al, 1994) as well as explored novel ways of studying inner speech (e.g. through private jottings in notebooks and diaries), both from an LI (John-Steiner, 1985a) and L2 perspective (John-Steiner, 1985b). Wertsch (1980, 1985b, 2000), one of Vygotsky's most important interpreters, has dealt with some of the most complicated issues in Vygotsky's conception of inner speech, such as its dialogic dimension, its role in consciousness, and the notion of internal word meanings. Because it would be impossible to do justice to all the available literature extending Vygotsky's work on inner speech, two major issues, highly relevant to the main focus of this book (inner speech and L2 learning), will be subsequently granted special attention: the dialogic nature of inner speech and the role of inner speech in consciousness, introspection, and self-awareness.

THINKING W O R D S IN O N E ' s FIRST LANGUAGE

37

The Dialogic Nature of Inner Speech. As mentioned, there are a few references to inner speech as a form of dialogue in Thought and Language (e.g. "inner dialogue," Vygotsky, 1986, p. 243). The idea of inner speech as dialogic is also implicit in Vygotsky's insistence on the social origin of inner speech. If inner speech results from the application of social forms of intercourse to the interpsychological sphere, then it follows that dialogic forms of discourse will be internalized, too. Vygotsky, however, did not pursue to the full the implications of a dialogic view of inner speech. As Wertsch (1980, 1985b) notes, this view can be found with greater specificity in the work of Bakhtin. According to Bakhtin (Volosinov, 1973),28 "the units of which inner speech is constituted . . . resemble the alternating lines of a dialogue" (p. 38). As Wertsch (1985b) points out, Bakhtin's notion of dialogue is related to his conception of voice and utterance. Voice in Bakhtin's work does not simply mean the speech of an individual but an amalgam of ideological, historically situated, social voices in a person's speech. And as Holquist (1981) explains, utterance for Bakhtin is "social, historical, concrete, and dialogized" (p. 433). According to Wertsch (1985b), Bakhtin's notion of inner dialogue as the point of contact for sociohistorical and ideological forms of discourse is consistent with Vygotsky's notion of sense, yet it is more concrete in specifying the sociohistorical underpinnings of inner speech. Thus, incorporating Bakhtin to Vygotsky's theories would suggest to Wertsch (1985b) that "the speech which is internalized carries with itself a great deal of sociohistorically specific ideological baggage. The voices to which a speaker is exposed in social life determine certain fundamental aspects of how reality can be represented in inner speech" (pp. 229-230). InnerSpeech: Its Role in Consciousness, Introspection, and Self-Awareness. Although the problem of consciousness was perhaps the overarching and most important concern for Vygotsky (2000/1925; see also Wertsch, 1985a), there are only a few obvious references to consciousness in relation to inner speech in his work. Specifically, at the beginning and end of Chapter 7 in Thought and Language, Vygotsky (1986) put forth his "historical theory of inner speech" (p. 255), most tersely contained in the closing statement of the book: "A word is a microcosm of human consciousness" (p. 256). According to this theory, the historical development of consciousness is reflected in the inner word. For Vygotsky, the inner fusion of thought and speech is a product, not a prerequisite, of the development of human consciousness (p. 210). Another allusion to the role of inner speech in the experience of being conscious made by Vygotsky was his reference to the child's attainment of "verbalized introspection" or "verbalized selfobservation," when thanks to internalized meanings he/she can perceive his/her own intellectual processes as meaningful (p. 170). No doubt, Vygotsky saw inner speech as indispensable for self-awareness, or what he called "the consciousness of being

28 Although published under the pen name of Bakhtin's colleague, V. N. VoloSinov, the book Marxism and the Philosophy of Language is attributed in its majority to the authorship of Bakhtin (Holquist, 1981, p. xxvi).

38

CHAPTER 2

conscious" (p. 170). Important extensions of the role of inner speech in self-awareness have been presented in Morin (1993) and McCrone (1994, 1995). An explicit treatment of the sociohistorical dimension of consciousness-as shaped through inner speech-was emphasized by Bakhtin (see Emerson, 1996). For Bakhtin, inner speech (the inner sign) is "the semiotic material of inner life-of consciousness" (Volosinov, 1973, p. 14), therein the equation of inner life with inner speech and consciousness (Emerson, p. 127). But, for Bakhtin, as for Vygotsky, consciousness is socially and historically constructed. The psyche of an individual is a social, ideological formation: The content of the "individual" psyche is by its very nature just as social as is ideology, and the very degree of consciousness of one's individuality . . . is ideological, historical, and wholly conditioned by sociological factors. Every sign as sign is social, and this is no less true for the inner sign than for the outer sign. (VoloSinov, 1973, p. 34)

It was in the context of his discussion on the inner sign and consciousness that Bakhtin raised the issue of introspection. According to Bakhtin, the inner sign is not only "accessible to" (p. 33) but also "the object of introspection" (p. 36). To introspect, in Bakhtin's view, is to hold the inner sign for inspection and to understand it: "In the process of introspection we engage our experience into a context made up of other signs we understand. A sign can be illuminated only with the help of another sign" (p. 36). For Bakhtin, understanding of one's inner sign always implies grasping its ideological baggage and the empirical context to which it is tied: "The sign and its social situation are inextricably fused together" (p. 37). Inner Speech as Activity: Luria's Neuropsychological View One of the major treatments of the problem of inner speech is to be found in the work of A. R. Luria, a contemporary and follower of Vygotsky who devoted his life to the application of a cultural-historical perspective to psycholinguistics and, in particular, to the study of neuropsychology and aphasiology. In general, Luria followed closely Vygotsky's postulates about the nature of inner speech, its genetic development, and its role in the process of verbal thinking, but Luria made two very important novel contributions. The first was his attempt to reformulate Vygotskyan inner speech theory within the framework of activity theory, an endeavor that reflected the major influence that A. N. Leontiev's theory of activity would have on Soviet psychology in postVygotskyan times (see Wertsch, 1981). The other important contribution was derived from Luria's interest in specifying the cerebral correlates of speech functioning in normal and abnormal cases. His novelty in this respect consisted in identifying the neurophysiological basis of inner speech and the consequences, in terms of speech disorders, that result from damage to the brain centers implicated in inner speech activity. Luria's (1973, 1981) conception of inner speech is congruent with his view of speech as activity: "As is true with all other forms of mental activity, we must identify in speech activity the motive that gives rise to it, the goal that it serves, and the task that arises as a result of setting this goal in a certain context" (1981, p. 155). Speech in

THINKING WORDS IN ONE' s FIRST LANGUAGE

39

particular is an "organized form of conscious activity" (1973, p. 306) that uses language as its tool. Luria investigated three main functions of speech activity: as a means of communication, as a tool for intellectual functioning, and as a method of regulating behavior (p. 307). In all of these, he found inner speech to play a fundamental role. In its social communicative function, according to Luria (1973), speech activity has two mechanisms: expressive speech (speech production) and impressive speech (speech reception). Both mechanisms are mediated by internal speech. As shown in Figure 2-2, Luria (1981) distinguished four stages in the process of expressive speech: (1) the motivation or need to say something, (2) an initial thought in the form of a nonsequential utterance plan or semantic graph, (3) inner speech, which captures the sense of the utterance and reduces it to a predicative structure, and (4) an expanded speech utterance, phonologically, lexically, syntactically, and semantically organized on the basis of a specific language code (see pp. 152-153, 244-245). The sequence is reversed in impressive speech, a process that basically consists of decoding perceived speech, identifying its inner speech scheme, and understanding the hidden meaning and motive of the perceived expression (Luria, 1973). In reading, in particular, comprehending the message is a question of going from the text, or "outer meaning," to the subtext, or "inner sense" (1981, p. 194). motive

initial thought (semantic graph)

inner speech (sense-based, predicative schema)

external speech (expanded, meaningbased utterance)

Figure 2-2. Luria's view of the speech production process from motive to external speech.

Luria was also interested in the intellectual or cognitive function of speech activity. His conception of thinking as "concrete mental activity" (1973, p. 327)-and not as a general faculty localized in some substratum of the brain-allowed him to concentrate on the mechanisms and stages of verbal thinking. Luria's account of the mechanisms involved in problem solving is paradigmatic of his approach to thinking and speech as activity. For Luria, "the origin of thought is always the presence of a task" (p. 327) that creates a motive for action. This initial motive to solve the task triggers a serial process of related activities: analysis of task conditions, consideration of alternate solutions, creation of a general performance plan, choosing the appropriate operations for carrying out the plan, and conducting these operations in the actual solving of the problem. As Luria explains, the operative phase is extremely complex. It involves not only external trial and error but also an internal process of searches through inner speech. At this stage, the individual usually obtains assistance from "well assimilated internal codes" or "ready-made algorithms (linguistic, logical or numerical) which have evolved in the course of social history" (p. 328). Another important aspect of speech activity researched by Luria (1973,1981) was

40

CHAPTER 2

its self-regulative function, that is, the role speech plays in regulating voluntary action. Luria's conception of the voluntary act rests on two main principles: (1) it is largely through speech that humans regulate their behavior and (2) the regulative function of speech has its origin in social interaction. Luria's (1981) ontogenetic analysis of the regulative function of speech reveals that it is the mother (or caretaker) who initially, with her speech, organizes the child's actions. The child's behavior is at first subordinate to the adult's verbal commands. Voluntary activity takes place as distributed, interpsychological activity, initiated by the adult's speech and completed by the child. At some point, the child internalizes the adult's speech and starts giving orders to himself/herself. The child develops the capacity to regulate his/her own actions intrapsychologically, first through external (egocentric) speech and then through inner speech: Inner speech retains all of the analytic, planning, and regulative functions found in external speech; it continues to fulfill the same intellectual role originally performed by the adult's speech addressed to the child and later carried out by the expanded speech of the child himself/herself. (1981, p. 106)

Luria's (1981) account of the development of inner speech took Vygotsky's work as point of departure but added extensively to it with research conducted by himself and colleagues. Luria's findings basically confirmed Vygotsky's observations on the social-to-the-individual progression of regulative speech and the changes in structure it undergoes: At first, this speech appears in very expanded form. During the following stage of ontogenesis, it gradually becomes more and more abbreviated, finally turning into whispered speech, in which children describe haltingly the difficult situation they find themselves in. In the next stage (after another year or two), external speech disappears altogether. Only by watching the child's lip movements can one surmise that this speech had turned "inward," had become abbreviated and "internalized" or transformed into what might be termed "inner speech, (pp. 104-105)

It is interesting to note Luria's inclusion of whispered speech as a stage in the evolution ofinner speech (see also 1981, pp. 107 and 153). Vygotsky( 1986) had been emphatic in his rejection of whispered speech as a transitional phase between external and inner speech. Reacting to Watson, who had posed a loud speech-whisper-inner speech sequence, Vygotsky asserted: "There are no valid reasons to assume that inner speech develops in some mechanical way through a gradual decrease in the audibility of speech (whispering)" (p. 84). Further, he stated: "Neither in function nor in structure can [whispered speech] be considered a transitional stage between external and inner speech. It stands between the two only phenotypically, not genotypically" (p. 85). Clearly, for Vygotsky, the presence of whispering did not constitute a sign of a genetically distinct phase in the evolution of inner speech; whispering is simply a manifestation (in a low voice) of external, self-directed speech. For Vygotsky, the crucial intermediate link between overt and inner speech was egocentric speech. Luria, unfortunately, did not explicitly address this discrepancy with Vygotsky. Given the overall widespread agreement between the two scholars, however, it is possible to speculate that Luria may have used the term whispered speech as synonymous to

THINKING WORDS IN ONE'S FIRST LANGUAGE

41

egocentric speech.29 Luria's description of whispered speech as fragmentary and abbreviated and not simply as speech-in-a-low-voice (p. 107) provides a clue that his conception of whispering was not too far apart from Vygotsky's notion of egocentric speech. Luria coincided with Vygotsky in pointing out the tendency for speech to become predicative as it is internalized (Luria, 1981, p. 107). Like Vygotsky, Luria recognized the predominance of the psychological predicate in inner speech structure. Actually, Luria used the terms "theme" and "rheme" to refer respectively to the nominative and predicative functions of speech, but in essence he was in complete agreement with Vygotsky on the fact that inner speech is organized around what is new and not around what is already known (the psychological subject). In line with Vygotsky, too, Luria (1981) believed that it is the predicative nature of inner speech-the preservation of the plan of action in regulative speech-that makes it possible to transform inner speech into expanded speech. For example, if I am going to a lecture to speak about the mechanism of inner speech, I carry only an abbreviated outline of the lecture in my mind. Items such as 'inner speech," "egocentrism," and "predicativity"-items that do not name the theme but indicate what I want to say about it (in other words, predicative items)-are precisely what enable me to go on to produce the external structure. Accordingly, on the basis of inner speech, the lecturer can expand a scant outline into an entire lecture, (pp. 107-108)

As mentioned, another novel contribution to the study of inner speech made by Luria was his identification of the brain areas and mechanisms involved in inner speech activity. Again, his view of thinking and speech as activity, and not as general faculties of the mind, allowed him to focus on the cerebral components and the various phases involved in the processes of speech production and intellectual functioning. To understand the cerebral organization of speech and cognitive activity, Luria turned to the study of speech and brain disorders. One of his most significant findings was that lesions in different parts of the brain affect different stages in the transition from thought to speech, resulting in a variety of speech disorders (1973, p. 309). For example, he found that the inability to produce spontaneous narrative speech results from damage to the brain's frontal lobes (p. 318). Damage to this area impairs the ability to formulate an intention or plan, which is the initial point of the spontaneous verbal utterance. Non-spontaneous speech, such as echolalic responses, however, remains intact. Lesions in the "inferior postfrontal zones of the left hemisphere" (p. 319), on the other hand, cause disturbances in the transition from the general plan or intention to a linear verbal scheme. Inner speech, with its predicative structure, plays an important role in this transition because it already provides the syntax for the general scheme of the sentence. Damage to regions supporting this transition, Luria observed, deprives subjects of the ability "to express a thought, or even to produce an

29 Vygotsky and Luria were coauthors of the paper "The Function and Fate of Egocentric Speech," which was presented at the Ninth International Congress of Psychology, Yale University, in 1929. In this paper, at least, Luria was espousing the traditionally Vygotskyan sequence of external speech-egocentric speech-inner speech. (See Vocate, 1987, pp. 78, 132.)

42

CHAPTER 2

elementary verbal expression" (p. 319). Luria's experimental studies showed that what is impaired in these subjects is not the formulation of a motive or plan but the conversion of thought into sequential speech that is made possible by inner speech. Luria (1981) also found that the cerebral mechanisms governing the regulative function of speech are not the same as those involved in speech perception and production. For example, damage to Wernicke's section (left temporal lobe), which is associated with speech perception, causes inability to distinguish phonemes. Additionally, lesions in the post-central zones of the speech cortex, which are responsible for articulatory processes, lead to motor aphasia. Luria's studies, however, showed that inner speech in these cases "remains relatively intact and that a patient whose perception of speech and phonemes is badly affected (in other words, a patient suffering from sensory or motor aphasia) may continue to regulate actions that are either self-initiated or are assigned by others" (pp. 108-109). On the basis of his research, Luria concluded that the neural basis for the regulative function of inner speech is localized in the frontal lobes, especially in the left hemisphere (p. 109). Lesions in this area may not affect simple, habitual verbal activity, such as greeting or answering routinely, but they would severely impair complex, voluntary speech production requiring inner speech programming. Luria's legacy to the construct of inner speech cannot be overemphasized. Following general Vygotskyan principles about the social origins of intellectual functioning, Luria was able, through neuropsychological investigation and the application of activity theory, to discover the brain materiality of inner speech. Through his research, he not only solidified notions about the social basis of inner speech but also advanced important hypotheses about the role of inner speech in speech production and regulative behavior. In particular, his approach to inner speech as mental activity with real correlates in brain mechanisms allowed him to show the functionality of inner speech and how utterly disruptive to normal mental life any tampering with inner speech activity could be. Inner Speech in the Planning of Speech Production: A. A. Leontiev's View An important contribution to an understanding of the role of inner speech in the production of speech acts was offered by psycholinguist A. A. Leontiev (1981, see also, Akhutina, 1978). Particularly interesting is Leontiev's specification of inner speech mechanisms in the planning of FL utterances. A brief discussion of his general view of stages of speech production is presented here to set the basis for a more elaborate treatment of his ideas regarding the L1-L2 connection in Chapter 3. A. A. Leontiev believed that Chomskyan, transformational approaches to psycholinguistic data on speech production would profit from reinterpretation based on the inclusion of a Vygotskyan inner speech stage (Akhutina, 1978, p. 22). Leontiev (1981) proposed dividing the verbal act, that is, "the act of generating a verbal utterance in the process of setting and solving a particular problem," (p. 105), in two main parts: a planning (or programming) stage and a post-programming stage (see Figure 2-3). In this view, the planning stage consists of three moments: the first is a

43

THINKING WORDS IN ONE'S FIRST LANGUAGE

need or problematic situation and the motivation to solve it, the second is the generation of a speech intention in response to the motive, and the third is the internal programming of the utterance through a code of subjective meanings. This is the stage when the speech intention is given subjective shape through the predicative and sense structure of inner speech. Once the inner program has been formulated there is an important transformation of the internal utterance into the external utterance through mediation of an objective linguistic code. At this post-programming stage, personal, subjective meanings are translated into objective meanings with the aid of external words, and the grammar of thought is transformed into the grammar of external speech. It is at the post-programming stage that an additional translating stage would be introduced in FL production. (See Akhutina, 1978, p. 23, for a similar diagram of Leontiev's six-step speech production schema). programming (planning)

motive

thought (speech intention)

inner programming inner speech

post-programming external speech lexical and grammatical elaboration

motor implementation

Figure 2-3. A. A. Leontiev 's schema of speech act stages from motive to external speech.

The Psychophysiology of Inner Speech: Sokolov 's View The significance of A. N. Sokolov within the study of inner speech is attested by the fact that, to this day, his book Inner Speech and Thought (1972) remains the only single-authored book exclusively dedicated to the subject. In his exhaustive treatise on inner speech, Sokolov not only traced the historical development of the concept of inner speech and provided a thorough overview of research on inner speech but also presented the findings of an impressive series of experiments on inner speech conducted by himself and colleagues. Whereas theoretically Sokolov's work was grounded on sociocultural principles of the Vygotskyan school, methodologically his approach was mostly psychophysiological. In other words, his research explored the physiological correlates of the psychological study of inner speech primarily by means of electromyographic, electroencephalographic, and speech interference techniques. (A detailed discussion of these techniques is offered in Chapter 4.) Sokolov's definition of inner speech, from the psychological point of view, is one of the most elegant and transparent in the sociocultural theory literature. Because of its clarity and force and its relevance to the study of inner speech from an L2 perspective (as will be seen in further chapters), it is worth quoting here at length: In psychology, the term "inner speech" usually signifies soundless mental speech, arising at the instant we think about something, plan or solve problems in our mind, recall books read or conversations heard, read and write silently. In all such instances, we think and

44

CHAPTER 2

remember with the aid of words which we articulate to ourselves. Inner speech is nothing but speech to oneself, or concealed verbalization, which is instrumental in the logical processing of sensory data, in their realization and comprehension within a definite system of concepts and judgments. The elements of inner speech are found in all our conscious perceptions, actions, and emotional experiences, where they manifest themselves as verbal sets, instructions to oneself, or as verbal interpretation of sensations and perceptions. This renders inner speech a rather important and universal mechanism in human consciousness and psychic activity Despite its specificity (soundlessness and fragmentariness), inner speech, far from being an independent entity is a secondary phenomenon derived from external speech-auditory perception of the speech of other persons and active mastery of all the forms of the spoken and written word. Seen from this viewpoint, inner speech represents a psychological transformation of external speech, its "internal projection," arising at first as a repetition (echo) of the speech being uttered and heard, but becoming later its increasingly abbreviated reproduction in the form of verbal designs, schemes, and semantic complexes operating not unlike "quanta" of thought. From these psychological descriptions, inner speech emerges as a rather intricate phenomenon, where thought and language are bound in a single, indissoluble complex acting as the speech mechanism of thinking, (p. 1)

This passage gathers, in a nutshell, Sokolov's view of the essential features and functions of inner speech as well as summarizes his social theoretic perspective of the phenomenon. Highlighted in the above is the silent, covert, fragmentary, and selfdirected character of inner speech; its involvement in various kinds of mental operations (planning, memory, perception, reasoning, etc.); its social derivativeness and thus ontogenetically secondary nature; and its crucial role as an instrument of thought. Sokolov's main task was to obtain empirical evidence for these assumptions. One of the greatest achievements of Sokolov's (1972) research is that it managed to render tenuous and fugitive inner speech almost palpable. By measuring-and interfering with-activity of the speech muscles during silent verbal tasks, Sokolov was able to document and analyze the phenomenon of inner speech and make inferences about its structure and functions. From a phonological point of view, Sokolov found inner speech to be inaudible to others but sonorous to the self. Although Sokolov agreed with the usual characterization of inner speech as soundless, he warned that the term is "justified only from the point of view of an outsider; for the thinker himself, however, inner speech remains linked to auditory speech stimuli even in the case of maximal inhibition of speech movements" (p. 55). From a structural point of view, Sokolov noticed two forms of inner speech: an abbreviated, condensed version that he regarded as thinking in allusions to words and an expanded, elaborated form that he referred to as inner talking or verbal reasoning (p. 121). Both forms are always interacting and are equally important: "While engaged in thought, we constantly pass from thinking-reasoning to thinking in allusions to words" (p. 121). In Sokolov's (1972) view, when mental habits become fully automatized, such as in reading or solving mathematical problems, the verbal reasoning that supports these operations is highly reduced and inner speech is in its most abbreviated form. This reduction in inner speech implies not only absence of vocalization but also a compression and rearrangement of the verbal structure resulting in an internal language of "semantic complexes" (p. 71), that is, "reduced verbal statements sometimes combined with graphic images" (p. 71). One of the functions of these verbal semantic

THINKING WORDS IN ONE'S FIRST LANGUAGE

45

complexes, which Sokolov explored through experiments involving translation of FL texts (to be discussed in Chapter 3), is to single out "semantic points of reference, or key words," (p. 78) which help generalize information during reading or listening. Sokolov referred to this function of inner speech as "semantic generalization" (p. 6). Sokolov's (1972) speech interference experiments clearly demonstrated that inhibition of inner speech during the realization of verbal tasks has detrimental effects not only on understanding but also on remembering. One of his experiments, for example, showed that it is very difficult to remember what a speaker says on the radio if one tries to listen to the speaker and simultaneously count to oneself (one, two, three, etc.). In this case, artificially suppressing the inward repetition of the speech heard results in "instantaneous amnesia" (p. 95). This type of experiment led Sokolov to stress the extreme importance of speech movements, or inward repetition, in memorization. Further tests, however, indicated that when the verbal act is sufficiently automatized, speech interference does not lead to forgetting. In these cases, automatization makes it possible to engage in maximally reduced inner speech, or extremely abbreviated inward repetition, and thus the mental act is not impaired. Sokolov called this function of inner speech "semantic memorization" (p. 6). To Sokolov (1972), the concurrent repetition of speech heard or read is crucial for understanding and retention. Some level of covert enunciation, even if minimal, is necessary for retention of words in written text or oral speech. What leads to retention, however, is not so much repetition per se, but the abstraction and generalization of meaning that occurs as words and phrases are singled out in the process of reading or listening. In other words, the importance of repetition resides in its power to generate understanding, and thereby memorization, of the general meaning of what is heard or read: In our experiments inner speech was a means of fixation not so much because it enabled one to reproduce the speech heard as because it generalized it, creating a logical scheme of material and fixing thereby the contents of the speech heard, (p. 112)

Inner speech, in the form of concealed verbalization, is not limited to the act of listening or reading. By means of electromyographic measurements of speech musculature, Sokolov was able to determine that strong speech tensions were also involved in the recollection of phrases heard or past recent events (p. 180). He thus recognized a "mnemonic function" by which inner speech aids in retrieving perceived stimuli from memory (p. 186). As mentioned, Sokolov (1972) found that "inner speech is not always abbreviated . . . . At times it may occur in a very unfolded [expanded] form. For instance, . . . when we reason or argue with ourselves" or when we read difficult texts (pp. 115-116). In his experiments, when the possibility of engaging in expanded inner speech was eliminated, the subjects experienced great difficulty in understanding texts of an abstract nature or written in an FL (see Chapter 4 for further details on these experiments). In these cases, the enunciation of fragmentary inner speech was not sufficient. "What is necessary is unfolded inner speech and, at times, even explicit articulation of words" (p. 117). According to Sokolov, suppression of unfolded inner

46

CHAPTER 2

speech articulation hinders understanding of difficult material for two reasons: (1) it prevents the analysis that comes from singling out important information and making generalizations about it and (2) it does not allow retention of the material. From the above, it is clear that, for Sokolov (1972), inner speech performs two essential functions: semantic generalization and memory fixation and retrieval. Yet another important function of inner speech identified by Sokolov is to serve as preparation for outward communication: "Although inner speech cannot, as such, serve as a means of direct communication among people and is chiefly a vehicle for thought..., it nevertheless carries out very important preparatory functions for human communication" (p. 66). Sokolov believed that external speech is dependent on inner speech. This is evident, according to him, when we stop to think before speaking. In this case, we first fix thoughts in our mind by means of inner speech, we plan what we want to say, and we formulate a synopsis of the future utterance (p. 65). This preparatory function is even more evident in writing. Planning a written statement usually requires an elaborate process of contemplating alternatives and selecting the appropriate words. In Sokolov's opinion, inner speech is always present in the planning of an utterance, no matter how immediate or fast the transition from thought to external word is. This transition is extremely complex because it involves the transformation of reduced, subjective inner speech to expanded, understandable external speech. Like Vygotsky, Sokolov (1972) believed that inner speech not just expresses but fulfills thought (p. 3). In Sokolov's view, human thinking is essentially verbal thinking (p. 2) because speech is instrumental in the "intellectualization" or shaping of perceptions and sensations (p. 3). Mastery of a language means that nonverbal forms of thinking (pictorial, practical, or concrete) "will occur within the conceptual framework of language, i.e., on the basis of previously acquired concepts that are retained in memory and are subsequently actualized in the form of concealed, or inner, speech" (p. 3). But, Sokolov admitted, thought and language continue to be separate entities. The fact that ideas can be expressed in different linguistic codes and that inner speech may sometimes be complemented by graphic images indicates that language is not inseparable from thought. Nevertheless, Sokolov found inner speech to have a pervasive role in all types of thinking. He demonstrated this in a series of studies involving mental tests, such as Raven's matrix problems, which were designed to observe concrete thinking and were based on visual, graphic stimuli. Increased electrical activity of the speech musculature during solution of these tests revealed that what was perceived visually-through the "first signal" system-was translated into speech-the "second signal" system (p. 236). The fact that all subjects displayed some level of motor speech activity, no matter how subtle, while solving the mental tests indicated that inner speech is "a necessary component of thought activity" (p. 237). In short, Sokolov's contribution to the study of inner speech lay in demonstrating the internal, psychophysiological reality of inner speech while emphasizing its origins in external sources. His research forcefully illustrated the involvement of inner speech in all forms of thinking, from verbal to conceptual to concrete thought, as well as its role in critical mental processes, such as comprehension, memorization, planning

THINKING WORDS IN ONE'S FIRST LANGUAGE

47

speech production, and problem solving. Sokolov's research also plainly revealed the essential structural characteristics of inner speech, from its soundlessness to its frequent verbal-pictorial nature to its shifting patterns of formal expansion-reduction. Inner Speech as Intrapersonal Communication: Vocate's View Invoking ideas from Vygotsky's and Luria's sociocultural theory and from Mead's symbolic interactionist theory, communication expert Donna Vocate (1994b) discussed inner speech in the context of intrapersonal communication. Vocate fundamentally distinguished between inner speech and intrapersonal communication, two constructs that are often, in her opinion, mistakenly equated. In her view, inner speech, as "the cognitive process linking thought and utterance" (p. 15), makes possible all levels of human communication (interpersonal, intrapersonal, and public communication). Inner speech is thus implicated in, though not restricted to, intrapersonal communication, a level of speech performance in which the "communicator is both the source and the object of the interaction" (p. 6). Specifically, Vocate was interested in self-talk, a particular type of talk occurring at an intrapersonal level in which the object of the interaction is oneself. Vocate defined self-talk "as a dialogue with the self existing in two forms: (a) the silent, internal dialogic process of inner speech, and (b) the audible, external dialogue addressed to self although others may hear it" (p. 6). According to Vocate, self-talk is characterized by four elements: self-awareness, a dialogic nature, a stimulus, and a response, the last two semiotic in nature and self-originated. The element of self-awareness is necessary, even at a minimal level, "in order to address the self as an object" (p. 7). (For other accounts of the relationship between self-talk and self-awareness, see Morin, 1993, 1999; Morin & Everett, 1990; and Siegrist, 1995). A major point in Vocate's (1994b) discussion of self-talk is its role in the creation and development of the self. Vocate, in line with Vygotsky and Luria, believed the self is a product of the internalization of social speech as inner speech. In the process of internalizing the spoken language of the community, the child constructs a social self based on the cultural attitudes and beliefs of those around him/her. At first, this self is "simply an abbreviated clone of its social milieu" (p. 8), but as the child becomes aware of his/her own speech and the responses he gets from others, he/she develops a dual concept of self based on his/her own individual perspective and that acquired from others (the /and the Me, in Mead's terms). It is the interaction between these two sides of the self, via inner speech, which creates self-consciousness. "Talk physically creates a new awareness of self... and elevates consciousness to the uniquely human level of self-reflection" (p. 10). According to Vocate (1994b), self-talk has two purposes: the creation of meaning and cognitive adaptation. Vocate described the first as an interpretive process by which new meanings are created. Self-talk, as an intentional dialectic between society and the individual, becomes the venue for the resolution of tensions between subjective and objective meanings {sense and meaning, respectively, in Vygotskyan terms). Selftalk is also carried out for cognitive adaptation purposes. In other words, through self-

48

CHAPTER 2

talk, the self mentally adapts; it adjusts to the environment and adjusts the environment to the self. Drawing from what she interpreted to be Luria's concept of inner speech as an essentially encoding/decoding process that moves from thought to utterance and vice versa, Vocate (1994b) proposed a view of inner speech as a coding process that involves three stages (see Figure 2-4). Transitionally located between an initial motive and a fully expanded utterance, the three stages of inner speech coding are a semantic set, which consists of purely semantic elements, simultaneous relationships, and no lexical or syntactic form; a deep structure stage, in which semantic relationships begin to develop a primitive sequential predicative form; and a surface structure stage, which includes some lexis and a syntactic structure, mainly elliptical and predicative.30 According to Vocate, self-talk is possible only at the deep and surface structure levels. At the deep structure stage, inner speech is sufficiently developed syntactically to support communication with oneself, though not with others. The surface structure stage is, however, the optimal level for self-talk in Vocate's opinion because it allows the subject maximal consideration of "sense" and "meaning" (p. 22). Inner speech, however, may not always progress through all of the three stages: "It may be arrested at any of these and not continue. For some self-talk messages, coding at the level of deep structure may be sufficient to address the self (p. 22). inner speech coding

motive semantic set

deep structure

surface structure

expanded utterance

Figure 2-4. Vocate's conception of the communicative act from motive to expanded utterance.

As Vocate (1994b) explained, not all inner speech activity involves self-talk; in other words, not all inner speech consists of a self-conscious dialogue with the self. Some inner speech coding may "occur passively below the level of self-awareness as in scripted or exclamatory utterance" (p. 23). But if self-talk is to occur, it is at the deep and surface structure stages that the subject can shift "to the dialogic function of inner speech" (p. 23) and consciously take an active role in the process of interpretation that accompanies the syntactic and semantic progression to a full expanded utterance. Vocate's argument is critical because it shows that, although not all inner speech occurs under voluntary attention and control, it can be intentionally 30

For her proposed stages of inner speech, Vocate (1994b) utilized as basis the thought-to-utterance schema that Luria developed "at the end of his career" (p. 20), more specifically in his 1976 book Basic Problems of Neurolinguistics. This schema includes constructs such as deep structure and surface structure, which reflect the impact that Chomsky's linguistic theory had at one point in Luria's career (see Vocate, 1987, pp. 117-118).

THINKING WORDS IN ONE' s FIRST LANGUAGE

49

accessed and manipulated in the active process of silently talking to oneself. Vocate's (1994b) treatment of inner speech and self-talk as aspects of intrapersonal communication is extremely valuable in its fine distinction of terms and in its detailed exploration of the implications of a view of self-talk as an inner speech phenomenon that can be intentionally monitored by the individual. Vocate's view of inner speech, it should be observed, however, is not the only one within an intrapersonal communication framework, although it is the one that most explicitly adheres to a sociocultural approach to inner speech. Several other perspectives on intrapersonal communication are to be found in Dance and Larson (1976), Johnson (1984), Roberts and Watson (1989), Vocate (1994a), and Aitken and Shedletsky (1997). An interesting outlook was offered by Honeycutt et al. (1989), who argued for the term imagined interactions, rather than self-talk, to refer to intrapersonal communication on the account that in their mental conversations individuals not only talk to themselves (or to versions of their own selves) but also imagine themselves talking to others. Overview of the Sociocultural Approach to Inner Speech In review, the picture of LI inner speech that emerges from sociocultural theory is of an extraordinarily complex and consequential phenomenon, deeply implicated in the construction and operation of the individual mind. As surmised from an analysis of some of the most representative literature, namely Vygotsky's, A. A. Leontiev's, Luria' s, Sokolov' s, and Vocate' s work, the sociocultural approach to inner speech rests on the following assumptions: • Inner speech results from the convergence of thought and speech at an intrapsychological level. • Inner speech is the mechanism for verbal (symbolic) thinking. Speech and thought come together in inner speech, but thought and speech remain independent phenomena. • Inner speech is not a natural faculty of the mind; it derives from social, external speech and is therefore secondary in nature. Inner speech is internalized social speech. • Inner speech in one's LI is the culmination of a developmental process involving the transition from social, communicative speech; to vocalized egocentric speech; to silent, self-directed internal speech. • On the basis of various sign systems (predominantly language, but including other semiotic resources), inner speech mediates mental activity making possible the operation of higher psychological processes. • The main function of inner speech is to serve as a tool for thinking. This implies a crucial role in a broad spectrum of mental functions: giving shape and control to one's own thoughts, reasoning, planning action, regulating behavior, solving problems, comprehending, and remembering. • Inner speech has an important intra-communicative function. It is the vehicle for self-communication or self-talk. • Inner speech is instrumental in the realization of verbal communicative tasks.

50

• •

•

•

•

•

•

•

•

• •

CHAPTER 2

Listening, reading, speaking, and writing, all involve inner speech. In listening or reading, inner speech represents the transformation of words into thoughts; in speaking and writing, it is the opposite, the turning of thoughts into words. In planning the spoken or written utterance, inner speech has an essential rehearsal or speech preparatory role. Inner speech is not overtly vocalized, and thus it is ostensibly a silent form of speech. Its realization, however, is frequently accompanied by subtle movements of the vocal organs as well as phonological traces in the speaker's mind. Inner speech has a sui generis syntax and semantics. Syntactically, inner speech is abbreviated, fragmented, and elliptical. Semantically, it is condensed but rich in allusions to the senses of words. Inner speech is typically abbreviated and condensed, particularly during automatized mental acts. It may become expanded and fully elaborated, however, as in preparation for a lecture or in rehearsing written production. As a secondary phenomenon derived from social forms of discourse, inner speech has a dialogic dimension. The Bakhtinian version of dialogism in inner speech would hold that, as the condensation of other people's sociohistorically situated voices in the individual mind, inner speech is always dialogic. For other sociocultural theorists, inner speech is dialogic in self-talk, that is, when adopting the style of a dialogue or conversation with the self. Inner speech is not an internal language system or even a network of verbal connections. Inner speech is the process of putting language in action for thinking; it is the use of language to mediate thinking. Inner speech is characterized by stages. There are ontogenetic phases in the internalization of social speech as well as microgenetic planes in the formulation of thought, in the reception of verbal stimuli, and in the production of the speech act. Inner speech is activity-intrapersonal, rather than interpersonal activity. It is goaldirected, purposeful mental activity, mediated by language and mobilized in response to specific motives or intentions. Consciousness, self-awareness, and introspection are to a large extent mediated by inner speech. Not all inner speech, however, occurs at the level of consciousness or takes place under voluntary control. Disruption of inner speech activity by mechanical means (as in suppression of subvocalization) or cerebral damage severely impairs normal mental functioning. Inner speech processes can be affected at various stages by differently localized brain disorders. COGNITIVE APPROACHES TO INNER SPEECH: A MISCELLANY

The existence of inner speech is hardly disputed. Its conceptualization, however, varies greatly among the different theoretical approaches to mind and cognition. How the mind is conceived, as a predominantly social formation or as a function of biologically specified mechanisms, determines what origin, roles, and particular

THINKING WORDS IN ONE ' s FIRST LANGUAGE

51

relationship to thought are attributed to inner speech. The following accounts of inner speech illustrate the various positions that have been adopted towards the phenomenon outside the purely sociocultural perspective. As will be seen, there is a wide range of approaches, from attempts to blend the external with the computational in inner speech (Frawley's sociocomputational and Clark's supracommunicative views), to decidedly a-social outlooks (Carruthers's modular hypothesis) and rather reductionist views of inner speech as an internal mnemonic device (information-processing approaches). A Sociocomputational Approach to Inner Speech In an effort to develop a Vygotskyan cognitive science, Frawley (1997) propounded the reconciliation of sociocultural theory and computationalism. His basic argument was that the mind is both social and computational. The mind is social to the extent that consciousness originates in cultural specific contexts and is mediated by language, a social artifact. But the mind is also computational to the extent that it needs a computing brain to register the social units of mind. Frawley claimed that social language, internalized as private or inner speech, is effective in mediating the brain's computational activity. Private-inner speech provides control of computational processes. In turn, the computational mental system allows for internal representation of socially derived units of information. According to Frawley (1997), when people talk to themselves during problem solving, they deploy language for thought in the form of private (audible) or inner (silent) speech. This language/or thought, which encodes the individual's internalized view of the social world, regulates thinking and frames the language of thought, in which the ultimate mental computations for problem solution are carried out (p. 31). Private-inner speech can only exercise control on the mind's computational system at the level of metaconsciousness (or self-consciousness). In metaconsciousness, defined as "deliberate awareness for mental and behavioral control" (p. 6), attention can be directed to and control can be exercised on a variety of under-specified or nonspecified representational and computational features through the language for thought. The language for thought, for example, can specify semantic properties not coded in mental representations or can provide context for decontextualized representations (p. 186). Because different languages have different ways of systematizing relationships between cognitive architecture (inner mental structure) and linguistic code, Frawley (1997) argued that the activity of language for thought at the level of metaconsciousness varies from language to language. Thus, Frawley reinterpreted Whorf s relativity hypothesis as a claim about how different languages, with their different ways of representing the world, affect the on-line thinking process that characterizes metaconsciousness (p. 232). In summing up, then, it might be said that Frawley's sociocomputational view of the language for thought, or the linguistic code utilized in inner-private speech, is an account of how certain cultural specifics of information coded in internalized social language(s) can impinge on the cognitive architecture of the mind.

52

CHAPTER 2

Clark's Supracommunicative View of Inner Speech Clark (1998) advocated a supracommunicative conception of language and cognition. This view holds that language, in addition to facilitating interpersonal communication, has the potential for "transforming, reshaping and simplifying the computational tasks that confront the biological brain" (p. 163). For Clark, as for Vygotsky, language is an external artifact designed to facilitate interpersonal communication as well as a selfdirected tool used for guiding and controlling one's own behavior, harnessing attention, and shaping one's thought processes. Clark's basic postulate is that language significantly complements and enhances the cognitive architecture of the brain by subtly modifying its basic computations and patterning abilities. Human cognition is extended, and thus transformed, by means of external linguistic resources. A fundamental way in which language affects cognition, according to Clark (1998), is by offering the possibility to stabilize thought by creating temporarily fixed linguistic structures in the mind through mental sentence rehearsal. Linguistically formulated and briefly held in working memory, a thought can become the object of attention in inner speech. We can think about our thoughts, hold them to inspection, and critically evaluate them. Through language, cognition becomes meta-cognition. Clark argued that language is particularly well-suited for this because, by being a highly conventionalized and permanent code, it reduces the effects of contextuality,31 allows for common representation of various input modes (visual, oral, tactile), and facilitates memorization. In Clark's view, then, inner speech would have a "freezing" (p. 178) metacognitive function whereby thoughts, rendered in linguistic form, would become temporarily the object of metaconscious scrutiny. A Modularist View of Inner Speech Carruthers (1996,1998a, 1998b) endorsed the cognitive conception of language, which holds that language is constitutive of some of our conscious thoughts, and that much of our thinking is done in language. In this way, he distanced from the communicative view of language, which assigns no essential role for language in thinking and sees language as a mere vehicle for thought. Despite the apparent affinity of the cognitive conception of language with Vygotskyan positions on the interdependence of thought and speech, Carruthers did not adopt a sociocultural approach to problem. In his view, mind and language, rather than being social constructs, are innately determined and modular in structure. Thus, his cognitive conception of language is nativist and modularist in kind, though he departs from the total separation of language and thinking that characterizes Fodor's (1975, 1983) modularity hypothesis. For Carruthers (1998a), language is not just an isolated module of mind, but "is directly

31 Clark's (1998) view of the contextual effects of inner speech is quite different from Frawley's (1997). Whereas Clark believes words minimize contextuality by virtue of being relatively stable in meaning in different sentences, Frawley holds that language, through inner or private speech, provides essential contextual clues not coded in mental representations.

THINKING W O R D S IN O N E ' s FIRST LANGUAGE

53

implicated in central cognitive processes of desiring, believing, and reasoning" (p. 458). In addition to an innate and modularized language faculty, Carruthers (1996) posed the existence of a reflexive thinking faculty, responsible for conscious thought, which has the capacity to access natural language knowledge and deploy it in the imagination in the form of inner speech. 32 Carruthers (1996) maintained that most of our private thoughts are done in inner speech by deployment of natural language sentences (p. 50). He used introspective evidence, such as that collected by Hurlburt (1990) through the thought sampling technique (to be discussed in Chapter 4) and L2 learners' reports about the moment when they begin to think in the language they are learning, to show that people's conscious thoughts are conducted in a natural language and not in a universal, symbolic, innate language of thought such as Mentalese (to use Fodor's label for such a language). According to Carruthers (1996), there should be no doubt that inner speech exists. The question is what role it plays. For him, inner speech has a thinking role (p. 225). It is the reflexive thinking faculty that makes it possible for humans to think about their thoughts without engaging in self-interpretation (p. 226). When we hear an utterance in inner speech, Carruthers contended, we do not just experience a phonological representation but an immediate semantic interpretation of it. What we hear "is already interpreted-I hear meaning in the words" (1998a, p. 463). So, the imaged sentence is not an interpretation of thought but is directly constitutive of it. In other words, inner speech is propositional, conscious thinking (1996, p. 238). Here Carruthers departs from Vygotsky's view of speech and thought as independent and distinct phenomena despite their convergence in inner speech. Information-Processing

Perspectives on Inner Speech

A substantial body of work has been devoted to the inner speech activity that occurs as a function of memory and learning. This research has an information-processing approach posing short-term memory as the locus of inner speech activity. From this perspective, inner speech is a form of covert rehearsal presumably influencing longterm storage of information. Based on psycholinguistic methods of experimentation, such as word imitation, articulatory suppression of repetition, word recall, and rhyming judgments, this line of research usually portrays inner speech as the interaction between the inner ear (how words are heard in the "head") and the inner voice (how words are inwardly articulated). There are thus two main foci of attention among these studies: sub vocal rehearsal and auditory imagery. The working-memory model (Baddeley, 1986) lies at the heart of this research. Working memory is composed of three elements: the central executive, a system that regulates the processing, storing and retrieving of information, and two subsystems

32 In Carruthers (1998b), it would appear that the reflexive thinking faculty is a component, or one of the functions, of the central executive, a central processing system of the mind that has the capacity to generate inner speech.

54

CHAPTER 2

specialized for the temporary retention and maintenance of information: the phonological loop, which handles verbal oral data, and the visuo-spatial sketchpad, which works with visual and spatial representations. The phonological loop, in turn, consists of a phonological store, which keeps information in phonological form, and an articulatory rehearsalprocess, which maintains and refreshes decaying information in the phonological store (Baddeley, 1986). The phonological store and the articulatory rehearsal process would correspond, respectively, to the terms inner ear and inner voice, while a third term, inner eye, would be reserved for the visuo-spatial sketchpad (Baddeley & Lewis, 1981, Gathercole & Baddeley, 1993). Baddeley, Gathercole, and Papagno (1998) proposed that the main function of the phonological loop is to mediate language learning. Specifically, "the function of the phonological loop is not to remember familiar words but to help learn new words" (p. 158). The task of the phonological loop would be to hold and rehearse new words in working memory while long-term representations of that material are formed. Thus, the phonological loop is thought to be extremely important in language acquisition because of its role in forming long-term representations of novel lexical material. Although tentatively, these researchers also suggested that the phonological loop may be implicated in the learning of syntax. Other research (to be discussed in Chapter 3) has found the phonological loop to be strongly related to FL acquisition. The phonological loop is thought to be involved in the processing of auditory as well as visual verbal input, such as in silent reading. It has been speculated that during silent reading subjects activate acoustic representations in inner speech (Abramson & Goldinger, 1997) and verbalize visual information through sub vocal repetition before storing it up as phonological representations (Gillet, Espagniet, & Billard, 1997). It has been shown, however, that subvocal rehearsal during silent reading is not indispensable for auditory imagery to take place (Baddeley & Lewis, 1981). What sub vocalization during reading may need is not actual verbalization but only some form of auditory rehearsing inside the head (Baddeley & Logie, 1992). The research thus speaks in favor of a phonological loop role during silent reading based on auditory images that make it possible to hear words even when not spoken. Auditory imagery has been recognized as an important function of phonological working memory. Its exact nature, however, is debated. Smith et al., (1992) argued that both the inner ear (phonological store) and the inner voice (rehearsal) are involved in auditory imagery: "Most likely, subjects produce the strings [they read] with the inner voice, 'listen' to themselves with the inner ear, and then make their judgments based on this auditory image" (p. 97). Campbell (1992) interestingly pointed out that even people who are born deaf have phonological inner speech (that they have signbased inner speech is not questioned) although this inner speech is qualitatively different from that of normal hearing adults. Her argument is that deaf individuals get their phonological (not acoustic) inner speech from lipreading, that is, from the visual

THINKING WORDS IN ONE ' s FIRST LANGUAGE

55

articulatory and phonetic information that lip movements convey.33 In Campbell's view, auditory input is not necessary for development of inner speech. Further evidence of this is the case of Rebecca, a well-educated normal hearing adult who does not have phonological inner speech (possibly, Campbell speculated, because of some disruption of the loop between the inner voice and the inner ear). Rebecca's case would indicate that auditory imagery is not indispensable for inner speech. According to MacKay (1992), theories on the auditory and articulatory nature of inner speech need to address a number of unexplained phenomena or insufficiently supported claims. Some of these constraints on inner speech theory are the nonauditory and nonarticulatory aspects of inner speech, in other words, why certain features of auditory imagery and rehearsal are not essential in inner speech. To illustrate his point, MacKay pointed out that several aspects of the acoustics of overt speech, such as loudness and pitch, "are normally absent from our awareness of selfproduced internal speech" (p. 128). He also noted what he calls the unrehearsability of language in inner speech: "Sentences heard in a foreign language are unrehearsable even if the phonology of these sentences is compatible with English phonology . . . . For example, a single phoneme such as a German trilled /r/, cannot be accurately rehearsed by someone unfamiliar with German" (p. 142). Other puzzling aspects of inner speech that need to be explained, according to MacKay, is why mental rehearsal is sometimes involuntary and uncontrollable and the differential effects of overt and internal rehearsal on speech production. Overview of Cognitive Approaches to Inner Speech Various approaches to inner speech as a cognitive mechanism have been proposed outside the strictly sociocultural theory paradigm. Some, as has been seen, do not stray too far apart from fundamental sociocultural principles about the social nature and origin of inner speech. Frawley and Clark, for instance, purported to extend the Vygotskyan view of inner speech as an internalized social artifact by complementing it with speculations about the way in which inner speech affects the pre-determined cognitive architecture. Both Frawley and Clark blurred boundaries between the social and the hard-wired mind by viewing inner speech as a tool for thought that can be effectively deployed in metaconsciousness for computational purposes. Frawley and Clark differed, however, on precisely how language, through inner speech, interacts with the basic properties of the mind. To Frawley, the language for thought is an effective mechanism in regulating thinking by providing essential contextual, semantic, and pragmatic information to the computational units of mind. To Clark, inner speech in the form of mental rehearsal enhances central cognition by stabilizing thoughts in

33

According to Campbell (1992), the fact that educated deaf people can judge rhyme in written words suggests that they too have some form of auditory inner speech. This, Campbell argues, may be the effect of literacy: "It is extremely important to teach the deaf child to read and write and, if this training is alphabetic,... it may enhance the deaf child's skill in using inner speech" (p. 86). Development of literacy would thus appear to encourage development of inner speech

56

CHAPTER 2

working memory and making them the object of attention and reformulation. Other approaches take a more internalist (as opposed to externalist, Frawley, 1997) approach to inner speech. Carruthers, for example, openly rejected the notion of mind as a cultural construct and therefore conceived of inner speech as a normal function of the reflexive thinking faculty that is part of the central executive system. Informationprocessing approaches simply bypass the debate between externalism and internalism by ignoring issues about the genesis and mediating role of inner speech and focusing exclusively on the phenomenon as a function of a biologically specified, self-sufficient memory system. Discussion of inner speech is reduced in this field of research to discrete analyses of language-based memory processes such as the interactions in the phonological loop between articulatory rehearsal (the inner voice) and the phonological store (the inner ear). Within this framework, inner speech-as instantiated in the phonological loop-attains a major role in language learning, especially in vocabulary acquisition. The exact nature of some inner speech processes such as auditory imagery and subvocal rehearsal is, however, still not clearly understood, but important advances have been made into specifying the perceptual and motor variables that constrain individual performance in inner speech. BRAIN IMAGING: TECHNOLOGY IN SEARCH OF INNER SPEECH The study of inner speech is being bolstered by new technologies. In particular, two brain imaging techniques, positron emission tomography (PET) and magnetic resonance imaging (MRI) allow researchers to examine the functioning of the brain during inner speech activity (Mazziotta, 1994;Posner&Raichle, 1994;Raichle, 1994, 1997). How can these technologies contribute to the investigation of inner speech? By mapping the human brain in action, these techniques make possible visualizing the on-line processes of inner speech. Brain imaging also offers precise information of brain areas involved in language processing. Utilizing evidence from brain imaging studies, Morin (1999), for example, has uncovered evidence suggesting that inner speech and self-awareness share a common neurological area: the left inferior frontal region. The application of these technologies has vast potential for confirming (or discontinuing) hypotheses, uncovering unknown aspects, or clarifying obscure features of inner speech. Following are some of the findings of brain-imaging inner speech research. (For a detailed description of these technologies and their possibilities in the study of inner speech, see Chapter 4.) A few studies have looked at activated brain areas during various forms of inner speech activity. Employing PET technology, Paulesu, Frith, and Frackowiak (1993) found neural evidence for two distinct components of the articulatory loop (Baddeley, 1986): the subvocal rehearsal system and the phonological store. Two silent verbal tasks (rehearsing letters and judging rhyme) in the study led to the localization of the phonological store in the left supramarginal gyrus and the subvocal rehearsal system in Broca's area. Both tasks also activated the left lingual gyrus, which is related to the processing of visual letters, and motor speech areas despite the absence of overt speech.

THINKING WORDS IN ONE ' S FIRST LANGUAGE

57

Using regional cerebral blood flow (rCBF) measurements, Ryding, Bradvik, & Ingvar (1996) compared activation of brain areas during inner and audible speech in both left and right brain hemispheres, while subjects were quietly counting to themselves silently or aloud. Multiple regions in both the left and right hemispheres, especially in the prefrontal and supplementary motor areas, were activated during silent counting, suggesting an internal feedback loop or interaction between both hemispheres. The study offered strong evidence that some forms of inner speech activity call for the concerted efforts of both sides of the brain. More recently, Shergill et al. (2001), using MRI, investigated the neural correlates and functional anatomy of inner speech and different forms of auditory verbal imagery. Silent vocalization of sentences was compared to imagining speech in different "voices" (as if the first, a second, or a third person's voice were heard in the mind). Inner speech during silent articulation "was associated with activation in the left inferior frontal/insula region, the left temporo-parietal cortex, the right cerebellum, and the supplementary motor area" whereas auditory verbal imagery in general was associated with the same areas engaged during the inner speech task "plus the left precentral and superior temporal gyri, and the right homologues of all these areas" (p. 241). Interestingly, imagining a second or third person's voice generated greater activation of areas associated with covert articulation. The three studies reviewed here are a small sample of the work already done in exploring the neural correlates of inner speech through brain scanning techniques. Studies such as these provide a glimpse of the amazing complexity of the brain activity involved in the mind-stirring business of silently talking to oneself. From the images that are emerging, one gets an inkling of that intricacy and fluidness of inner speech that Vygotsky recognized in his time but, admittedly, did not yet fully understand. Despite the complexity that is unfolding, however, this type of research, like the information-processing approach on which it is predicated, is still highly reductionist. What seems to be lacking is placing the new discoveries within a larger context of interpretation that includes questions about the role played by inner speech in consciousness and shaping the human mind. REVIEWING THE LI LITERATURE: IMPLICATIONS FOR L2 INNER SPEECH This chapter has presented several propositions about the nature of inner speech. While theoretical perspectives varied, the constant focus of the research has been inner speech in the LI. With a few exceptions, references to inner speech in an L2 were left out. The purpose of this last section is then to explore some of the major implications for L2 inner speech that can be drawn from the LI literature. They are presented here in the form of queries, many of which will be addressed in the next chapters. • Thought and speech develop along separate lines until they come together in (children's) egocentric speech. When this occurs in LI acquisition, the L2 is frequently out of the picture. At what point do thought and the L2 converge in L2 acquisition? • Development of inner speech occurs in one's LI as an evolution from social, to

58

•

•

•

•

•

•

•

• •

CHAPTER 2

egocentric, to inner speech. Does development of L2 inner speech recapitulate this sequence? Is egocentric (private) speech a phase in the internalization of the L2? Inner speech in the LI is characterized by brevity, compactness, and fluidity. Is L2 inner speech similar in structure? Does L2 inner speech lack any of these characteristics? In what circumstances? Inner speech goes through phases as it moves from thought to external speech. The LI enters early in this transition by providing the basic semantic component of inner speech. What is the L2 point of entry? Can the L2 be accessed as early as the LI, or is it always a posteriori? Does L2 inner speech involve a translation phase, from the LI code to the L2 code? For a bilingual or L2 learner, the LI is always in the mind-or rather, the mind is constructed on the basis of the LI. What then is the role of the LI in L2 inner speech? In advanced bilinguals, or in cases of LI attrition, can a different language for thought become an alternate, or even the primary, mediating code? Does a different mind, mediated by the L2, emerge? As an instrument of thought, inner speech in the LI is involved in a multitude of mental operations. To what extent is the L2 involved in the myriad mental functions that are normally mediated by the LI? Adult human beings talk to themselves in their LI. What role does the L2 play in intrapersonal communication? What purposes does talking to oneself in the L2 serve? If one's own thoughts in the LI can be attended to and manipulated in metaconsciousness, can the same occur with the L2? Would this activity have positive effects on L2 learning? Articulatory sub vocal rehearsal has an important role in L1 acquisition, particularly in the area of vocabulary. What is the role of inner speech in covert rehearsal of the L2? Auditory imagery (hearing a voice) in the L1 is a real phenomenon for most human beings. What is the nature of auditory imagery in the L2? Inner speech in the LI contributes to our experience of consciousness, selfawareness, and introspection. What is this experience like when mediated by the L2? Does a different self emerge when thinking words in another language? CONCLUSION

The focus of this chapter has been inner speech in the LI. Various conceptions of inner speech from different theoretical perspectives have been explored in a review of the most representative literature. A thorough basis for an understanding of inner speech within the sociocultural theory framework has also been established. Finally, implications for the study of inner speech in L2 learning have been presented. Many of the issues raised in this chapter on the nature of inner speech, its form and functions, its evolution, its role in shaping consciousness and a sense of self, as well as the methodology for its research, will be addressed again in the following chapters from anL2 point of view.

CHAPTER 3 THINKING WORDS IN A SECOND LANGUAGE Inner Speech: The L2 Perspective

The purpose of this chapter is to critically review research conducted on inner speech in an L2. Studies dealing strictly with the inner speech-L2 connection are hard to find, a fact that suggests that the role of inner speech in L2 learning has received only minimal attention from L2 researchers as an empirical problem to be pursued in a straightforward manner. Careful scrutiny of the L2 acquisition literature shows, however, some indirect concern with the topic in the form of speculations on the nature of verbal thought in the L2 or about learning strategies that involve the internal use of the L2. A good number of studies, particularly those framed within a Vygotskyan sociocultural perspective, do make direct allusions to inner speech and L2 learning but usually with the purpose of theoretically framing another main focus of investigation. This ancillary literature on inner speech and L2 learning is very important, however, because it offers insights into various aspects of the phenomenon, suggests hypotheses, and may serve as the basis for firsthand research on the topic. The leading question throughout this chapter is "What does the literature say about inner speech and L2 learning?" In answering this question, a variety of research studies more or less directly concerned with inner speech and L2 learning are reviewed. The studies are critically appraised for their relevance to and potential in illuminating the phenomenon of L2 inner speech. The analysis also includes a critical examination of the diverse theoretical rationales underpinning the studies. The chapter has been organized around five main thematic groups: (a) inner speech as the mechanism for verbal thought in the L2 and the related issue of "thinking in a second language;" (b) the internalization of social speech as inner speech in the L2; (c) the role of inner speech in reading and writing in the L2; (d) mental rehearsal of the L2 in its various forms: spontaneous playback or Din, covert practice of the L2, and language play; and (e) L2 inner speech activity as revealed through brain imaging technology. Overall, the chapter offers a comprehensive view of the existing research related to L2 inner speech.

60

CHAPTER 3

INNER SPEECH AS VERBAL THOUGHT IN THE L2 The idea of inner speech as the psychological interface between thought and language and as the mechanism for turning thoughts into words has not only concerned researchers working from an LI perspective but also those interested in the acquisition or teaching of L2s. The question that naturally arises from the notion of inner speech as the mechanism for thinking in or through words is how this mechanism materializes when learning or using an L2. Two major aspects of this question have been probed in the literature: (a) What does it mean to "think in an L2," and (b) what is the nature of verbal thought in the processes of L2 acquisition and use? Thinking in a Second or Foreign Language Coding Thoughts in an L2/FL: A Translation Process? In his book Psychology and the Language Learning Process, A. A. Leontiev (1981) devotes a whole chapter to the issue of "thinking in a foreign language." Leontiev is interested in the familiar claim among educators that students need to think in an FL if they want to learn it. In his opinion, this belief is based on the erroneous assumption that thinking in a particular language involves a different way of thinking: "Man does not think in a way determined by language: he mediates his thought through language to the extent to which language answers to the content and to the tasks of his thought" (p. 108). Leontiev is clearly not a linguistic relativist in this respect, but neither does he limit the idea of thinking in an FL to "attaching verbal labels to thought essences" (p. 109). For him, the idea of thinking in an FL as simply the expression of thoughts in another language is a limited and incorrect approach. He considers it is an "indisputable" but "trivial" fact that the realization of a speech act in an FL always implies utilizing a (foreign) linguistic code (p. 109). What appears to Leontiev (1981) as a significant point of contention is the extent to which the FL is involved in the programming stages of speech production. According to Leontiev, in the early stages of FL learning, the transition from inner program to external utterance in the FL is not as direct as in the LI. In his view, there is an intermediate translating stage before the actual FL utterance is produced. The post-programming stage involves first rendering the utterance in the native language and then turning it into the FL (p. 26). Leontiev holds it is the task of teachers "to 'get rid' of the intermediate stage as quickly as possible" (p. 27) by acquainting students with the rules-to be discussed in Chapter 7-governing the transition from speech operations in the LI to those in the FL. Figure 3-1 accommodates the L2 translation phase into Leontiev's schema of speech act stages as presented in Chapter 2. Two points deserve attention in Leontiev's understanding of the speech act in an FL. First is the fact that the current method used in teaching FLs in Leontiev's working context was translation from the LI to the FL (p. 27). It is possible that heavy reliance on translation as a method for learning FLs may be directly related to the presence of an intermediate translating stage from native to foreign code between the planning and the actual utterance stages of speech production, such as the one

THINKING WORDS IN A SECOND LANGUAGE

61

postulated by Leontiev. This leads to at least two questions: Would such translating stage always take place independently of the methodology used? Would this translating stage occur at all levels of acquisition or only at the beginning? The second point is related to an intriguing comment made by Leontiev in an endnote, which opens the door for the possibility of FL access at the inner speech stage: "It should be remembered that even at the programming stage foreign words and structures may be used as elements of the subjective code. But this use is of a purely external and random nature" (italics added, p. 158). Again, questions come to mind: Is Leontiev suggesting that this use of the FL in inner speech, because it is external and not subjective in form and because it is accidental and not systematic, does not have any real impact on the process of generating the crucial internal, personal senses of an utterance? With greater mastery of the FL, could the use of FL words and structures at the level of inner programming become more pervasive and therefore more influential in generating the inner syntax and semantics of a verbal utterance? programming (planning)

motive

thought (speech intention)

inner programming

post-programming application of LI code

translation into the L2

motor implementation

external speech

Figure 3-1. A. A. Leontiev's schema of speech act stages in a foreign language. Leontiev seems to assign a more influential role to the FL than simply providing a linguistic code for ready-made ideas. In line with Vygotsky, Leontiev believes that, at the programming stage of speech production, language "fulfills" rather than expresses thought. Through words, thought attains fulfillment or completion rather than expression (p. 105). Thus, to Leontiev, speech in an LI or FL is "a creative intellectual activity" (p. 109) that completes a cognitive operation and contributes to its realization. Leontiev suggests that to deal with the "methodological aspect of the idea of thinking in a foreign language" (p. 109) it is necessary to go beyond the traditional psycholinguistic approach of only examining the coding (postprogramming) stage of a speech act and to delve more deeply into the relationship between thought and speech. In his book, Leontiev touches on several critical issues and triggers numerous questions related to the idea of "thinking in an FL." First is the notion of thinking "directly" in a language in order to speak it. Is it possible? Is it necessary? Or, does thinking in an FL always need "switching" mentally from one language to another? Another crucial issue is to what extent the nature of thought is affected or determined by language. Does learning an FL imply changing the structure of thoughts? Is there a "core" way of thinking that is unperturbed by the language of its realization? Finally, is it a "trivial" fact, as Leontiev believes, that in the realization stage of speech

62

CHAPTER 3

production various languages or linguistic codes may be used? What are the methodological and pedagogical implications of "coding" thoughts in an FL? Although highly theoretical in nature, Leontiev's essay is a rich source for speculation and reflection on inner speech and L2 learning. Learners' Preferences for a Language of Thought In his lengthy discussion of L2 learners' preferences for a "language of thought,"34 Cohen (1998) attempts to clarify the idea of thinking in a target language by resorting to Vygotsky's notion of inner speech. According to Cohen, when we think in a language, "our thoughts reflect inner speech-that is, the thinking that we do in our minds that is in the form of words rather than images or symbols" (p. 160). Unfortunately, Cohen does not pursue the implications of viewing the question of thinking in a target language from a Vygotskyan standpoint, and the connection to inner speech-as Vygotsky conceived it-is left unexplored. Cohen's (1998) insights into L2 learners' preferences for a language of thought and his collection of empirical and anecdotal evidence on thinking in a target language are useful, however, because they offer light on what Leontiev (1981) might call the "coding" stage of L2 inner speech, that is, the moment when thoughts are put in some definite foreign linguistic form. One of the most important suggestions made by Cohen is that the question of which language the learner uses to think depends on mastery of the specific discourse domain of the task being conducted. If the learner has greater control of a particular discourse domain in the L2 than in the LI, it is likely the learner will use the L2 to think through on that matter. Learners who have learned certain academic subjects in an L2 may naturally think about those subjects in the L2 rather than in the LI. Likewise, non-native speakers who have learned certain workrelated jargon in the L2 may have difficulty thinking about those concepts in the LI. On the other hand, a learner or non-native speaker may feel more comfortable to think in the LI about other domains in which LI control is greater. This may result, in Cohen's words, in a "diglossic" thinking situation, "where the speaker has the capability of thinking in two or more languages and uses these languages for distinctive and largely complementary purposes" (p. 162). Cohen (1998) presents data from a mini-survey he conducted and his own multilingual thinking experiences to show the various factors determining the language of thought (see also Cohen & Olshtain, 1993). Sometimes, he found, the choice of which language to use mentally is done in an unplanned, automatic way. For example, a trilingual native speaker of English reported: Sometimes when something triggers a memory of being abroad where I spoke an L2 (i.e. Guatemala, Poland, etc.), I think in the language I used at the time, especially if the memory involves conversations or encounters with native speakers in those places. (Cohen, 1998, p. 164)

34 Cohen's use of the term language of thought does not seem to imply adherence to any Fodorian notion of language of thought as an innate mentalese. Rather, Cohen seems to be using the term to refer to any natural language used as a medium for thinking.

THINKING WORDS IN A SECOND LANGUAGE

63

Other times, Cohen points out, learners deliberately think in the L2, either voluntarily or because the instructional method requires them to do so. He considers this a sort of enhanced "din in the head."35 The planned use of the language of thought is evident in the following comment from a quadrilingual English-Ll speaker: I do this all the time, for the purpose of practicing my other languages. I'll take an English thought, and ask myself, "How would I say this in Spanish, or Ukrainian?" Then, additionally, I might ask myself, "Which language seems best to express that idea, or that thought, or feeling the best?" (p. 169)

In relation to the planned choice of language of thought, Cohen (1998) brings up the issue of the metalinguistic use of the LI as an aid in learning the L2 and the limited use of the L2 as the language of thought that may occur in instructed language situations. To learn the formal rules of the L2, learners may actually think about them in their LI. In fact, as Cohen speculates, "it is possible that the learners do not think complex (e.g. metalinguistic) thoughts through the LT [target language] at all, but rather make passing reference to the LT in the form of fleeting or limited thoughts" (p. 167). How much a learner thinks metalinguistic thoughts in the L2, Cohen suggests, may depend on the learners' motivation to become submerged in the LT culture, on the teaching method, or even on the type of task. His own personal experience of learning Japanese through a grammar-translation rote-memory approach showed him how certain classroom methods and textbooks actually promote the metalinguistic use of the LI. Cohen (1998), like Leontiev discussed earlier, explores the popular assumption-and sometimes the instructional dictum-that "it is beneficial for foreign language learners to think as much as possible through the language that they are learning" (p. 169). Learners are discouraged to translate into their LI in the belief that thinking in the LI will be adverse to L2 acquisition. Cohen believes it is "unrealistic" (p. 170) to expect learners to put their LI aside, yet 50% of the respondents in his survey indicated that they were advised by their teachers to think in the LT and 82% reported that they made the effort to do so. The issue in this deliberate use of the L2 to think through is whether it equals "thinking in the LT." According to Lantolf (cited in Cohen), "when non-natives plan and rehearse what they want to say subvocally in an LT [target language]..., this does not really constitute thinking in the LT" (p. 173). To Lantolf, this activity looks more like thinking about the LT Cohen (1998) presents evidence from a study he conducted among children (aged 8 to 11) attending a Spanish immersion program. The purpose of the study was to investigate how the "external language environment" (instructional goals, classroom activities, and school discourse) affected the "internal language environment" of the students. The internal language environment referred to the children's inner language processes concurrent with the performance of numerical and mathematical cognitive tasks. Results of the study showed that the students were using the L2 (Spanish) to solve the problems to some extent, though not as much as it would have been expected

35

A phenomenon described by Krashen (1983), discussed later in this chapter.

64

CHAPTER 3

of learners who had been immersed in the L2 school environment, many of them for several years. Only 3 out of 15students reported mentally doing the problems totally in Spanish. The rest of the students tended to read the problems in Spanish, try to solve them in Spanish, and switch to English (the LI) when things got rough. On the basis of these results, Cohen observes that there is "an underground of [LI] use, out of earshot of the teacher, in pupil working groups, and, most importantly, in the pupils' minds" (p. 208). Cohen believes there is a great deal of "reprocessing" of L2 material into the LI in students' minds in language immersion programs. A recommendation he makes to conclude his discussion on the learners' preference for language of thought is for educators to start taking more seriously the role of the LI in FL learning and use. Additional evidence on L2 learners' preferences for a language of thought is provided by Larsen et al.'s (2002). The authors investigated the inner speech of a group of 20 Polish immigrants in Denmark, half of whom had migrated at an "early" (though adult) age and the other half at a "late" date, and all of whom had learned Danish (the L2) after arrival. A questionnaire revealed that the early immigrators had more inner speech behaviors in the L2 than the late immigrators. Furthermore, when asked to retrieve autobiographical memories and to indicate the language in which those memories "came to them," both groups reported more inner speech in Polish (the LI) for memories of events before immigration than in the L2 and more inner speech in Danish for memories after immigration. The results have interesting implications on the type of change that late bilinguals undergo at the level of conceptual and semantic representations. As the authors contend, the memories of bilinguals who acquire the L2 in their adult life seem to be cut along cultural and language lines. In other words, bilinguals seem to have different, co-existing sets of memory representations at the conceptual level, one set of memories for events occurring in their LI culture and linked to the semantic and lexical system of the LI and another set for events in the L2 culture linked to the semantic and lexical system of the L2. Finally, in this section, it is pertinent to cite evidence from Cook (1998). Cook focused on the functions of the language used for "internal purposes" by a group of bilingual students and teachers of mixed language backgrounds. In this study, seven internal or private functions of language were tested through of a questionnaire: (1) self-organizational uses, such as making appointments or shopping lists; (2) performing mental tasks involving arithmetic, such as counting or adding up; (3) remembering everyday information, such as days of the week, phone numbers, or historical dates; (4) unconscious uses, such as dreaming, singing, or talking to oneself aloud; (5) displaying emotions, such as feeling happy, sad, or tired; (6) non-communicative uses, such as talking to nonspeakers like small babies and animals; and (7) praying to oneself.36 Although there were some clear language preferences, for example, using the LI for praying and unconscious uses, and the L2 for self-organization, there was 36 Although some of the above functions (such as #6 and probably #5) seem to apply to private, audible speech, most of them could be carried out through silent, inner speech. Cook's results are therefore relevant to this discussion.

THINKING WORDS IN A SECOND LANGUAGE

65

great variability depending on the tasks involved and the language background of the participants. For example, within the mental tasks category, remembering historical dates had a high preference for the L1, whereas remembering phone numbers displayed a balanced use of either the LI or the L2. There were also differences depending on the particular languages spoken the participants. Francophone Africans reported an overall preference for the L2 for six of the internal functions and a preference for the LI for only one (the noncommunicative function). Finnish subjects, on the other hand, had a clear overall bias toward the LI. Cook raises an important question at the end of his article: whether public uses of the L2 are learned before private uses or whether both are learned simultaneously. He suggests looking at proficiency as a variable to investigate the development of private uses. Nature of Verbal Thought in the L2 LI as the Basis for L2 Inner Speech The interface between thought and language that occurs at the level of inner speech has also been explored from an L2 perspective in studies dealing with the nature of verbal thought in the L2. One study in particular (Ushakova, 1994) addresses the issue of inner speech structures in the L2.37 In this study, Ushakova elaborates her thesis that L2 acquisition is a process of incorporating the newly acquired L2 into the already established system of the LI. This process implies drawing upon the preformed semantic structures of inner speech developed in LI acquisition. Ushakova reports two experiments conducted by her research team to identify mechanisms of word acquisition in the L2. In both experiments, participants were trained to recognize the sound-meaning correspondence of twenty new words from an artificial language. Meaning of the new words was rendered in Russian, the participants' native language. One of the most salient results of the experiments was that the processes of memorizing and recognizing the new words depended heavily on the category to which the words belonged, nouns being learned faster and more consistently than verbs, for example. In addition, errors in meaning occurred only within words of the same semantic category, for instance, dog and animal. Association by sound of new words with words in the LI was another result of the experiments, indicating that the influence of LI structures is not only semantic but also phonological. The research group interpreted these findings as showing that words of anew language are classified

37 There is a certain ambiguity in Ushakova's (1994) conception of inner speech. Despite her frequent reference to inner speech processes, she defines inner speech as "the mechanism hidden from direct observation, comprisingfunctional structuresfor word storage, relations between words, semantic fields, grammatical rules, and rules for discourse production [italics added]" (p. 135). In her view, inner speech constitutes an organized "system" (p. 154) or "foundation" (p. 151) of conceptual structures, verbal networks, and grammatical structures (pp. 146-147, p. 154) stored in memory and developed in the course of LI acquisition. In other words, it is not clear whether, in Ushakova's view, inner speech constitutes the activation of those structures and networks (as she suggests on p. 144) or the structures and networks themselves. As discussed in Chapter 2, a conceptualization of inner speech as a set of structures stored in memory would run counter to the sociocultural notion of inner speech as activity and process.

66

CHAPTER 3

and grouped into the pre-established inner speech organization of the LI. We have come to the conclusion that the grouping of memorized words is subject to the semantic foundation realized in the system of the earlier acquired (first) language. Acquisition of the second (and other) language(s) occurs as a "plugging" of the new lexicon into the already established linguistic structures, which allows for categorization and linking of structures when necessary, and, most importantly, interpretation. This is the fundamental way in which the first language influences the second, (p. 151)

Ushakova's conclusions suggest that development of inner speech in the L2 is strongly dependent on the existing LI foundation. A major implication of this is that, once a native language is acquired, any subsequent language that is learned will inevitably be associated with the LI. These are quite formidable inferences. A few methodological problems with Ushakova's research, however, should make us cautious in the interpretation of her results. The first drawback is that, because the participants had to supply the newly acquired word at the sound of its Russian equivalent, the association with the LI was imposed at the outset. The strong dependency on the native Russian the researchers found may have been a function of the methodological setup of the experiment. Grabois (1996) has similarly exposed this shortcoming: "The nonce words were part of an artificial language, and were taught as the exact equivalents of Russian words. The learning of the words really represents the memorization of phonological forms within the Russian conceptual system" (p. 80). Another weakness in Ushakova's study is the high degree of artificiality that characterized the experiments. The new words were learned in a totally decontextualized way,38 and the only referent to the meaning of the new words the subjects had was via the words' Russian translations. No other associations to conceptual structures or world experiences-in Vygotsky' s terms, no other senses-werQ generated for the words. It is therefore not known what other types of connections the subjects could have made, or whether the critical LI association would have resulted, had the words been learned in another way. For example, had the L2 words been learned in the context of the L2, would they have been associated with the LI or with the L2? Furthermore, the experiments involved only the learning of a few words at what may be called a very initial stage of word acquisition. It is not known whether the same results would hold with longer and more complex exposure and use in the L2 and with a more massive load of lexical items. An open question therefore is whether the internal organization of an L2 is static or subject to change, and whether it would always respond to the architecture laid out by the LI. Commenting on this aspect of Ushakova's findings, Lantolf (1999) points out: "Under certain circumstances (i.e., cultural immersion), adults are capable of appropriating new cultural models and in so doing modifying their conceptual organization" (p. 37). The major problem with Ushakova's research, in short, is that her claims seem overly ambitious on the basis of the data and the methodology used. Her conclusions are not implausible, but they should be taken as hypotheses to be 38

That the words were "artificial" is not a problem since they could have been new sound-meaning combinations in any natural language.

THINKING WORDS IN A SECOND LANGUAGE

67

tested rather than as definitive findings. Unifying the LI and the L2 in Verbal Thought Like Ushakova, John-Steiner (1985b) is concerned with the nature of verbal thought in the L2 and the role of the LI in L2 development.39 Unlike Ushakova, however, John-Steiner does not believe the process of L2 acquisition is one of simply plugging in new labels to LI concepts. The lead question in John-Steiner's work is how two languages, the native one and a new one, become united at the level of verbal meaning and thought. Her data consisted of self-reports gathered through interviews with adult L2 and FL learners. The interviews centered on a variety of questions concerning learning strategies, affective issues, and the role of verbal thinking in dual language development. Her main conclusion was that "the relationship of language to thought is not one of static connections; it changes with the shifting lines of development of the two languages" (p. 357). In other words, there is no single or fixed thought-language relationship for dual language learners; there are in fact several stages in this relationship. The first stage, according to John-Steiner, is "leaning on the known" (p. 359). This is a stage of reliance on the LI in which the learners tend to obtain meaning from L2 utterances by translating into the LI. The native language provides the internal conceptual structures that help learners understand new words. The second stage consists of "an uneasy alliance of two languages" (p. 363). At this stage, learners make great efforts not to translate in order to break their dependence on the LI. The third stage is "toward unity of thought and diversity of expressions" (p. 366). This stage is achieved by fully developed and balanced bilinguals, such as interpreters. John-Steiner believes "a largely unified meaning system" emerges at this stage. This "internal system supports the varied and flexible uses of multiple codes among bilinguals" (p. 366). Interpreters describe the process of translating as one of going beyond the individual meanings of words, phrases and sentences and reaching at the ideas underlying the words. While discussing these stages, John-Steiner makes the following comment about "thinking in a second language": The task of internalizing a second language and weaving it together with the existing fabric of verbal thought is a complex one. The relationship of language and thought changes with the shifting lines of development of the two languages. In the early stages of secondlanguage acquisition, there is a strong dependence upon the native language as the primary processor for both comprehension and production. But the assumption that competent bilinguals "think in English" while they speak English and "think in Spanish" when they speak Spanish is based on a simplified notion of thought. The shift in internal processes referred to in these comments is one of subvocal process of planning an utterance; it is the increasing ability to plan in one's weaker language that corresponds to the notion of "thinking" in it. (p. 365)

39 Elsewhere, John-Steiner makes clear the connection between inner speech and verbal thought: "In inner speech, thought and language intersect and influence each other dynamically" (John-Steiner et al.,1994, p. 16).

68

CHAPTER 3

What John-Steiner is saying is that what we commonly call "thinking in a second language" is actually just the planning stage of producing an utterance, a process that is subvocal. This is, according to her, a simplified notion of thought. To her, verbal thought is a much more complex issue than just planning in one's L2. John-Steiner's reasoning echoes that of A. A. Leontiev (1981), who also believes that the popular idea of "thinking in a language" is limited in that it merely focuses on the precommunicative stage of coding thoughts in linguistic form. On the relationship between language and thought among dual language learners, John-Steiner concludes that, as learners develop competency at the production level, "the two languages are increasingly separated into autonomous systems of sound and structure, while at the level of verbal meaning and thought the two languages are increasingly unified" (p. 368). This finding supports Vygotsky's claim that "the processes of the native and foreign language have between them a great deal in common . . . they are internally united" (cited in John-Steiner, 1985b, p. 349). JohnSteiner illustrates her hypothesis about the unification of languages at the level of verbal meaning and thought with some developmental data showing the note-taking behaviors of adults studying in their non-dominant languages. At first, the learners tried to translate and write their notes in the LI as they heard the L2, or they used a mixture of both languages. Their next strategy was to take notes in the language of the speaker, as rapidly as possible, even if they didn't understand everything that was said. Finally, they were able to listen carefully and write down in the L2 only the core concepts of what they heard. "Thus, as learners are increasingly able to comprehend, condense, and store information in their weaker language, they start the process of weaving two meaning systems together" (p. 365). In the process of integrating the two languages, it is also frequent for individuals to experience difficulties in production in either language. They seem to undergo "temporary losses" and "inhibition" (p. 366) in their language behavior.40 Eventually the learners overcome their verbal losses, inhibitions, and fears, and the consolidation of the two languages as a unified conceptual system takes place. One learner comment reported by John-Steiner exemplifies this level of development: "It is interesting, now that I speak some English as well as Hungarian, I am no longer sure which language I am thinking in" (p. 366). Self-Reconstruction and Reconceptualization through the L2 The process of developing two (or more) languages in an individual's mind is complicated, as John-Steiner suggests. Pavlenko and Lantolf (2000) and Pavlenko (1998) add another angle to the issue of development by focusing on the role of the LI and the L2 in self (re)construction. In their research, Pavlenko and Lantolf highlight the role of inner speech, as a point of convergence of thought and language, in the formation of a personal identity. For their studies, the researchers analyzed self-

40

These experiences are very similar to those described by Pavlenko and Lantolf (2000). See the ensuing discussion of their work.

THINKING WORDS IN A SECOND LANGUAGE

69

narratives of late bilinguals in which the writers depicted their L2 acquisition experiences. Unlike other L2 learners, who do not achieve full mastery of the L2, the writers of these narratives, who started learning the L2 in their adulthood or adolescence, were able to achieve "ultimate attainment" (p. 169) of their L2. In this process of developing very advanced levels of attainment in the L2, these bilingual/bicultural individuals (all writers, mostly of East European origin, learning a prestigious language, such as English or French) underwent a multilayered and conflicted process of self-reconstruction, largely characterized by loss of an LI identity and creation of a new L2 one. For these individuals, learning a second language involved first a series of losses related to the native language. This phase is characterized by five stages: (1) loss of one's linguistic identity, (2) loss of all subjectivities, (3) loss of the frame of reference and the link between the signifier and the signified, (4) loss of the inner voice, and (5) first language attrition. Some of the changes experienced during this phase are poignantly narrated by the writer Eva Hoffman (1989): I wait for that spontaneous flow of inner language which used to be my nighttime talk with myself. . . . Nothing comes. Polish, in a short time, has atrophied, shriveled from sheer uselessness. Its words don't apply to my new experiences, they're not coeval with any of the objects, or faces, or the very air I breathe in the daytime. In English, words have not penetrated to those layers of my psyche from which a private conversation could proceed, (p. 107)

From the above excerpt, Pavlenko and Lantolf deduce that the narrator has suffered a loss of her LI inner speech-the rich inner voice in one's own native language that draws from dense personal meaning and that one uses to organize and make sense of the world-and has no language to make sense of her new experiences: In the above passage, Hoffman seems to be in a semantic twilight zone in which her inner speech in Polish has ceased to function, while the inner speech sparked by English, her new language, has yet to emerge. From a sociohistorical perspective, then, she has no way of organizing and making sense of her experiences. In some sense, she has no experiences, because, as both Vygotsky and Bakhtin agree, it is through inner speech that we create our experiences; that is, in inner speech we organize and integrate the events that occur in space and time into the plot of our life narrative. Without inner speech, this organization and integration are impossible. (Pavlenko & Lantolf, 2000, p. 165)

Another passage by Hoffman illustrates one of Vygotsky's major points about the semantic nature of inner speech, that is, its reliance on (personal) sense rather than (public) meaning. The words I learn now don't stand for things in the same unquestioned way they did in my native tongue. "River" in Polish was a vital sound, energized with the essence of riverhood, of my rivers, of my being immersed in rivers. "River" in English is cold-a word without an aura. It has no accumulated associations for me, and it does not give off the radiating haze of connotation. It does not evoke. (Hoffman, 1989, p. 106)

In this case, Hoffman is quite transparently alluding to what Vygotsky calls sense in inner speech. The senses of the word "river" are the "things" the word stands for in her native Polish. On the contrary, words in her L2, English, do not have senses; that is, they are empty labels with no referents. Her inner speech has "no accumulated

70

CHAPTER 3

associations" for English words, no "aura." Her English does not "evoke" senses. The last stage of the loss phase in L2 learning is LI attrition. Learners at this stage have difficulty remembering words in their LI, as the new language takes over as the means to name things and experiences in the new environment. The process is often accompanied by strong feelings of guilt, embarrassment, and failure, as John-Steiner (1985b) had observed. The following excerpts from writers Jan Novak and Kyoko Mori describe these feelings: My Czech had begun to deteriorate. There were times now when I could not recall an everyday word, such as "carrot," "filer," or "sloth." I would waste the day probing the labyrinthine recesses of my memory because to get help from the dictionary seemed only to legitimize the loss. (Novak, 1994, p. 263) Trying to speak Japanese in Japan, I'm still thinking in English . . . . Flustered, I try to work out a quick translation, but my feelings are untranslatable and my voice is the voice of a foreigner. (Mori, 1997, p. 16-17)

As Pavlenko and Lantolf (2000) suggest, the second phase of these late bilinguals' linguistic journey is the recovery and reconstruction of their selves. This phase consists of four states: (1) appropriation of others' voices, (2) emergence of one's own new voice, (3) translation therapy or reconstruction of one's past, and (4) continuous growth into new positions and subjectivities. One of the first symptoms of this phase is the "appropriation" of other people's voices in the L2, a phenomenon called "ventriloquation" by Bakhtin (Wertsch, 1991). Through the new language, these individuals construct a new "me." They reinvent themselves with a new voice, but in order not to lose their old self they need to reconstruct it (sometimes re-write it) in the new language. Pavlenko and Lantolf (2000) call this stage "translating oneself into the L2, a necessary step to unite the two selves together. As this occurs, the individual finally finds him/herself more and more comfortable in his/her new self and develops full identity with the new language. Pavlenko and Lantolf s research offers a lot of insights-and prompts a lot more questions-into the interrelatedness of inner speech and the "self." Is language identity defined by the nature of one's inner speech? How does inner speech develop with more L2 acquisition and participation? Certainly, the study of late bilinguals' development brings to surface the issues of change, shift, incrementation, attrition, and levels of inner speech. What is the inner speech of a bilingual like? In the case of these late bilinguals, who have suffered considerable loss of their LI, can we say they continue to be "bi-linguals," or have they become more like "mono-linguals of an L2"? And what does that imply for their inner speech? Obviously, the study throws into serious question the notion of "native speaker." After having adopted a new L2 identity and having been "re-born," so to speak, into the new L2 culture, are not these late bilinguals "native" speakers of the L2 to some extent? Of course, as Pavlenko and Lantolf also explain, there are levels of bilinguality, and not all learners achieve the ultimate attainment of the late bilinguals depicted in the study. As the authors cogently argue, attaining full competence and building a new L2 identity is to a great extent a matter of personal choice. Learner agency is an important factor in the acquisition of the various levels of the L2 language and culture. Agency and choice, it will be

THINKING WORDS IN A SECOND LANGUAGE

71

argued, are also factors that determine the nature of an L2 learner's inner speech. The consequences of becoming fully bilingual go beyond the issues of loss and reconstruction of the self. Mental reorganization on the basis of conceptual change may also result. As Pavlenko and Lantolf (2000) suggest, "participation can also lead to changes in what is more traditionally dealt with under the heading of cognition and cognitive mediation" (p. 177). Gestures and Conceptual Change in L2 Inner Speech Additional insights into the nature of inner speech and the thought-language relationship in L2 learning are offered by McCafferty and Ahmed (2000), who explored the role of nonverbal forms of expression, such as gestures, gaze, and body movements, in L2 learning. Working within a Vygotskyan framework and following McNeill (1987), McCafferty and Ahmed contend that thought, language, and gesture are intertwined at the level of inner speech. Evidence of this is that often gestures reveal the psychological predicate of what the person is thinking, as when a person searching for words completes a sentence with a gesture. Gestures can also express thoughts in ways that are not possible through speech, thus reflecting inner speech better than words themselves. In their study, McCafferty and Ahmed (2000) pursued the question of appropriation of L2 gestures. In particular, they were interested to see whether Japanese-Ll learners of English of two different types (naturalistic and classroom-instructed) appropriated abstract gestures of the L2 culture (the U.S.), such as metaphorics (pictorial gestures that represent ideas rather than objects, such as the splitting-the-space hand movement to indicate two contrasting ideas) and beats (hand or foot movements that signal important discourse-pragmatic content). These types of gestures were selected because they are associated with cultural expression and could throw light on the issue of intrapersonal change with exposure to a different L2 culture. The researchers speculated, If gestures of this type are appropriated by L2 learners, given the interconnections discussed above, it would seem possible that aspects of inner speech change as well, and moreover, that the concepts imaged by gesture in connection with inner speech change as part of a process of remediation41 involving use of the second language within the settings of the culture, (p. 200)

The results showed that typically American gestures of the abstract were indeed appropriated by the naturalistic learners, but not by the classroom-instructed learners, despite their high proficiency in the L2. These findings indicate that when learners are immersed in the L2 culture, it is possible for them to change, or infuse with new sense, the gestural substratum of their inner speech. The question of conceptual change as related to L2 learning is a recurrent one, as

41 The word "remediation" is used in the sense of "re-mediation," that is, as an alternate or different form of mediation that is made possible through the newly appropriated L2 gestures, and not in the sense of "providing a remedy." (Personal communication, Steve McCafferty, June 25, 2002)

72

CHAPTER 3

we have seen in the writings of Ushakova (1994), John-Steiner (1985b), Lantolf (1999), Pavlenko and Lantolf (2000), and McCafferty and Ahmed (2000). Obviously, the question is too complex to be treated in all fairness here, but when one considers the great diversity of language learners and learning situations that exist, it is clear that outcomes other than those obtained in situations of "subordinate bilingualism"42 are possible. As Pavlenko (1999) explains, while co-existence (without interaction) of two (or more) separate conceptual representations is possible (as in the case of bilinguals who keep their languages apart by using them in different contexts), multiple interactions between conceptual stores resulting in conceptual change may take place. Pavlenko's research (1996, 1999) showed that changes in the bilingual conceptual store may involve any of the following processes: (a) internalization of new concepts, (b) conceptual shift from LI to L2, (c) convergence or the unification of LI and L2 conceptual domains, (d) restructuring or modification of particular conceptual domains, and (e) attrition of certain concepts learned through the LI. To sum up, a review of the literature on verbal thought in the L2 throws light on various aspects of the inner speech-L2 connection. Some of these are the idea of thinking in another language, the question of students' preferences for a language for thought, and the role of the LI in the verbal thought of L2 learners. A recurrent concern among several of the studies reviewed is the issue of change at a semantic and conceptual level and how this may affect the thought-language relationship of L2 inner speech. INNER SPEECH AS INTERNALIZATION OF THE L2 In Vygotskyan sociocultural theory inner speech represents the culmination of the process of internalizing social speech. In this view, language acquisition is essentially a process of privatization of language, from the interpsychological (social) space to the intrapsychological (individual) sphere. A transitional phase in this process that has been observed among children learning their L1 is private, audible, self-directed speech {egocentric speech, in Vygotsky's terms). Several researchers have documented the occurrence of private speech among L2 learners, both as a symptom of externalization of their inner speech (see, for example, McCafferty, 1994a, 1994b; Frawley & Lantolf, 1985) and as evidence of L2 internalization processes (Centeno-Cortes, 2003; Lantolf, 2003; Lantolf & Yanez, 2003; Ohta, 2001; Saville-Troike, 1988). What is not clear yet is whether all learners engage in some form of audible private speech in their acquisition process. In other words, it has not been established that overt private speech is a necessary phase preceding inner speech in the internalization of an L2. Despite their focus on vocalized, rather than covert, self-directed speech, private speech studies offer important insights into how an L2 is internalized and how it may eventually result in inner speech. Various private speech processes were common to Centeno-Cortes (2003), Lantolf and Yanez's (2003), Saville-Troike's (1988), and 42

This occurs when the L2 is learned as an FL and L2 words become attached to conceptual representations through LI words (Pavlenko, 1999, p. 110).

THINKING WORDS IN A SECOND LANGUAGE

73

Ohta's (2001) investigations: repetition or imitation (verbatim or with variants) of the teacher and other students' utterances, vicarious response (answering a question or providing a response to a teacher prompt addressed to another student), production of new linguistic forms, rehearsal (practicing what to say), and experimentation (manipulating and playing with language forms). Two additional functions of private speech were also evident in Lantolf and Yanez (2003): the heuristic use of the LI to make sense of the L2 and the use of LI metalanguage to understand the L2. Several important implications can be derived from the "private speech as internalization" studies: First, learners are not as passive as they may seem in class but are actively involved processing the language at an intrapersonal level; second, their focus of attention may not be the same as the one the teacher wants to emphasize-in other words, focus is a subjective, self-directed, experience in private speech (Frawley, 1997); and third, as students appropriate the L2 through imitation and repetition, they not only reproduce in their minds the social artifact that is the L2 but also transform it into an individual cognitive tool. (See Chapter 6 for further discussion of the implications drawn from studies dealing with internalization through private speech.) Studies of L2 private speech are valuable because they provide clues on the internalization process that culminates in inner speech. Yet, they remain studies of an externalized form of self-directed speech. If, as Vygotsky postulated, external speech suffers a transformation in form and function as it internalized into inner speech, can private speech-as a transitional phase-be assumed to be identical with inner speech? For example, if a progressive abbreviation of external speech is posed, should we expect inner speech to be even more abbreviated than private speech? How does L2 external speech become "inner" L2 speech? What is "inner" L2 speech like? And how do L2 learners develop an "inner" voice in the L2? Although private speech among L2 learners may offer important insights into the nature of L2 inner speech, the latter merits attention in its own right. A study that strictly pursued the process of internalization and the development of inner speech among L2 learners of English was conducted by Guerrero (2004). (Because this study will be discussed in detail in Chapter 5, only a brief summary is offered here as part of the general review of research on inner speech and L2 learning.) The purpose of Guerrero's study was to explore the early stages of L2 inner speech; in other words, her interest was to document what might be taken as the initial phase in the transformation of social or private L2 speech into inner speech. Guerrero collected verbal reports from 16 pre-basic ESL college students. Reports came in the form of diaries where the participants recorded their notes on the inner speech they experienced during class and outside the classroom. Just as in Lantolf and Yanez's (2003), Saville-Troike's (1988), and Ohta's (2001) research on private speech, repetition and imitation of the L2 were found to be prominent covert behaviors. Subvocal or silent repetition of language being heard or read was the most frequent reported type of internal L2 use whereas mental verbalization of private thoughts was the least common. As suggested by their verbal reports, these beginning students seemed to be very busy turning external language into inner language and transforming English into a tool for verbal thought.

74

CHAPTER 3

INNER SPEECH IN L2 READING AND WRITING The link between inner speech and L2 learning has also been pursued in a small group of studies on L2 reading and writing (Sokolov, 1972; Upton & Lee-Thompson, 2001; Huh, 2002). The first of these studies to be reviewed is the research reported by Sokolov in his book Inner Speech and Thought (1972). Although the book is about inner speech in general and does not have an L2 or FL learning perspective, some of the experiments described in it involved the resolution of mental tasks by FL learners and are therefore relevant to a discussion of inner speech in the L2. Sokolov and his colleagues used FL texts in many of their experiments. One set of experiments was designed to document the formation of verbal semantic complexes and verbal generalizations in reading (Sokolov, 1972, pp. 78-88). To this end, subjects with varying degrees of knowledge of English as an FL were asked to "think aloud" as they translated English texts into Russian. The texts themselves varied in terms of difficulty. As the subjects read and translated the English texts into Russian with the aid of dictionaries, it was evident that, regardless of the readers' level of knowledge of English and of the text level of difficulty, the subjects' main strategy was to single out key words or groups of words, find out their meaning (via dictionary or guess), and arrive at a more or less tentative hypothesis about the general meaning of the phrase. For the subjects, these key words acted as "reference points, landmarks in the process of understanding" (p. 80). On the basis of these reference points, the subjects were able to form separate semantic groups and make generalizations about the text. An important finding was the grammatical structure of these semantic groups. It was found that sequences of nouns and verbs served as the main carriers of meaning. In general, all subjects formed their primary general notion of the passage by establishing the meaning of these noun-verb semantic landmarks. In some cases, "these generalized semantic verbal associations . . . evoked various graphic images which, too, became carriers of the general meaning" (p. 87). Thus, Sokolov hypothesized that the principal structural element of inner speech consists of verbal semantic complexes of a generalized and abbreviated nature, possibly substituted at times by graphic symbols. On the basis of these paradigmatic experiments involving translation of FL texts, Sokolov was then able to demonstrate the internal processes of generalization and reduction of speech that occur in the realization of complex mental operations. An ancillary observation during the above experiments was that beginning students of English evidenced their difficulty in translating the foreign text by reading the text aloud, whispering, or moving the lips as they read. Sokolov's hypothesis in this respect was that "for people who are in the initial stage of mastering a foreign language, inhibition of articulation should represent a substantial obstacle to their translating of foreign texts" (p. 148). A study conducted by Babalova (as reported by Sokolov, p. 148) tested this hypothesis. The purpose of the study was to observe the effects of inhibited articulation and motor speech interference on thinking. As in the previous experiments, FL learners were used as subjects. In this case, two groups of

THINKING WORDS IN A SECOND LANGUAGE

75

students of English, in their first and fifth year of college respectively, participated. The fifth-year students were asked to translate English texts at two levels of difficulty: adapted (simplified) and non-adapted pieces of fiction. The first-year students translated only the simplified texts. All subjects were asked to read (and presumably mentally translate) the texts under varying conditions of speech articulation and then retell the contents in Russian. There were three conditions of articulation during reading: (1) mechanical retardation of articulation (such as clamping the tongue between the teeth), (2) concurrent articulation of extraneous syllables (such as "la, la, la"), and (3) concurrent recitation of a poem. A fourth condition used as control was free articulation. Outcomes were compared in terms of (a) amount of time taken in reading (and translating) the text and (b) number of semantic units translated correctly. Results were complicated, but some definite patterns arose. In the translation of simplified texts, mechanical retardation of articulation and enunciation of syllables had almost no negative effects for any of the subjects, whether beginners or advanced. Adverse effects were noticed, however, in the poem recital condition, most impressively amongst the first-year group. In the translation of the more difficult texts, the fifth-year students evidenced negative effects in only the recital condition. It appeared then that retardation or interference with speech articulation affected the most when it was maximally instantiated (as when subjects were concurrently reciting a poem) and when verbal mastery of the FL ("verbal connections," in Sokolov's words, p. 150) was least strong. How did Sokolov explain these findings? When texts are difficult, as in the case of the first-year students and of the fifth-year students when reading complex texts, suppressing the possibility of inwardly articulating words causes serious troubles in comprehending, translating, and remembering. When texts offer no major difficulties, either because of their inherent simplicity or because of the reader's high level of competence, no extensive inner articulation of speech is necessary. Advanced students in fact indicated experiencing at times immediate comprehension of texts, even when articulation was obstructed. As they reported, these students were able to "grasp the text with their eyes and enunciate the words inwardly" and to "read easy texts without translating them" (p. 151). With less simplified texts, however, Sokolov observed, "verbal interference rendered the translation more difficult by hindering the articulation of the words of the text" (p. 152). Sokolov's experiments highlight the role of inner speech in vital mental processes involved in using an FL, such as reading, translating, comprehending, and retelling. The evidence suggests that inner speech is instrumental in aiding understanding and memorization through the formation of semantic complexes and verbal generalizations. The extent and nature of inner speech in processes of FL comprehension, however, is relative to the subject's mastery (degree of automatization) over these processes. The greater the mastery, the less need for unfolded and intense inner speech activity. In this case, inner speech is reduced to the auditory or visual recognition of words in the stimuli and to incomplete articulatory reproduction of key words. The more active the efforts in understanding the verbal data, however, the greater the reliance on complete articulatory reproduction of the speech heard or read. Among the conclusions derived

76

CHAPTER 3

from these foreign text reading experiments, Sokolov points out "the extreme importance which the fixation and reproduction of words and their grammatical connections within sentences by means of articulation has for the study of foreign languages" (p. 263). The relative, and to some extent paradoxical, role of inner speech in FL understanding is thus explained: The student has to learn to understand oral and written speech quickly and accurately, with a minimal participation of articulatory movements (the only factor capable of ensuring a rapid rate in understanding utterances or a text being read). But the student can only achieve this by means of a maximally unfolded articulatory fixation of speech in the course of preceding practical experience in the spoken language. This may seem contradictory, yet there is no other solution for this problem, (p. 263)

Another study that made connections between inner speech and L2 learning in the context of reading was that by Upton and Lee-Thompson (2001). Arguing that mental translation is related to Vygotsky's notion of inner speech, the authors observed the role of the LI in reading L2 texts. Twenty ESL college students (10 Chinese and 10 Japanese native speakers) representing three proficiency levels participated in the study. As the students read a passage in English, they were encouraged to verbalize their thoughts in the language (LI or L2) they were thinking those thoughts. The researchers found that as proficiency increased there was a decline in use of the LI as the language of thought for processing the L2 text. When the LI was used, it was found that it mainly served the purpose of wrestling with or confirming meaning. Upton and Lee-Thompson argue for the need to go beyond the view of the role of the LI as a mere translating (decoding) strategy in L2 reading. They propose instead a Vygotskyan sociocultural perspective, "which sees inner speech as the foundation of thought [and] suggests that the LI would quite naturally serve as a tool to help students think about and make sense of (i.e., mediate their thinking about) the structures, content, and meaning of the L2 texts they read" (p. 491). The notion of mental translation has also been challenged in study by Huh (2002) that focused on inner speech in L2 writing. In the study, framed within Vygotskyan sociocultural theory, Huh found that the subjects-2 Korean-Ll advanced ESL learners-did not compose their texts by writing first in Korean and then translating into English. Rather, the learners used LI (Korean) inner speech to generate ideas and then tried to formulate them into English sentences. Through the stimulated recall technique, Huh discovered that much of the planning stage of composing for the two subjects consisted of turning thoughts into words through the use of LI inner speech. The writers, however, did not fully develop a text in Korean that would then be translated word for word into English. Their inner speech consisted of fragmentary and elliptical forms, images condensed into words, keywords saturated with sense, and the verbalization of self and other inner voices, all in Korean. From these sketchy verbal representations rendered in LI inner speech, the subjects went on to identify the linguistic structures of the L2. One of the subjects explained this process: "I could only briefly sketch the diverse images and associations . . . in Korean. Then, I thought how to communicate them in English" (p. 6). The process of turning LI inner speech into fully developed English text was not a smooth one, however. It consisted of

THINKING WORDS IN A SECOND LANGUAGE

77

"successive approximations of L2 texts through expansion, rephrasing, and editing [of the subject's] LI inner speech" (p. 9). On the basis of these data, the researcher thus contended that LI inner speech is a critical mediational tool in the process of writing in the L2, enabling writers to construct and convey meaning to their audience. In short, the few studies reviewed in this section suggest that inner speech plays a significant role in reading and writing L2 texts. It was shown that inner speech aids reading comprehension through the formation of abbreviated semantic groups and generalizations and the inward articulation of text read. Again, as in the studies on the verbal nature of thought, the LI was found to be implicated in the inner speech of L2 learners, both in reading and writing. Through these studies, mental translation emerges as an extremely complex process in which inner speech has a critical meaning making role. INNER SPEECH AND MENTAL REHEARSAL OF THE L2 Further insights into L2 inner speech are provided by studies dealing with mental rehearsal of an L2, a language learning strategy that has been defined as "the covert practice of the L2" (Guerrero, 1994, p. 84). Mental rehearsal appears to be a major learning strategy among L2 students. It has been associated with repetition (Chamot, 1987; O'Malley, Chamot, Stewner-Manzanares, Russo, & Ktipper, 1985), practice (Rubin, 1987), advance preparation (Chamot, 1987), and production (Tarone, 1983) of the L2. O'Malley et al. (1985) found that within eleven cognitive strategies reported by beginning and intermediate ESL students, repetition-a strategy that involves silent rehearsal-had the highest percentage of use. Of course, verbal rehearsal in short-term memory has been for years associated with long-term recall (Bransford, 1979; Craik & Lockhart, 1972; Houston, 1986). More recently, research within the working-memory model (as discussed in Chapter 2) has stressed the importance of rehearsal in long-term retention of FL vocabulary among adults (Baddeley et al., 1998; Gathercole & Thorn, 1998). Several studies (Ellis & Beaton, 1993; Ellis & Sinclair, 1996; Papagno, Valentine, & Baddeley, 1991; Service, 1992; Service & Kohonen, 1995) have found that audible and subvocal repetition have positive effects on long-term retention of words43 and that suppressing rehearsal produces negative effects. It has also been suggested that gifted language learners, or people that appear to have a natural talent for learning FLs, have superior phonological loop skills, that is, excellent use of the mechanism involving articulatory rehearsal (inner voice) and phonological memory (inner ear) (Baddeley et al., 1998). Evidence for this link between language ability and phonological memory was presented in Papagno and Vallar's (1995) study of "polyglots." References connecting mental rehearsal and inner speech abound. One of the earliest is an observation made by Vygotsky (1986) himself. In his comment that inner speech "serves as preparation for external speech-for instance, in thinking over a

43

Effects may even extend to grammar accuracy and fluency (see Ellis & Sinclair, 1996).

78

CHAPTER 3

lecture to be given" (p. 88), Vygotsky was assigning a rehearsal role to inner speech. In 1983, Smith equated rehearsal with the usual practice of "talking to oneself," a form of inner speech "by which we prepare for something we might want to say" or by which "we recapitulate and elaborate upon conversations after the event" (p. 90). In 1987, Rohrer argued that inner speech is "the language of the mind" (p. 92), used in various mental operations, one of which is rehearsal. In 1990, Murphey reviewed studies conducted on the "din in the head," linking this phenomenon to Vygotsky's concept of inner speech: "What [Vygotsky] calls inner speech may have a strong connection to what is now being called the Din" (Murphey, 1990, p. 55). In all these references, it appears that mental rehearsal, both in the LI and the L2, is tightly linked to inner speech. Mental Rehearsal as a "Din " Mental rehearsal gained distinction as an SLA phenomenon when Krashen (1983) called attention to the "Din," which he defined as "an involuntary rehearsal of second language words, sounds, and phrases" (p. 41). Krashen had read about this phenomenon in an article by Barber (1980), a linguist and FL learner of Russian, French, and German. Barber reported experiencing a "din" as she was traveling in Russia: By the third day . . . the linguist in me was noticing a rising din [italics added] of Russian in my head: words, sounds, intonations, phrases, all swimming about in the voices of the people I talked with. . . . The sounds in my head became so intense after five days that I found myself mindlessly chewing on them, like so much linguistic cud, to the rhythm of my own footsteps as I walked the streets and museums. Whenever I noticed this din, the linguist in me would demand to know what I was saying. Half the time I had to look what I was saying up, or somehow reconstruct what it meant from the context in which I had heard it hours or days earlier. The constant rehearsal [italics added] of these phrases of course was making it easier and easier to speak quickly and fluently; things popped out as prefabricated chunks. But I had no control over what my subconscious fed into my "chewer" each day. It fed me what it considered to be memorable-usually from a surprising or stressful or isolated incident-not what I considered maximally useful. Nonetheless, my overall command of Russian improved more in a single week than it would have in a month or two of intensive reading, (p. 30)

Krashen (1983) recognized the Din as an experience he himself had undergone at a conference after listening to several hours of presentations in German, an FL to him. After the conference, "on the plane, walking to the hotel, [Krashen] felt the Din rattling in [his] brain, exactly as Barber described it" (p. 42). On the basis of these anecdotal reports and in harmony with his SLA theory, Krashen hypothesized that "the Din is a result of stimulation of the Language Acquisition Device" (p. 43), that it is triggered by comprehensible input of the i +1 variety, and that it will not occur in very advanced learners "since they will receive less input containing i +1, having acquired most of the language" (p. 43). Perhaps also in keeping with his usual dismissal of monitoring as a successful acquisition activity, Krashen did not contemplate the possibility of rehearsal occurring voluntarily, that is, as a result of the learner's own desire to control

THINKING WORDS IN A SECOND LANGUAGE

79

input or production. This possibility would later be found to be real by Guerrero (1987). Bedford (1985) was the first to test Krashen's Din Hypothesis empirically. In the questionnaire Bedford constructed for his survey study, he described the phenomenon as follows: You have spontaneous rehearsal if you sometimes "hear" a clearly noticeable din or jumble of Spanish words, sounds, phrases, or even characteristic melody patterns in your head. (This is very normal!) These words and phrases are usually things you have been hearing recently (in class or on Spanish tapes, etc.). Often you "hear" the words or phrases in the voice qualities of your teacher or of the people who made the language lab tapes, or maybe even in your own voice. These random "snatches" of Spanish just pop into the head at nearly any time or place, and it's usually quite involuntarily. At times it may be active enough to be described as a "constant rehearsal in the head." (p. 286)

On the basis of responses by 160 L2 college and FL adult learners, Bedford was able to confirm that the Din was a widespread phenomenon rather than restricted to a few individuals. He found no difference by amount of previous study, thus being unable to support Krashen's prediction that the Din would disappear with more proficiency. Bedford clarified, however, that none of the subjects in his sample could be described as a "very advanced acquirer" (p. 283). In 1987, Guerrero replicated Bedford's study with a sample of 52 ESL college students at three levels of proficiency. Again, the Din was confirmed to be a wellknown phenomenon for the language learners, 79% admitting to having experienced it. Guerrero (1987) found no difference in frequency of Din activity among the three levels, although there was a slight (nonsignificant) increase with proficiency. Thus, Guerrero concluded that mental rehearsal could occur at any moment during acquisition and that, contrary to Krashen's prediction, even very advanced learners mentally rehearse. Furthermore, Guerrero observed what Krashen had not mentioned, namely, that, in addition to experiencing involuntary playback of the L2, the students also engaged in voluntary rehearsal. During this deliberate, purposeful rehearsal, the students were consciously retrieving information from memory in order to prepare for future production. This preparation included "organizing the material to be utilized, checking correctness, locating words in memory, refining meaning, etc." (p. 544). In 1986, Parr and Krashen published the results of two studies testing Krashen's Din Hypothesis. The first study tested the prediction that the Din is a widespread phenomenon. The data obtained from 150 high school students of Spanish and 216 college students of Spanish confirmed the prediction: 78% of the high school students and 69% of the college students answered affirmatively the question "Have you experienced the 'Din in the Head'?" after reading Barber's (1980) description of the phenomenon. The second study, however, supported the claim that advanced performers do not experience involuntary rehearsal. In this study, the data came from a group of 28 "advanced graduate students and faculty in foreign language education who had acquired their second language as adults" (p. 276). Only 10% percent (3 subjects) of these speakers answered Yes to whether they had experienced the Din. There are a few problems with these data, however. Although the authors claim these

80

CHAPTER 3

were "advanced performers of the second language" (p. 276), the proficiency level of the participants in either study was not measured in any systematic way. Moreover, the "advanced performers" were interviewed orally rather than surveyed through a questionnaire, as was done in the first study. Krashen explained the discrepancy between these results and Bedford's (1985) and Guerrero's (1987) saying that maybe the subjects in his study "were even more advanced, professors and teachers of the language" (personal communication, March 14, 1988). In summation, studies on the Din suggest that it is a common and widespread occurrence among L2 and FL learners, both children and adults, and that it can arise spontaneously or be triggered deliberately by the learner. Sounds, words, phrases, and sentences of the L2 may suddenly pop out and ring insistently in the student's head, or may be deliberately retrieved, repeated, and analyzed. Though Krashen predicted that the Din would occur as a result of comprehensible input, the Din has also been documented as an internal mechanism to cope with language that is not fully understood. There are reports of learners experiencing the Din after listening, talking, or even reading in the L2. On the question of whether the Din disappears with proficiency, there is contradictory evidence. This may possibly be an effect of the methodology used in the various studies testing the Din. A close scrutiny of the sampling procedures reveals great differences in the surveyed populations and in the way linguistic ability was determined. These variations render comparison among proficiency levels problematic. The study by Guerrero (1999) described below aimed at clarifying these discrepancies and elucidating the question of whether mental rehearsal of the L2 wanes or disappears with proficiency. Mental Rehearsal and Inner Speech Development The role of mental rehearsal in the development of inner speech in the L2 was explicitly pursued in a wide-ranging investigation conducted by Guerrero (1990/1991, 1994,1999). In two separate studies based on questionnaires and interviews, Guerrero investigated the nature-in terms of form and functions-of inner speech as it manifested itself during mental rehearsal of the L2. The first study (Guerrero 1990/1991, 1994) focused on learners at three levels of ESL proficiency-low, intermediate, and high. The second study (Guerrero, 1999) replicated the first study with a population of advanced ESL learners. Both studies were able to confirm the occurrence of L2 inner speech as various forms of rehearsal among the participants. The studies also found a positive correlation between inner speech and proficiency; in other words, as the proficiency level increased, so did the frequency of L2 inner speech. Several phonological, lexical, syntactic, and semantic characteristics of L2 inner speech were identified through these studies as well as various different functions of inner speech. It was found, for example, that, as students were mentally rehearsing in the L2, their inner speech performed a very important mnemonic role of recalling and fixing words in memory. Their inner speech was also used for instructional, preparatory, and affective purposes, among others. In addition, the studies revealed that, though generally the occurrence of inner speech increased with proficiency, certain specific

THINKING WORDS IN A SECOND LANGUAGE

81

functions decreased. Very advanced learners, for example, reported engaging in less rehearsal to store words in memory, to imitate pronunciation, and to prepare for future production than lower proficiency learners.44 The results suggest, however, that L2 inner speech can function, in advanced levels of proficiency, as a powerful and efficient tool for thought. On the other hand, the studies also suggest that at low levels of L2 proficiency, mental rehearsal, rather than representing the activation of fully fledged L2 inner speech, is a mechanism operating towards L2 internalization and the development of L2 inner speech. Succinctly, these are the findings of Guerrero's research, which are offered here in summarized form to situate her studies in perspective with other studies on inner speech in the L2. Chapter 5 will provide an in-depth description of Guerrero's two studies on inner speech and mental rehearsal as well as a detailed account of the author's later research on L2 inner speech based on learners' diaries. Mental Rehearsal as Language Play Although the concept of language play is normally associated with audible language production and has been related particularly to children's discourse (Weir, 1962; Kuczaj, 1983; Cook, 2000; Broner & Tarone, 2001), at least one researcher (Lantolf, 1997) has suggested the possibility that language play, as it occurs among adult L2 learners, may also take place as a form of non-audible, covert, private speech. It has been noticed that Lantolf s concept of L2 language play stresses the rehearsal function rather than the ludic (fun) function of language play (Broner & Tarone, 2001). Actually, Lantolf s language play phenomenon is a mixture of what has been identified in the literature as the Din, subvocal rehearsal, mental rehearsal, and self-talk (pp. 8-9). Some instances of language play included in his research instrument are talking to oneself in the L2, silently repeating L2 phrases to oneself, making up L2 sentences or words, silently practicing the L2, imitating L2 sounds, and hearing the L2 in the head. Lantolf s theoretical rationale for the occurrence of language play among L2 learners draws from Vygotsky's ZPD concept and MacWhinney's (1985) dialectic competition model of language learning. The questionnaire used in Lantolf s study, modeled on Bedford's (1985) instrument testing the Din, asked students to identify whether they played with the languages they were learning (Spanish or English). The participants were 156 college students, some taking first and second year classes of Spanish as an FL (SFL), some enrolled in advanced third and fourth year SFL classes, and others enrolled in advanced ESL classes. Although placement procedures and program requirements differed between the SFL and ESL students, making proficiency comparisons difficult, Lantolf estimated that the ESL students' level of proficiency in English was higher than that of the SFL group in Spanish and therefore took his sample to represent three levels of proficiency: SFL elementary, SFL advanced, and (more advanced) ESL.

44

Similar findings were obtained in Gutierrez (2000)'s replication study among FL college students.

82

CHAPTER 3

Based on the above categorization, Lantolf found a tendency for language play to decrease with proficiency. Despite this tendency, the results of the study unmistakably show that L2 students do engage in language play, sometimes covertly within their own heads and sometimes more overtly, as their inner speech resurfaces in the form of audible private speech or as they are in the process of internalizing the social speech they are exposed to. Why do students do this? According to Lantolf, "language play is the activity of regaining lost equilibrium" (p. 25), a situation that arises when L2 learners are confronted with some feature of the L2 that does not match their own knowledge of the L2. To overcome this conflict, learners will resort to language play. In a conversation, for example, a learner may notice something that does not agree with his acquired system. Because language play while conversing would be out of the question, the learner may later reflect on this discrepancy and make the necessary adjustments in his own linguistic system through private language play. As to why advanced learners presumably play less with the L2, Lantolf hypothesizes: As learners become more advanced, the potential conflict between their internal system and external models declines, thereby reducing the chances of the learner being thrown into a state of disequilibrium. Consequently, the need for advanced learners to engage in language play, as we have already seen, is greatly diminished or eliminated altogether, (p. 26)

While granting that language learning is far too complex a process to be reduced to one single mechanism, Lantolf supports the idea that language play is an essential condition of language learning: "Without language play learning is unlikely to happen" (p. 19). NEUROIMAGING RESEARCH OF L2 INNER SPEECH ACTIVITY Growing evidence of inner speech processes in the L2 is coming forth from research involving the application of neuroimaging techniques. As in the LI field, L2 neuroimaging has the potential for identifying brain areas implicated in L2 processing and for revealing patterns of activation in L2 neurolinguistic activity. The field of L2 neuroimaging research is thus instrumental in confirming or disconfirming long-held hypotheses about the organization of languages in the brain and in unraveling the complex architecture and functioning of the bilingual or multilingual mind. Two areas of neuroimaging research appear to be highly relevant to the topic of this book, inner speech in the L2. One is the investigation of the spatial allocation of language functions in the bilingual brain and another one is the question of how variables such as age of acquisition and L2 proficiency affect the distribution of L1 and L2s in the brain. A study by Price, Green, and von Studnitz (1999), for example, revealed different patterns of activation for different L1-L2 tasks, namely, translation and switching. Using PET, the researchers uncovered different neural systems underlying translation and language switching among proficient German-English adult bilinguals, suggesting at least partially independent mechanisms. Increased activity in the anterior cingulate and subcortical structures was found for translation, but not switching. Overall, the study indicated that switching and translation between languages involve different

THINKING WORDS IN A SECOND LANGUAGE

83

components of the language system: phonological recoding in the case of translation and semantics and articulation in the case of switching. Rodriguez-Fornells, Rotte, Heinze, Nosselt, and Munte (2002) sought to investigate how bilingual subjects are able to handle two languages in the brain without interfering with each other. Using behavioral, electrophysiological, and fMRI (functional magnetic resonance imaging) measures, the authors examined word processing in a group of bilingual Catalan-Spanish subjects who had acquired both languages early in life and in a group of monolingual Spanish subjects. Words in both languages and pseudo words were presented on a computer monitor. Subjects were asked to press a button for only the words of the language being tested, whereas no response was to be given to the other language or pseudo words. Results indicated that the meanings of the words in the experimental irrelevant language were not accessed and that these words were rejected at an early stage before semantic analysis. The study thus showed that bilinguals can effectively avoid interference between languages and turn off semantic analysis of the language not being processed. The study also evidenced greater activation among the bilingual subjects of the planum temporale, an area of the brain that has been linked to phonological processing. The authors hypothesize that bilinguals can block access to the irrelevant language through the use of an indirect access route to the lexicon; in other words, when processing the relevant language, bilinguals do not go directly from orthography to lexical access but indirectly from orthography through phonological representation to lexical access. According to the researchers, it is possible that the more convoluted pathway to meaning through phonology in the relevant language prevents semantic interference from the irrelevant language. Kim, Relkin, Lee, and Hirsch (1997) applied fMRI to determine the spatial allocation of Lls and L2s in the cerebral cortex. In the study, bilingual subjects with a variety of L1-L2 combinations performed silent (internal speech) expressive tasks with similar semantic content across languages. In different sessions alternating LI or L2, the subjects were asked to imagine sentences describing events that had occurred the previous day. The researchers found that, within the frontal-lobe region known as Broca's area, L2s learned in adulthood and Lls were located in separate regions, whereas L2s learned in infancy shared common frontal areas with Lls. In the temporal-lobe region known as Wernicke's area, however, little or no separation of activity was found for L2s and Lls regardless of age of L2 acquisition. The study suggests that, at least for language functions located in Broca's area, late L2 acquisition does result in the anatomical separation of LI and L2. Two investigations have confirmed the presence of shared cortical regions for languages learned in infancy. In a word comprehension study, Chee, et al. (1999) found through fMRI that proficient Mandarin-English bilinguals exposed to both languages early in life used common neuroanatomical areas (specifically in the prefrontal, temporal, and superior parietal regions and in the anterior supplementary motor area) for both languages. Similarly, a recent study by Chee, Soon, and Lee (2003),which investigated through fMRI the effects of word repetition within and across languages among English-Chinese bilinguals who had been exposed to the L2

84

CHAPTER 3

before the age of 4, found that common semantic neuronal networks were activated for both languages. The authors speculate, however, that some of the network components could be language specific. Utilizing PET, Klein, Milner, Zatorre, Meyer, and Evans (1995) found no evidence supporting the hypothesis that an L2 learned after the age of 5 and the LI engage different neural substrates. In a study of word generation in L1 and L2 among EnglishFrench bilinguals, the authors found common neural areas involved in within- and across-language searches. Similar findings were obtained by Chee, Tan, and Thiel (1999) in an fMRI study involving word processing in Mandarin and English. In this investigation, no differences in areas of activation for LI or L2 were identified regardless of whether English (the L2) was learned early (before the age of 6) or late (after the age of 12). While age of L2 acquisition appears in some studies to be a significant factor in the organization of LI and L2 in the brain, Perani et al. (1998), employing PET, observed no different cortical regions for early or late bilinguals as they performed story listening tasks. In the study, two groups of highly proficient bilinguals participated: (1) Spanish-Catalan bilinguals who had acquired the L2 before the age of 4 and (2) Italian-English bilinguals who had acquired the L2 after the age of 10. The researchers found that several brain areas, similar to those observed for the LI in low proficiency bilinguals, were activated by the L2, findings that suggest that attaining a high proficiency level in the L2 is more important than the onset age of L2 acquisition in the organization of Lls and L2s in the brain. Neuroimaging, in short, is beginning to reveal an extremely complex picture of the localization and handling of more than one language in the brain. Already it is evident that very sophisticated cerebral mechanisms of language storage, coordination, and segregation allow bilinguals and multilinguals to function successfully while processing one language or another. It is also clear that to fully understand inner speech activity among L2 learners, bilinguals, and multilinguals it will be necessary to untangle an intricate maze of variables that are at play in L1-L2 language processing: age of L2 acquisition, level of proficiency, nature of the verbal task, and types of languages involved, to name just a few of the most salient ones. SUMMARY OF RESEARCH ON INNER SPEECH AND L2 LEARNING Few research studies exist specifically on L2 inner speech There is, however, a considerable amount of scholarly work on verbal thought and mental processes in L2 learning that throw light on various aspects of the inner speech-L2 connection. Five areas of research were reviewed in this chapter: (a) inner speech as the mechanism for verbal thought in the L2, (b) inner speech as internalization of the L2, (c) inner speech in L2 reading and writing, (d) inner speech and mental rehearsal of the L2, and (e) neuroimaging research on L2 inner speech activity. The first area to be reviewed was that of studies dealing with the nature of verbal thought among L2 learners. These studies particularly illuminate, in fact expose, the complexity of the semantic and conceptual aspects of inner speech. An important

THINKING WORDS IN A SECOND LANGUAGE

85

question arising from these studies is what it means "to think in a foreign/second language." The literature reveals there is a simplistic tendency to equate "thinking in another language" with the process of encoding ready-made thoughts into words or with the stage immediately preceding speech externalization. Some researchers (Leontiev, A. A., 1981; John-Steiner, 1985b), on the contrary, believe that coding, or putting thoughts into words, is just one step in the process of speech production and that "to think in another language" involves a much more complex operation of formulating thoughts through the medium of the L2. Thoughts, in this view, are completed, rather than expressed, by language. This means that language has an enormous effect at a conceptual level in that it helps shape thoughts and that, particularly at high levels of proficiency, an L2 could exert a great influence on verbal thought. The research conducted by John-Steiner (1985b) with competent bilinguals in fact points to the unification of two (or more) languages at the level of verbal thought. It would appear then that any equation of inner speech with "thinking in a second/foreign language" would have to encompass not only the idea of verbalizing thoughts in some linguistic "externalizable" code but also the notion of thought completion through the L2 at a deep conceptual level. Some researchers, however, are not so much concerned with what it means "to think in another language" as with its practical considerations: Is it beneficial for learners to think "directly" in an L2? Should students be encouraged to think in an L2? Can thinking in a target language be promoted? Why do learners prefer to think in one language or another? Cohen (1998) has addressed several of these questions, and the answers he provides are by no means simple. Preferences for a "language of thought," as Cohen puts it, may depend on a variety of factors, such as the student's level of proficiency in the L2, mastery of the specific discourse domain, the content of the thoughts themselves (e.g. the thought of an event may trigger a particular language associated with it), the method or program used to instruct the learner, and the student's motivation and goals in learning the language. Even the particular purpose for which the person is privately using the language, whether for praying, solving mathematical problems, or remembering life experiences, for example, may determine a preference to think in the L2, as demonstrated by Cook (1998) and Larsen et al. (2002). All these factors could contribute to stimulate or inhibit the intramental use of a certain language at a certain moment. Moreover, the switch from one language to another may be planned or unplanned. It may be completely spontaneous and automatic, or it may be deliberate, even sometimes enforced (by the method, the teacher, etc.) Whether it is beneficial for students to try to think in the L2, this has not been proven. What is evident, however, is that no matter how much educators insist that students think in their L2, they will resort to their LI, if necessary, to mediate their thinking. To ban the LI from the students' minds is an unrealistic expectation, as Cohen (p. 170) suggests. Not only that, it may deprive students of a critical mediating cognitive tool, for example, in reading and writing L2 texts. An interesting question arising from the literature is whether the deliberate efforts that students sometimes make in using the L2 in their minds, as for example, when a person is thinking how to say something in a language, represent thinking in that

86

CHAPTER 3

language. It has been argued that this voluntary use of the L2 may be better characterized as "thinking aboutthe language" (Lantolf, cited in Cohen, 1998, p. 173). There certainly exists a metalinguistic use of the L2, which is not the same as thinking in or through a language. In this case, the use of the L2 in the mind may be so fragmentary and fleeting that it may not amount to thinking in the L2. An important point that emerges from the literature is whether the semantic basis of inner speech changes with the acquisition of an L2. According to Ushakova (1994), the basic foundation of inner speech structures built in LI acquisition remains fixed, and any language learned subsequently will be stored and organized on the lines laid by the LI. John-Steiner (1985b) adds some complexity to the issue of change by suggesting that the way two (or more) languages are related to each other in the mind shifts with development. At the beginning stages of acquisition of a new language, there will be a great dependency on the LI, as Ushakova suggested, but at a high degree of proficiency in the L2, there will be a consolidation of the native and the other language at the level of verbal thought. Like John-Steiner, Pavlenko and Lantolf (2000) and Pavlenko(1998) suggest there are stages of development in verbal thought. In their studies of identity formation among bilinguals, the authors have documented significant shifts, losses, and reconstruction in the semantic and conceptual layers of the bilinguals' languages. In addition, their studies suggest that agency and personal choice are important factors in the kind of inner speech an L2 learner experiences. Finally, further evidence that changes in the mental organization and conceptual structures of an L2 learner may affect the nature of inner speech are provided in studies of the bilingual lexicon (Pavlenko, 1999, 1996) and on the appropriation of gestures of the L2 for self-mediation (McCafferty & Ahmed, 2000). The second area of research reviewed concerns the process of internalization of the L2 that culminates in inner speech. Studies on private speech among L2 learners (Centeno-Cortes, 2003; Lantolf & Yanez, 2003; Saville-Troike, 1988; Ohta, 2001) throw light on some of the ways in which students try to appropriate the language they are learning. In these studies, it is shown how learners, in their self-addressed low vocalizations, select aspects of the language they are exposed to and work on them privately through imitation, repetition, manipulation, and language play. Other uses of private speech include rehearsing for future production, answering questions addressed to other students, and drawing upon the LI to make sense or evaluate knowledge. Although not all students may resort to private speech as a transitional phase in the internalization of the L2, it is possible that the private speech they display may share many of the functional and structural aspects of inner speech. Therein, the value of private speech research in providing insights into inner speech. The study of inner speech per se, however, may yield different or additional findings. The diary study conducted by Guerrero (2004) on inner speech development among beginning learners revealed that the early process of L2 appropriation and transformation into inner speech is characterized by a great deal of sub vocal and mental repetition, recall and rehearsal of language heard or read, but little use of the L2 for private thinking purposes. A third group of studies concerned the role that inner speech plays in the reading

THINKING WORDS IN A SECOND LANGUAGE

87

and writing of L2 texts. The most outstanding of these studies is by far Sokolov's (1972) research on reading in an FL. Using a variety of techniques (think-aloud protocols, interference with speech articulation, and electromyography), Sokolov was able to draw several conclusions about the function of inner speech in understanding FL texts. One of his most important findings was that inner speech aids in the formation of key semantic complexes, in making generalizations about the text, and in memorization. Sokolov was also able to show the relative need for concealed articulation of speech during reading among FL learners. The greater the difficulty of the text or the lower the level of proficiency of the learner, the greater the need for inner speech in the form of unfolded concealed verbalization. In other words, before processes of reading comprehension in an FL become automatic, learners first go through a phase of intense inward articulation of the text read. Sokolov's experiments highlight the importance of motor, auditory, and visual mechanisms involved in inner speech while reading in the L2. The study by Upton and Lee-Thompson (2001) also throws light on the nature of inner speech during L2 reading. Analysis of learners' think-aloud protocols revealed that as proficiency increased there was a decline in use of the LI to facilitate comprehension and an overall reduction of verbalizations, which is taken as a symptom of reduced inner speech activity with greater automatization of L2 comprehension. Finally, in this group, Huh (2002) applied the concept of inner speech to L2 writing. Her study, also based on learners' think-aloud protocols, revealed the use of LI inner speech as a cognitive tool to make meaning while writing in the L2. Both Upton and Lee-Thompson's and Huh's research challenge the notion of mental translation in reading and writing as a simple process of coding or decoding language in the mind. Rather, they present an intricate, dynamic process of meaning making in which LI inner speech plays a crucial mediating role. The fourth area of research discussed in this review focused on mental rehearsal of the L2, a language learning strategy that has been associated with the development of L2 inner speech (Guerrero, 1994, 1999). Studies within the working-memory model have suggested that silent articulatory rehearsal is effective in foreign vocabulary acquisition (Baddeley et al, 1998). In L2 acquisition, Krashen (1983) was one of the first to call attention to a form of mental rehearsal that he characterized as a spontaneous (involuntary) Din of the L2 in the head. Further studies explored more elaborate uses of mental rehearsal in L2 learning. Guerrero (1987), for example, documented the occurrence a voluntary form of rehearsal that L2 learners engage in when they want to practice the language. Lantolf (1997) investigated the rehearsal function of language play and its role in appropriating the L2. Lantolf claims that language play is an essential mechanism in L2 learning used by learners to solve conflicts between their acquired system and external models and to advance intramentally in their mastery over the L2. A point of contention around the notion of mental rehearsal is whether it increases or decreases with proficiency in the L2. There seems to be enough data in the literature to support both hypotheses. Guerrero (1999) explored this discrepancy and found that whereas L2 inner speech in general increased with proficiency, certain functions of mental rehearsal, such as memory storage and retrieval, decreased. Guerrero (1999)

88

CHAPTER 3

hypothesized that in very advanced stages of proficiency, L2 inner speech may be as powerful and effective a mediating cognitive tool as the LI. The fifth group of studies reviewed in this chapter involves the application of neuroimaging techniques such as PET and fMRI to L2 processing. Research utilizing this technology has begun to expose an extremely intricate picture of L2 inner speech mechanisms and areas of activation. Some research (Price et al. 1999) has shown that different internal L2 tasks, such as translation and language switching activate distinct neural systems. According to Rodriguez-Fornells et al. (2002), bilinguals are able to process an L2 without interference from the L1 by effectively blocking semantic access to the LI through an indirect route to the lexicon. A number of studies have looked at the way Lls and L2s are organized in the brain according to age of L2 acquisition. Whereas it appears that early bilinguals develop common neural areas for the LI and L2 (Chee, Caplan, et al., 1999; Chee et al., 2003), at least one study (Kim et al., 1997) found segregated LI and L2 zones of activation when the L2 was learned in adulthood and common brain areas for the LI and L2s learned in infancy. Other research (Chee, Tan,etal., 1999; Klein etal. 1995;Peranietal., 1998), however, has found that similar neural substrates are involved in L2 processing regardless of age of L2 acquisition. According to Perani et al. (1998), ultimate degree of attainment in L2 proficiency is more important than age of L2 acquisition in determining whether same or different brain areas will be assigned for the LI and L2. CONCLUSION This chapter has dealt with the issue of "thinking words" in an L2. In researching the L2 perspective on inner speech, only a few studies have been found to approach the topic of inner speech in the L2 in a straightforward manner. A review of the literature, however, has uncovered a host of related investigations dealing with the nature of verbal thought in the L2, the question of internalization of the L2, the role of inner speech in L2 reading and writing, mental rehearsal as a mechanism involved in the development of L2 inner speech, and brain areas of activation during L2 inner speech activity, as shown through neuroimaging technology. Throughout this chapter, relationships among areas of investigation, salient aspects of the issue of L2 inner speech, implications for future research, as well as critical commentaries have been offered. The contents of this chapter provide a background for many of the ideas and arguments presented in Chapters 5 through 7 and serve as springboard for discussion of further research in Chapter 8.

CHAPTER 4 METHODOLOGY OF RESEARCH ON INNER SPEECH The Challenge of Studying Covert Verbal Activity

This chapter deals with the problem of "method" in researching inner speech. As Vygotsky put it, "the area of inner speech is one of the most difficult to investigate" (1986, p. 226). Because inner speech is covert language activity, it has remained mostly inaccessible to direct methods of observation. Vygotsky himself opted for the "genetic method," an indirect though objective approach that focuses on the analysis of private speech, the transitional phase mediating the internalization of social speech as inner speech. The chapter reviews Vygotsky's rationale for the methodology of private speech research and subsequent efforts to extend Vygotsky's proposals in this respect. Alternative methods for studying inner speech are then examined. One of these is collecting data from verbal reports. Although some aspects of inner speech are not amenable to metaconscious analysis, subjects' verbalizations on inner speech have been elicited through questionnaires, diaries, interviews, thinking aloud, and autobiographical narratives. Some additional verbal report techniques, not so widely used but worth considering for their relative merits, are Q-methodology, cued recall, and thought-sampling. Laboratory tools and techniques employed in researching inner speech are also explored, among these the psychophysiological measurement of covert verbal responses, for example, through speech interference and electromyography, and the more recent brain scanning procedures, effective in producing powerful images of inner speech in vivo. The purpose of this chapter is to orient researchers interested in taking up the challenge of pursuing the study of inner speech on the scope of methodologies available, on how they have been used in the past, and on the possibilities they offer for future research. Each method is appraised for its advantages and limitations as well as for its potential uses in the study of inner speech. The chapter ends with a call for a multimodal, non-reductionist, and collective approach to the problem of method. THE GENETIC METHOD Throughout his professional career, Vygotsky was as much concerned with the object as with the method of his research: "The search for method becomes one of the most

90

CHAPTER 4

important problems of the entire enterprise of understanding the uniquely human forms of psychological activity" (1978, p. 65, italics used in original). For Vygotsky, the optimal methodological route to inner speech was through analysis of its genetic predecessor: egocentric speech. Vygotsky rested his defense of egocentric speech as "the key to the study of inner speech" (1986, p. 226) on two grounds: (1) inner speech is inaccessible to direct methods of observation and (2) higher mental processes must be studied developmentally. Because egocentric speech is directly observable and genetically linked to inner speech, egocentric speech was thus, for Vygotsky, the most appropriate approach to the investigation of inner speech. The study of inner speech through egocentric speech has, according to Vygotsky (1986), two advantages. First, because egocentric speech is vocalized and audible, it is accessible to observation and experimentation. "To study an internal process, it is necessary to externalize it experimentally, by connecting it with some outer activity; only then is objective functional analysis possible" (p. 227). The second advantage is that, as a transitional phase, egocentric speech shows the process of transformation that gives way to inner speech. By observing how certain characteristics of egocentric speech become strengthened while others tend to disappear, it is possible to predict "which traits are essential to inner speech and which are only temporary" (p. 227) and thus to determine the ultimate nature of inner speech. For example, from his observations, Vygotsky inferred that as egocentric speech goes underground at around school age, it loses vocalization-a transitory feature-but retains predicativity (abbreviatedness, ellipsis)—a regular trait of inner speech (p. 229). Vygotsky's (1978) concept of the method of experimentation as applied to the study of higher psychological functions stems from the assumption that mental development, like the general development of the species, must be studied as a historical process. Experimentation, in Vygotsky's view, entails explanation of "actual causal-dynamic relations" (p. 62) and not merely description of phenomena at a point in time. Traditional experimentation and introspectionism, as practiced in his times, did not allow for the observation of the genesis of mental functions-that is, their origin and development over time (pp. 59-60). The goal of Vygotsky's experimentaldevelopmental (p. 61) method is to document process, whether in ontogenesis (development over a lifetime) or in microgenesis (the short-term formation or on-line changes of psychological functions) (see Wertsch, 1985b). Whereas Vygotsky applied ontogenetic principles to his examination of inner speech development throughout childhood, he followed a microgenetic approach in his account of the externalization of thought from inner to public speech (Wertsch, 1985b, p. 55). Despite its objective basis, Vygotsky's experimental-genetic method implies making inferences on the nature of inner speech. As van der Veer and Valsiner (1991) note, this method extrapolates the properties of inner speech from those of egocentric speech, thus following an "indirect route" (p. 364). Vygotsky, nevertheless, thought it was an "excellent method" because it allowed for the investigation of inner speech "live," that is "while its structural and functional peculiarities are being shaped" and because it was objective; in other words, it depended on audible egocentric speech, which is "accessible to observation and measurement" (1986, p. 86). The fact that the

METHODOLOGY OF RESEARCH ON INNER SPEECH

91

method ultimately rests on interpretation and inference was not a drawback for Vygotsky because, in his view, such is the nature of psychological research (van der Veer & Valsiner, 1991, p. 149). If explanation is sought, scientific research has no choice but to transcend the observable. According to Vygotsky (citing Engels), "although we can never see the world through the ant's eyes, we can reconstruct its view of the world" (van der Veer & Valsiner, 1991, p. 149). Vygotsky (1986) made frequent references to the experiments he and his research associates conducted to determine the fate and functions of self-directed speech. Though lacking much methodological detail (in terms of data collection, procedures, and analysis of raw data) in their reporting, these experiments basically involved introducing some type of difficulty or frustration, for example, removing a color pencil while a child was drawing (see Vygotsky, 1986, pp. 29-34), or changing the circumstances surrounding the experiment, for instance, having the experimenter leave the room while the child was solving a problem (see pp. 232-234), to observe effects on the amount and quality of egocentric speech. Aside from a few short samples of speech produced during these experiments, however, Vygotsky "did not illustrate the syntax of inner speech by giving examples from protocols of egocentric speech registered in his own research" (van der Veer & Valsiner, 1991, p. 366). Instead, he chose to quote literary pieces (from Tolstoy, Dostoevsky, etc.) showing features of external (oral or written) discourse that might resemble inner speech. Notwithstanding certain limitations, probably due more to the quality of the reporting than to its actual conception and execution (but see Wertsch's 1985b critique of some shortcomings in its application, pp. 17-57), Vygotsky's experimental-genetic approach to inner speech is a remarkable effort to overcome the difficulties inherent in the study of inner speech. Though scholars before him had investigated inner and egocentric speech, no one (not even Piaget) had so ingeniously and effectively synthesized both strands of research into a coherent methodological and substantive approach to the issue of verbal thought (van der Veer & Valsiner, 1991, p. 364-370). By merging subject matter and methodological procedure in the study of egocentric speech, Vygotsky was able to escape the obstacles posed by the covert, subjective nature of inner speech and at the same time remain consistent with the overall framework of his sociohistorical approach to mental development. The Study of Private Speech Following Vygotsky's suggestion to investigate inner speech indirectly through its vocal, objective, manifestation, numerous researchers have dedicated themselves to what has come to be known as the study of private speech. If, as Vygotsky argued, private (egocentric) speech is similar in form and function to inner speech and different in only one aspect, outward vocalization, private speech, both among LI and L2 speakers, should provide important clues about the nature of inner speech. The study of private speech has become an important testing ground for many of Vygotsky's, Piaget's, and others' hypotheses concerning mental and linguistic development. By now, the body of work on LI and L2 private speech is quite

92

CHAPTER 4

voluminous (for an estimate of the breadth and depth of this field of study, see reviews of the literature in Berk, 1992; McCafferty, 1994a; Ohta, 2001; Winsler, Carlton, & Barry, 2000). Although all private speech studies are potentially relevant to inner speech, the main focus of the majority of these investigations is private (audible) speech as an interesting phenomenon in itself. In other words, the researchers' interest lies in finding out the structural features, amount and age of occurrence, and functions of private speech, without much concern for the implications that this form of vocalized, self-directed speech poses for inner speech. Only a few studies deliberately pursue connections between private and inner speech, that is, purposely look at private speech to find evidence that might illuminate the phenomenon of inner speech (Fry, 1992; Goudena, 1992; John-Steiner, 1992; McCafferty, 1998; Pomper, 1990). Although the first wave of private speech studies dealt primarily with private speech in the L1, a growing number of researchers have turned their attention to private speech among L2 learners or in bilingual scenarios. Because of its mediating role in the construction of the linguistically-supported, social mind, private speech has become a particularly appealing topic within the sociocultural theory approach to L2 learning. A large number of studies within the L2 sociocultural field have focused on the regulatory function of private speech during solution of cognitive-verbal tasks (Appel & Lantolf, 1994; Brooks & Donato, 1994; Brooks, Donato, & McGlone, 1997; Centeno-Cortes & Jimenez, 2004; Frawley & Lantolf, 1985; Lantolf & Frawley, 1984; Lantolf, DiCamilla,& Ahmed 1997; McCafferty, 1992,1994b, 1998). Within this line of research, private speech is conceived as externalized inner speech, which is deployed in an effort to self-regulate behavior and gain control over a task. This type of study is thus particularly relevant to inner speech as it relates to the issue of externalization. Until recently, the role of private speech in the internalization of the L2 had been little explored. A seminal investigation by Saville-Troike (1988) provided evidence, however, that learners sometimes use audible private speech in order to learn, or internalize, the L2. Following Saville-Troike, several researchers have begun to look at the internalizing or L2 learning function of private speech (Centeno-Cortes, 2003; Lantolf & Yanez, 2003; Ohta, 2001). Collecting samples of private speech is facilitated by the fact that it is overtly vocalized and therefore recordable. Private speech studies have thus usually relied on video and audio recordings and the use of such technology as video cameras and portable or wireless individual microphones. Audio and video recordings of private speech data are typically turned into written transcriptions that can be subsequently analyzed. Using microphones and video cameras, however, involves some technical difficulties that need to be taken into account. The first is that the equipment must be sensitive or positioned close enough to the speaker's mouth to record very low volume utterances (see Ohta, 2001, p. 25). The second is that mikes and cameras introduce an element of artificiality that might affect the speaker's degree of self-consciousness. To avoid some of these effects, some researchers have resorted to taking quick notes of the private speech they overhear (Berk & Garvin,1984; Kronk, 1994; Winsler et al., 2000). This alternative has two drawbacks: (a) Some low volume utterances may be out of ears' reach, and (b) no permanent records of the utterances are produced that

METHODOLOGY OF RESEARCH ON INNER SPEECH

93

could be played repeatedly for analysis. Whereas the majority of private speech studies have dealt with the oral modality of private speech, a few studies have looked at its written form, alternatively referred to as private writing (DiCamilla & Lantolf, 1995; Roebuck, 1998, 2000), introspective writing (Jensen, 1989), or inner speech writing(John-Steiner, 1985a; 1992). As with the studies of oral private speech, research on the written modality views this type of self-oriented discourse as externalized inner speech in the face of challenging cognitive tasks or as a vehicle for self-communication. There are differences among the terms used to refer to this modality, though. As DiCamilla and Lantolf (1995) point out, inner speech cannot be literally "written down" (p. 351). Therefore, their preferred term private writing reflects the fact that only "portions of one's inner dialogue with the self (p. 351) are evidenced in the written mode. Samples of private writing have been obtained from novice LI writers' compositions (DiCamilla & Lantolf, 1995) and the written recall protocols of L2 learners (Roebuck, 1998, 2000). The emphasis in these studies is on the linguistic properties of private writing, such as modality and reference. Private writing as the externalization of inner speech has also been investigated through observation of jottings addressed to the self. Pomper (1990) analyzed scribblings written by students on the margins of books to support her hypothesis that marginal notes usually reveal typical inner speech features The term inner speech writing, preferred by John-Steiner (1985a, 1992), responds to the telegraphic and regulatory features of writing for the self. John-Steiner found evidence of inner speech writing in the notebooks of creative thinkers, expert writers' journals, and scientists' laboratory records. This written mode usually consists of cryptic, highly condensed notes addressed to the self, jotted down quickly to anchor thoughts that might later be given greater attention or serve as the basis from which more expanded and elaborated text is produced. Written samples of inner speech are not just found in the writings of experienced thinkers. A quickly written grocery shopping list is also a highly condensed reflection of inner thought acting as an effective externalized mnemonic device (John-Steiner, 1992, p. 293). The third term, introspective writing, focuses on the private, ideational content of self-directed writing. Jensen's (1989) thesis is that personal forms of writing such as autobiographies, diaries, journals, and memoirs often reflect the processes of introspection that characterize intrapersonal communication.45 Both John-Steiner (1985a) and Jensen (1989) compare their inner speech written modalities to the ongoing stream of consciousness described by William James (1950) and the literary style derived from this concept, which is usually associated with the works by James Joyce, Virginia Woolf, and William Faulkner. The study of private speech is not without its hurdles. Diaz (1992) has pointed out four methodological difficulties: (1) collecting enough quantities of spontaneous private speech from children to support research, (2) distinguishing private from social 45 Although introspective writing may be described as self-directed, it should not be taken as a form of inner speech. Diaries, memoirs, and autobiographies may be primarily written for the self, but in form they follow the conventions of externalized public speech.

94

CHAPTER 4

speech (see also Wells, 1998), (3) categorizing private speech utterances, and (4) specifying the relationship between private speech and task performance. Another problem is the distinction between private discernible speech and private unintelligible speech. As it happens, private speech, though audible to an observer and obviously articulated by the speaker, may be so low or telegraphic that it is practically incomprehensible. This may occur in child as well as in adult private speech. References in the literature to data consisting of inaudible mutterings, soft whispers, or incomprehensible utterances abound. The problem is further complicated by the fact that, in some cases, incomprehensibility is caused not by the quality of the utterance per se but by the method of data collection. In other words, a muttering may appear to be so because the tape or tape recorder used to record the utterance was not sensitive enough in capturing its phonological features or, if not tape recorded, because it was not clearly heard by the observer (see Ohta, 2001, for a categorization of reduced volume in private speech utterances, p. 37). The incomprehensibility of some forms of private speech results in problems of categorization. What function does an unintelligible uttering have? If the contents of a private speech utterance are unknown, its function (other than that of self-regulation) cannot be discerned. Beyond the intractable issue of function, private speech mutterings, however, may yield important information on the moment-to-moment changes that attend the microgenesis of speech internalization. By attending to certain features of mutterings and whispers (the circumstances in which they arise, their relationship to preceding or following private speech, the subject's simultaneous gaze and engagement with objects or task), researchers may gain insights into the transformation that private speech undergoes as it turns into inner speech. Advantages and Limitations of Researching Inner Speech through Private Speech Researching private speech was Vygotsky's method of choice for the exploration of inner speech. To its advantage, the method enjoys objectivity-it is based on audible speech that can be heard or recorded by others-and is consistent with a sociogenetic approach to mental development. The study of private speech has thus become an important methodological window not only into the realm of inner speech but also into the genesis of higher mental functions. As Diaz and Berk (1992a) put it, "private speech plays a central role in understanding, validating, and expanding the ideas contained in [Vygotsky's sociocultural approach]" (p. 2). The method, however, is by force limited, as it reaches inner speech only in a roundabout, indirect way, a problem plaguing the majority of the approaches to the phenomenon. Private speech is verbalized speech and as such only an outward reflection of what inner speech might be actually like. Thus, as with all methods that pretend to explain the unobservable through its visible manifestations, the study of inner through private speech is inferential and predictive (Wertsch, 1979b). The method is nonetheless a valid and useful means for the generation of hypotheses and predictions about inner speech. Where private speech appears to be most valuable is in providing clues into the issues of internalization and externalization of inner speech. Several questions

METHODOLOGY OF RESEARCH ON INNER SPEECH

95

concerning these issues could be or have been partially pursued through analysis of private speech: • When does inner speech take over as the prevalent mode of self-regulation and primary means for thought? • On what occasions does inner speech reemerge as externalized private speech? • Is the absence of private speech an indication of higher levels of self-regulation through inner speech? • What are the linguistic features of private speech that might be reasonably inferred to be preserved in inner speech; and vice versa, what linguistic features of inner speech are preserved in private speech as it is externalized? • What is the role of mutterings in the internalization of inner speech? • What mental functions, not adequately served by private speech, at least among adults, are taken over by inner speech? In the case of L2 learners in particular, the study of private speech can be equally powerful as a source of insights into L2 inner speech. Following are some of the questions that might be profitably explored through private speech: • When does L2 inner speech need to be externalized as L2 private speech? • To what extent is the LI a medium for private thinking among L2 learners? • What do the private gestures (as a variant of private "speech") of L2 learners reveal about the relationship between inner speech and thought in the L2? • Do less proficient learners use more private speech than more advanced learners? • Is private speech a "universal" prior-to-inner-speech phase among L2 learners? VERBAL REPORTS One of the major sources of information on inner speech has been verbal reports. In this chapter, the methodological category referred to as verbal reports includes all methods of data collection in which the source of the information is the subject's own statements or verbalizations about the object of study, in this case, inner speech. Verbal reports have had their ups and downs in the eyes of researchers as legitimate methods of mental data collection (Ericsson & Simon, 1984; Lieberman, 1979; Lyons, 1986; Nisbert & Wilson, 1977). Favored at the turn of the twentieth century, experimental introspectionism and its techniques for "observing" private, mental experience became under attack in behavioristic research. With the advent of cognitive psychology, however, verbal reports, which were once thought unreliable and inadequate as sources of evidence about cognitive processes, found renewed support as ways of gaining access into covert, mental phenomena. A return to introspective data was also bolstered by the appearance of information-processing models of memory, which contributed to delimit the scope and conditions of verbalizing mental processes. This renewed interest in verbal reports was accompanied by a critical awareness of the strengths and limitations of verbal data and of the need to follow rigorous methodological procedures. In short, verbalization of mental processes is back, but not without a judicious realization of the events and conditions in which verbal reporting can be most usefully and appropriately applied.

96

CHAPTER 4

Verbal data are often categorized into two kinds: introspective and retrospective. Introspection typically requires the subject to verbalize thoughts while performing a task, that is, while the data are still heeded in short-term memory. Retrospection calls for the subject to report on cognitive processes related to an activity immediately after they occur or some time after their occurrence. A useful way of classifying verbal report data that incorporates the notions of introspection and retrospection is Cohen's (1987, 1998) distinction into three types: self-report, self-observation, and selfrevelation. In this chapter, such typology will be used to describe reports on inner speech. Self-report Data based on subjects' statements about what they tend to do or "think" they do; generalized observations about processes of learning or using language, not referring to any specific experience or event; information is by nature retrospective and potentially interpretive. Example: "When I have to make a phone call [in my L2], I rehearse in my mind what I want to say." [author's data] Self-observation: Comments based on observation of specific language behavior; observations are made introspectively while task is being performed or retrospectively some time after the event (immediate, early, or delayed retrospection). Example: "At this point I think I was thinking that this word must start with a vowel because of the apostrophe" (comment made during interview some time after think-aloud; Feldman & Stemmer, 1987, p. 259) Self-revelation: Usually referred to as "thinking aloud;" externalization of on-going thoughts as task is being performed; verbalization is made introspectively without analyzing or editing thoughts. Example: (Subject is looking for word "publicitaires" while filling in blanks) " A les pet films public A les petits films public A A ation (?) yes les petits films publica" (A pause, AA longer pause, Feldmann & Stemmer, 1987, p. 260) Pros and Cons of Using Verbal Reports Some general observations concerning the strengths and weaknesses of verbal reports are due before a critical appraisal of their application in the study of inner speech. To begin with the disadvantages, one major problem limiting the use of verbal reports is that most cognitive processes occur without conscious awareness and are therefore unavailable for reporting. Another serious drawback is that verbal reports rely on memory. Unfortunately, memory is not always reliable. Three main problems associated with memory have been identified: (1) Even with conscious awareness of the process, the information may have been forgotten and may thus be irretrievable; (2) the information may be recalled or reported incompletely; and (3) the information may be recalled or reported inaccurately (Ericsson & Simon, 1980; Lieberman, 1979; Nisbett & Wilson, 1977). In relation to the last two points, an acknowledged shortcoming of verbal reports is that subjects may fill in memory gaps with inferences rather than report their actual thoughts. Similarly, subjects may offer interpretations

METHODOLOGY OF RESEARCH ON INNER SPEECH

97

of their recollections in the form of plausible causal theories (Nisbett & Wilson, 1977) or culturally stereotyped accounts reflecting "folk psychology" (Lyons, 1986). Another complication that can affect the quality of verbal reports is the use of faulty elicitation techniques. Instructions that are difficult to comprehend or that push for information that is beyond accessibility are some examples. Finally, there may be the problem of insincerity. Out of pressure, desire to impress, or sheer dishonesty, subjects may fabricate their reports. Why would researchers still resort to verbal report data in the face of such serious limitations? The principal allure is that self-reports provide information about mental processes that cannot be obtained through external observation. Not only do verbal reports offer glimpses of how certain cognitive operations are conducted, they are particularly valuable as sources of data on learners' "metacognitive knowledge, i.e. what they know and can report about their language learning" (Wenden, 1986, p. 197). Verbal reports have the additional attraction of constituting the participants' perspective; they are first-hand accounts of mental activity. As such, verbalizations may offer important insights into learners' subjective experiences, perceptions, and beliefs that are beyond access to an observer. By providing "insider" testimony, selfobservation reports help researchers understand the learners' stance, their attitudes or motivations. Furthermore, when naturally contextualized, first-person reports have a great potential for ecological validity; that is, they may significantly inform on real life, socially-embedded mental behavior (Hurlburt, 1997; Pavlenko & Lantolf, 2000). Finally, as in the case of private speech, the information gathered through verbal reports can yield hypotheses, serve as points of departure, or suggest avenues of research that might be pursued through other methodological means. Methods of Verbal Data Collection in the Study of Inner Speech Five major techniques for gathering verbal report data on inner speech have been employed: questionnaires, interviews, diaries, first-person narratives, and think-aloud. Three additional, not so well-known, verbalization techniques-Q-Methodology, cued recall, and thought sampling-have also been found to generate inner speech data. Each one of the techniques reviewed in this chapter lends itself for certain types of verbal report data. For example, first-person narratives typically take the form of retrospective self-reports whereas think-aloud protocols generate introspective accounts of the self-revelation type. Some of these techniques have been utilized to obtain various types of verbal data; for example, interviews can produce both selfobservational introspective data as well as self-report retrospective accounts. Questionnaires Questionnaires have been extensively used to obtain learners' verbal reports in the L2 field. In a questionnaire, a series of written questions or prompts probe the subjects for answers on specific topics or areas of concern. Questionnaires are quite flexible in structure, ranging from the strict closed-question format with answers marked on a

98

CHAPTER 4

Likert scale to the less controlled design based on open-ended questions. In most cases, questionnaires elicit generalized responses about habitual behavior or nonrecent events, thus generating self-reports based on delayed retrospection. It is possible, however, to administer a questionnaire promptly after a task to obtain some early or task-specific retrospection (see Cohen, 1984). The questionnaire has been the most widely applied method of verbal data collection in the study of inner speech. Smith (1983) used a short open-ended questionnaire in an informal survey he performed on the self-talking habits of college students. Researchers working within an intrapersonal communication perspective have resorted to the questionnaire to explore people's "imagined interactions" (Allen, David, & Kung, 1997; Honeycutt et al., 1989). Researchers have also developed questionnaires to pursue aspects of intrapersonal communication, such as people's inner voice experiences and their role in creating an alter ego (Hamilton, 1997) and the correlations of inner speech with self-deception, depression, and self-consciousness (Siegrist, 1995). In the field of L2 learning, Bedford (1985) designed a questionnaire to gather data on the din phenomenon, an instrument that served as the basis for Guerrero's (1987) study of mental rehearsal in the L2 and Lantolf s (1997) research on L2 language play. Guerrero (1990/1991,1994,1999) again used questionnaires in her wide survey on inner speech during mental rehearsal of the L2 among learners of various ESL proficiency levels (see Chapter 5 for full details of this research), a methodology that was replicated in Gutierrez's (2000) inner speech study among FL learners. Finally, researchers have utilized questionnaires to explore L2 speakers' preferences for a language of thought (Cohen, 1998; Larsen et al., 2002) and for internal purposes (Cook, 1998). Q-Methodology Verbal reports on inner speech were obtained by Aitken (1997) by means of Qmethodology, a widely used technique in the social sciences that is aimed at measuring subjective experience. The basic procedure of Q-methodology i s card sorting (Brown, 1996), which involves having subjects sort out cards with statements on the topic of interest according to extent of agreement. In Aitken's study, 66 statements of opinions or feelings on the nature of self-talk were elicited from college students' essays and interviews with non-student subjects. The statements fell under three categories: purpose of self-talk, target of the inner speech, and type and frequency of self-talk. The subjects were asked to sort out cards containing the statements into three groups: agree, disagree, and neutral. Thus, a typology of inner speech patterns was constructed. Three types of self-talk user emerged: the Planner, who uses self-talk to manage interpersonal relationships; the Playful type, who uses self-talk creatively, and the Quiet Talker, who uses self-talk to cope with reality. The author warns, however, that these types do not represent real individuals but rather composite pictures of similar people.

METHODOLOGY OF RESEARCH ON INNER SPEECH

99

Cued Recall A cued recall methodology was used by Larsen et al. (2002) to investigate the question of what language-Ll or L2-was used by early and late bilinguals as the inner speech of their autobiographical memories. The researchers elicited self-reports from the participants by showing them cards containing cue words (e.g. "house"). One group of words was in the LI and another group of words was in the L2. Upon being shown the word, the participants had to respond by giving a memory of a related event. The participants had to write a brief description of the event they recalled. After each recall, the participants also had to (a) indicate whether the memory came to them in their LI or their L2 or if they were not sure and (b) assign a date to the event they remembered. Because the cued recall methodology elicited memory of an event and, soon after, information on the language in which the memory was recalled, the verbal reports obtained were of a self-observation, immediate retrospection type. An additional procedure employed in this study was a brief questionnaire in which the participants had to rate-on a frequency scale-their current use of the LI or the L2 to conduct several mental activities that typically require inner speech, such as writing a note to oneself, thinking in words, and using language in dreams. The questionnaire served to elicit delayed retrospective verbal reports on the "languages" of inner speech. Interviews In an interview, a subject responds orally to a series of questions posed by the researcher. Depending on the recency of the learning task, interviews may be introspective or retrospective; they may also be more or less formal in structure, although interviews are usually more flexible than written questionnaires. Interviews are sometimes employed as a follow-up procedure after an activity, thus generating immediate or early delayed retrospection; frequently, they are an integral part of the stimulated recall technique (to be discussed below; Gass & Mackey, 2000). Several studies have made use of the interview technique. John-Steiner (1985b) obtained retrospective self-reports on the role of verbal thinking in dual language development by conducting interviews with adult bilinguals. Rohrkemper (1986) elicited immediate retrospective reports by interviewing children on their use of inner speech during math problem solving. Flavell, Green, Flavell, and Grossman (1997) used oral questions to assess the extent of children's knowledge of inner speech. In their study, children were first probed on whether they could tell if a person was thinking up in his head and thus infer the presence of inner speech in others. Children were then asked to perform a task silently (for example, to think how old they were) and subsequently respond on how they had thought about the task (verbally or visually). Guerrero (1990/1991, 1994) included two post-activity interviews in her study of inner speech during mental rehearsal of the L2. The same was done in Gutierrez's (2000) replication study of FL inner speech. Interviews were carried out one day after the activity to allow for reporting of any after-the-task rehearsal. Cohen (1994) utilized a questionnaire-based interview to examine the "internal language environment" (language used for performing mental tasks) of learners enrolled in a

100

CHAPTER 4

language immersion program. The study involved both self-observation data (elicited in post-activity interviews) and self-reports (through general questions about the learners' attitudes and preferences concerning thinking in the L2). Interviews have been frequently used as part of the stimulated recall procedure. To stimulate recall of a specific cognitive activity, a tangible record (visual, written, oral) of the activity as it was being performed is presented to the subject as a stimulus to recall and thus comment on the thought processes involved in carrying out the activity (Gass & Mackey, 2000). Cohen and Olshtain (1993) videotaped L2 students' interactions with native speakers and then, after playing back the tapes, asked the students about their efforts in planning and executing the interactions, their language search and retrieval, and their selection of a language of thought. In her diary research on the early stages of inner speech, Guerrero (2004) carried out brief stimulated recall interviews for purposes of clarification and expansion of entries. In order to examine the role of inner speech in the L2 writing process, Huh (2002) conducted stimulated recall interviews by asking subjects to comment on the compositions they had just written. Upton and Lee-Thompson (2001) employed retrospective interviews on thinkaloud protocols to find out about the use of the LI as a tool for thought in the L2 reading process. Diaries Diaries have been extensively used as a method of data collection in the language learning field (see reviews in Allison, 1998; Allwright & Bailey, 1991; Bailey & Oschner, 1983; McDonough & McDonough, 1997). In language learning diaries, learners systematically retrospect and write about their experiences in learning the language. Wilga Rivers's language learner diary, one of the earliest in the L2 literature, offers much anecdotal evidence of inner speech activity. In the following examples, she reports internally deploying Spanish, the sixth foreign language she was learning: / I . . . make mental translations and these give me a feeling of security, yet I do find myself thinking directly in the language when I read, or go over an assignment, or create utterances in class. (Rivers, 1979, p. 71) / I find I'm thinking in Spanish when I make my little sentences, although I consciously construct them in my head, with due attention to any rules, (p. 75)

Although incidental comments concerning the use of L2 inner speech, such as Rivers's above, are frequent in personal learning accounts, only one study (Guerrero, 2004) has deliberately exploited the diary as a technique to gather data specifically on L2 inner speech. Guerrero's (2004) diary study (to be discussed at length in Chapter 5) had two purposes: (1) to investigate the early stages of L2 inner speech development and (2) to test the effectiveness of the combined diary and stimulated recall technique in the elucidation of L2 inner speech phenomena. The diary yielded self-observation reports in the form of immediate and delayed retrospection. Immediate retrospection occurred as students wrote down reports of the inner speech events they had just experienced in class. Outside the classroom, students sometimes recorded their inner

METHODOLOGY OF RESEARCH ON INNER SPEECH

101

speech events immediately after experiencing them and sometimes in a more delayed fashion, within a few hours or days. The diary also offered instances of retrospective self-reports as students generalized on their inner speech experiences in the past. First-Person Narratives A novel methodological approach involving verbal reports that may yield important insights into inner speech is the examination of learners' first-person narratives. Pavlenko and Lantolf (2000) adopted this procedure in their study of late bilinguals' autobiographical memoirs narrating the experience of becoming members of an L2 community. In their data, Pavlenko and Lantolf found evidence of inner speech processes related to loss of a linguistic identity and reconstruction of a new self on the basis of the L2. First-person narratives typically produce delayed retrospective reports; in fact, most of them are written long after the circumstances that prompted them or are generalized reflections on non-specific past experiences. Autobiographical reports are therefore highly interpretive and reconstructive. Indeed, an unstated purpose of this type of narrative is for the writer to self-interpret events in his or her life, frequently to bring order and make sense of confusing and potentially traumatic life experiences. As Pavlenko and Lantolf explain, "personal narratives . . . allow [individuals] to make their own lives cohesive; that is, to understand what they are and where they are headed" (p. 160). The approach to the data is therefore hermeneutic; in other words, the emphasis is on interpretation, rather than objective explanation. Think-Aloud Protocols The think-aloud technique requires subjects to vocalize their thoughts while performing an action. When thinking aloud, subjects must verbalize what is present in their minds under the focus of their attention. As mentioned, verbal reports generated by thinking aloud are of the self-revelation, introspective type (Cohen, 1987, 1998). In other words, the method calls for subjects to introspect and reveal their current thoughts without simultaneously trying to interpret or comment on them. Instructions ask subjects to simply vocalize what's on their minds: "Don't plan what to say or speak after the thought, but rather let your thoughts speak, as though you were really thinking out loud" (Silveira, cited in Ericsson & Simon, 1984, p. 81). Think-aloud protocols are rich sources of data on inner speech. Early in 1969, Benj afield found evidence suggesting that thinking aloud makes overt the normally covert processes of inner speech. Benjafield's study revealed that thinking aloud, when compared to immediate retrospective reports, was more elliptical, presentoriented, and contained more indefinite referents, all typical features of inner speech. Thinking aloud, among other techniques, was employed by Cohen (1994) to gather data on the language used internally by children to work on math problems. Some studies have employed thinking aloud to investigate inner speech during reading. In Sokolov (1972)'s research on understanding texts, for example, "the subjects were instructed to think aloud when working with the texts, i.e., to express their thoughts no matter how redundant, extraneous, or absurd they might seem to them" (p. 78). Using

102

CHAPTER 4

think-aloud followed by retrospective interviews, Upton and Lee-Thompson (2001) found evidence of LI inner speech in L2 reading. Similarly, think-aloud reports have been utilized to gain access into inner speech processes during writing. The following excerpt from a think-aloud protocol in Woodall (2002) shows language switching (switching back and forth between the L2 and the LI) in the composing process of an English-speaking student of Japanese: [As student is reading the prompt] Khaki that's summertime no gogaku English program no a-ku-ti-vi-ty ok program activity de rek ta de- direction? De aru su-mi-su shi ga sumisu shi I'm assuming that's like Smith-san is [looking up in the dictionary] s h i . . . shi [inaud] n o t . . . n o t . . . n o t . . . teacher? Don't have it in here . . . I don't know what that means . . . Well it's part of the paper . . . toku betsu katsu doo some sort of life style toke betsu katsudoo(pp. 20-21)

Thinking Aloud: Verbalization of Thought or of Inner Speech? In their response to Smagorinsky's (1998) claim that verbalization of thinking involves articulation of thoughts through speech, Ericsson and Simon (1998) argue that under some circumstances it is possible for adults to think aloud without altering the structure and course of their thinking. Only when asked to describe and explain their thoughts, would subjects be forced to interrupt the flow of thinking and focus on the thoughts they just had, thus generating additional thoughts. According to Ericsson and Simon (1998), thinking aloud reflects inner speech. To these researchers, the verbalizations produced during think-aloud are consistent with Vygotsky's conceptualization of inner speech as disconnected and incomplete: "With think-aloud . . . the participants do not appear to monitor their overt verbalizations of thoughts, mostly generate incomplete sentences and phrases, and rarely correct their verbalizations including speech errors" (p. 181). The subjects' overt vocalization of their thoughts coincides, in these researchers' view, with Vygotsky's description of inner speech as thinking almost in pure meanings. A subject's audible speech produced while mentally multiplying 36 times 24 would constitute overt evidence of the inner speech going on in the subject's mind: O K . . . 3 6 t i m e s 2 4 . . . u r n . . . 4 t i m e s 6 is 2 4 . . . 4 . . . c a r r y t h e 2 . . . 4 t i m e s 3 i s 1 2 . .. 1 4 . . . 1 4 4 . . . 0 . . . 2 t i m e s 6 is 1 2 . . . 2 . . . c a r r y t h e 1 . . . 2 t i m e s 3 is 6 .. . 7 . . . 7 2 0 . . . 7 2 0 . . . 1 4 4 p l u s 7 2 . . . s o it w o u l d b e 4 . . . 6 . . . 8 6 4 ( E r i c s s o n & S i m o n , 1 9 8 7 , p . 34)

Although one might concur with Ericsson and Simon that thinking aloud "reflects" inner speech, under a Vygotskyan view it could never be claimed that thinking aloud is a direct verbalization of silent thinking, which is Ericsson and Simon's (1998) primary contention about think-aloud protocols that are produced when there is no intervening description or explanation of the subject's own thinking. Smagorinsky's (1998) main point in his critique is that the process of verbalizing thought is always a mediated process and that the act of mediation itself is a transformative and creative act. As Vygotsky( 1986) expressed: Thought is mediated by signs externally, but it is also mediated internally, this time by word meanings. Direct communication between minds is impossible not only physically

METHODOLOGY OF RESEARCH ON INNER SPEECH

103

but psychologically. Communication can be achieved only in a roundabout way. Thought must first pass through meanings and only then through words (p. 252)

Think-aloud protocols, then, are not and could not be taken as direct encodings of thought. In Smagorinsky's (1998) words: If thinking becomes rearticulated through the process of speech, then the protocol is not simply representative of meaning, but an agent in the production of meaning. What is studied, to use Vygotsky's . . . metaphor, is the shower of words and not the storm cloud of thought, (p. 173).

Ericsson and Simon (1998) graphically represent thinking aloud as involving the following transition: thought—>verbalization of thought (p. 180). The act of verbalizing thoughts that is called upon during the think-aloud procedure, however, by necessity renders wordless thought into inner speech. From a sociocultural theory point of view, therefore, a more accurate representation of the transition involved in thinking aloud would be thought—> inner speech—> verbalization. (This is not to suggest that there is a straightforward, smooth progression between these phases. As has been seen in Chapter 2, the processes of thought formulation and verbalization through inner speech are extremely complex.) It seems the notion of mediation through inner speech is implicit in Ericsson and Simon's model, but a clear sociocultural position should make this idea explicit. Randomly Sampling Thinking A method that holds potential for the study of inner speech is the one described by Hurlburt (1997) as randomly sampling thinking. This methodology, which comes in three variants-thought sampling, experience sampling, and descriptive experience sampling-requires informants to carry beepers into their normal spheres of life. Ringing at random intervals, the beepers signal the informants to focus on the thought they are having at the moment of the beep and to report aspects of this thought experience by filling out a short questionnaire or writing a brief description. The main purpose of randomly sampling thinking is to obtain information on various features of thinking (such as participant's mood, visual imagery, and interior monologue) and on the context of the thought experience (whether the participant was alone, what he/she was doing at the moment, etc.). Because participants carry the beepers as they move around in their natural environments and activities, thought sampling is claimed to be "ecologically valid" (p. 941) rather than laboratory controlled. Hurlburt (1997) believes that thought sampling studies "are useful because they provide insights that are difficult or impossible to obtain by other methods" (p. 943). Hurlburt exemplifies his point by citing the case of a patient, Donald, who was suffering from severe anxiety attacks. When clinically interviewed, no apparent cause for the anxiety had been detected. Donald had reported an excellent life situation both at work and in his relationship with his wife and children. The randomly sampling technique, however, revealed that one third of Donald's thoughts were occupied with feelings of annoyance toward his children. Donald's case raises two important points: (a) people are not completely reliable in their long-term retrospective reports insofar

104

CHAPTER 4

as they may neglect or misrepresent aspects of their inner experience, particularly if these may cause pain or embarrassment, and (b) introspection or immediate retrospection (as in thought sampling) may provide a more direct, and perhaps more accurate, route to conscious cognitive activity than other types of self-reports. Although this procedure has been carried out to determine the thought characteristics of specific populations (older people, bulimic patients, schizophrenics) for clinical or therapeutic purposes, some of the studies reported by Hurlburt (1990; Hurlburt, Koch, & Heavey, 2002) make reference to findings on inner speech experience. These studies, for example, suggest that verbal thoughts and visual imagery are more prominent than other types of thinking. They also show that inner speech, or symbolized thinking, is not the only type of inner experience. In other words, though inner speech may be a frequent occurrence among normal individuals, thinking is sometimes unsymbolized, that is, unmediated by words or other symbols. In the case of autistic individuals, the descriptive sampling technique yielded only visual imagery and no experience of feelings or inner speech. So far, however, the full potential of randomly sampling thinking in the study of inner speech has been largely untapped. The type of introspective, self-observation reports generated by this methodology could throw light on various aspects of inner speech, particularly function. In short, randomly sampling thinking appears to be worth experimenting with in the study of inner speech. Advantages and Limitations of Researching Inner Speech through Verbal Reports Like all types of verbal reports on cognitive or learning processes, verbalizations on inner speech display both strengths and weaknesses. Rich as sources of first-hand subjective experience and insights into inner speech features and functions, verbal reports are nonetheless limited in the extent and quality of the information they can yield. The main factor affecting the success or failure of verbal reports in providing reliable data on inner speech is the availability of information on the phenomenon in consciousness or memory. Only when inner speech is attended to can it become the object of verbalization. Unfortunately, a large number of inner speech processes are beyond metaconscious attention and control (Frawley, 1997). Likewise, only when sufficiently well remembered can an inner speech event be accurately and precisely described. The more distant in time a self-report is from the actual event, the greater the probability of inaccuracy, impreciseness, or incompleteness in the report. Let us analyze the appropriateness of applying introspective techniques such as thinking aloud to the study of inner speech. Think-aloud verbalizations are purported to reveal thought processes at the moment they are occurring. According to Ericsson and Simon (1984), there is a very close, almost simultaneous coordination between thinking and speaking: "For many speakers, normal speech seems to be uttered as it is organized. The conceptual arrangements behind speech can be worked out at nearly the same time the sentence is produced" (McNeill, as cited in Ericsson & Simon, 1984, p. 225). From this it could be inferred that think-alouds and the protocols they generate are direct records of inner speech. Certainly, the fragmentary nature of

METHODOLOGY OF RESEARCH ON INNER SPEECH

105

verbalizations evokes the brevity and ellipticality hypothesized for inner speech. Inner speech, however, is by definition internal, nonaudible speech behavior. Any verbalization of inner speech, as close to and concurrent as it may be with on-going inner speech, is a disruption of the internal flow of symbolically mediated thinking and thus potentially different from its internal precursor. Several things may happen as inner speech is verbalized: It may become expanded, or it may be reduced; it may be edited, diverted, or ignored. Consider the example given earlier in this chapter (see p. 102) of a think-aloud protocol from Woodall (2002). How can one be sure that words such as "I'm assuming," "I don't know what that means," "Well" are not just elaborations of on-going inner speech put forth for public display? Think-aloud protocols should not be taken as direct records of inner speech. As mentioned, the process of thought—> inner speech—> verbalization does not involve smooth, direct transitions, nor is one stage an exact replica of the next one. Thinking aloud may be the closest one can get to the experience of inner speech, but it is only an indirect, externalized reflection of the phenomenon. Adding a twist to Labov's (1984) "observer's paradox" (p. 61), one might say that thinking aloud poses an "intraobserver" paradox: To observe ourselves as we go naturally about our mental functions, we have to stop the flow and disrupt this mental process. Caution should also be applied when it comes to interpreting just what type of information about inner speech thinking aloud produces. As mentioned, introspection can only have access to thinking that is being or has been heeded. It would not be appropriate, therefore, to ask or expect subjects to report accurately on inner speech processes they were not aware of, say, for example, the mental operations that led to the encoding of a particular thought into verbal form. The following example from Nisbett and Wilson (1977) illustrates this point: If a person is asked, "What is your mother's maiden name?", the answer appears swiftly in consciousness. Then if the person is asked "How did you come up with that?", he is usually reduced to the inarticulate answer, "I don't know, it just came to me." (p. 232)

The problem of asking such a question would be not just the absence of an answer but the production of a response that has little to do with the actual processes of the mental operation under discussion. It would be legitimate, however, to ask the person about the contents of her inner speech. As Nisbett and Wilson (1977) point out, although the process of thinking is not available to introspection, the product of that thinking may be.46 This does not mean that the verbal report will be a direct, exact copy of the inner speech it externalizes. As mentioned earlier, the process of externalization of inner speech may entail several changes. However, when it comes to the "contents" of the

46 It is interesting to note Vygotsky's (2000/1995) comments in this respect. Vygotsky did not believe that consciousness could focus on itself; in his words, "one cannot think one's thoughts, i.e., [one cannot] grasp the very mechanism of consciousness;" however, "as soon as a thought is completed, i.e., as soon as a reflex is formed, it may be consciously observed." (Part V, paras. 21-22). In other words, whereas one does not have access to the processes that lead to consciousness because they are unconscious, one may consciously observe the resulting thought of that consciousness, which is the point that Nisbett and Wilson (1977) were trying to make.

106

CHAPTER 4

subject's inner speech, it is safe to assume that verbal reports may produce faithful information, of course, provided the verbalization is gathered sufficiently close in time to the experienced inner speech. According to Nisbett and Wilson (1977), the fact that people do not have direct access to cognitive processes does not mean they do not possess a great deal of private knowledge. The individual knows a host of personal historical facts; he knows the focus of his attention at any given point in time; he knows what his current sensations are and has .. . "knowledge" at least quantitatively superior to that of observers concerning his emotions, evaluations, and plans. (Italics added, p. 255)

Think-alouds may thus give important clues about the contents and consequently the functions of inner speech. Verbalizations of inner speech may even reveal the circumstances and the motives (as Nisbett & Wilson, 1977, put it, "the stimuli," p. 251) of inner language behavior. Actually, as Nisbett and Wilson suggest, too, "the individual's private access to content will sometimes allow him to be more accurate in his reports about the causes of behavior than an observer would be" (p. 256). Immediate retrospection is one step removed from introspection when it comes to timing in verbal reports. Post-activity interviews, questionnaires, or diary entries promptly assigned after a task or written after an inner speech event generate immediate retrospective reports that may have much to say, again, on the contents, functions, circumstances, and motivations of inner speech. The same may be said of the self-observation, thought-sampling technique proposed by Hurlburt (1997). The fact that immediate retrospective reports are not concurrent with on-line inner speech activity, however, makes them even less likely than think-alouds to be accurate on aspects of form. As is well-known, form (in contrast to meaning) is notoriously prone to decay in verbal recall tasks (Sachs, 1967; Slobin, 1979). Still, some remnants of the original inner speech structure may be preserved in memory at the time of the reporting and thus reflected in the reports. Promptly administered retrospective instruments are also useful for clarifying data reported during think-alouds. They may also indicate the language in which a memory came to mind, as in the cued recall technique (Larsen et al., 2002), and are useful complementary methods in the psychophysiological study of inner speech (as will be discussed in the next section). Early retrospective reports-that is, those removed a few hours from the mental event-can yield much the same type of data as immediate retrospective accounts. The farther removed from the event, however, the less accurate and precise the report will be. The stimulus prompting the inner speech event and the circumstances surrounding it may begin to fade, and the likelihood of reconstruction on the basis of "plausible" explanations (Nisbett & Wilson, 1977) increases. Questionnaires, interviews, diaries, and other techniques such as Q-methodology and first-person narratives, which are not related to any recent or specific inner speech activity, produce delayed retrospection. Because of their considerable removal in time from any inner speech occurrence and because they depend on self-reflection, such instruments are appropriate for eliciting attitudes, beliefs, preferences, and certain general tendencies toward the phenomenon. Delayed retrospective tools are

METHODOLOGY OF RESEARCH ON INNER SPEECH

107

particularly sensitive to information about what subjects know or think they know, therefore yielding so-called "metacognitive" information on inner speech. Like all other forms of verbal reports, questionnaires, interviews and diaries take the insider's perspective, a view that may help researchers better understand the role of inner speech in L2 learning. A benefit for students derived from these retrospective reports is that, in producing them, learners may begin to develop a self-awareness of the phenomenon. Inner speech, however, is notably fast and elusive and thus very vulnerable to forgetfulness.47 The main problems with delayed self-reports on inner speech are therefore related to memory. All the memory limitations mentioned earlier in connection to verbal reports are applicable to delayed retrospective tools: irretrievability, incompleteness, and inaccuracy. As a result of memory decay, verbal reports may be susceptible to inferencing and creative re-interpretation. Another problem with delayed retrospection is that subjects tend to generalize. As Wenden (1986) put it, a "possibility is that subjects remember what they have done in particular instances, and turn this information into a general procedure" (p. 196). It is important to keep these limitations in mind when analyzing delayed verbal reports. In review, questionnaires, interviews, and diaries have relative merits and shortcomings: • Questionnaires, because they are easy to administer, may offer large amounts of information from multiple participants. This may show tendencies of inner speech behavior across various learner populations. • Diaries and interviews take longer to be obtained and analyzed, yet they have the advantage of offering in-depth information from individual cases. • Face-to-face interviews and to some extent diaries, where contact between researcher and respondent is close, may be embarrassing to the interviewee or diarist. Because people are usually not inclined to discuss their private thoughts with strangers, the truthfulness of responses may be affected in these cases. • An advantage of open-ended questionnaires, unstructured interviews, and nonfocused diaries, in particular, is that they may bring to light important insights into inner speech and the circumstances surrounding it not anticipated by the researcher. • In general, entries in a diary, especially when written soon after the inner speech event, have more recency of recall than those obtained through a questionnaire. • In contrast to class-related verbal reports, diaries taken out of the classroom may yield spontaneous data on inner speech events not-related to the class. • When assigned over an extended period of time, diaries may yield a great deal of information on individual subjects; this could help document inner speech development. 47 Korba (1990) has calculated the rate of inner speech as of above 4,000 words per minute. Frawley (1997), however, interestingly points out that metaconscious control, characteristically mediated by language for thought, is particularly slow and deliberate. It seems reasonable to infer, then, that some inner speech processes, particularly those that are automatic and unconscious, must be remarkably fast, while others, such as those operating at the metaconscious level, might be much slower and amenable to selfinspection and recall. On the automaticity of many language processes and their unavailability for selfreporting, see also Pressley and Afflerbach (1995).

108

CHAPTER 4

•

Responses to a diary are self-selected; in other words, the diarist is the only person who has control over what is reported and the frequency of the reporting. Diaries are therefore susceptible to much idiosyncratic reporting, not amenable to generalization.48 Finally, first-person narratives of advanced bilinguals, such as the ones researched by Pavlenko and Lantolf (2000), are quintessential^ delayed retrospection and therefore vulnerable to all the problems associated to memory and interpretation of past events, especially as it pertains to elusive inner speech processes. What is interesting about these reports is that what may seem their greatest drawback-self-interpretation of long past mental events-constitutes their strongest value. Because they are written to make sense of and bring coherence to past traumatic events resulting in inner language upheaval, first-person tellings throw light on the often deliberate, subjective process by which people choose to define themselves on the basis of their inner language experiences, a process that is instrumental in determining the subjects' attitudes toward the L2 and in shaping their linguistic identities. Furthermore, as Pavlenko and Lantolf (2000) point out, personal narratives have ecological validity, inasmuch as they represent attempts to grasp the social and environmental forces that influence subjective experience. LABORATORY TOOLS IN THE STUDY OF INNER SPEECH Technology has had great impact on the methodology of research on inner speech. The influence of technology has been most noticeable in the study of the psychophysiology of inner speech, which examines the relationship between physiological events (such as muscle tension in the speech, arms, and eye musculature; brain activity; blood flow; heart rate; respiration depth; and skin responses) and covert language behavior, as manifested, for example, during silent reading or writing, memorization, and word retrieval.49 Various laboratory tools and techniques have been employed in this type of research; foremost among them are speech interference, electrophysiological measures, and neuroimaging. Speech (Articulatory) Interference Some laboratory methods measuring the psychophysiology of covert verbal behavior consist of relatively simple articulatory interference procedures. Speech interference involves the simultaneous performance by the subject of some silent intellectual activity (listening, reading, writing, or solving mathematical or concrete problems) and an interfering verbal or auditory operation. Interference techniques include (a)

48

In Guerrero's (2004) diary study (N = 16), more than half of one learner's reported inner speech occurrences (16 out 28) fell under the category of "spontaneous recall" of L2 words. 49 See McGuigan (1978) for a fine-grained classification of psychophysiologically measured covert processes. McGuigan distinguishes between covert responses that pertain exclusively to the glands and muscles and neurophysiological processes that are strictly neural events.

METHODOLOGY OF RESEARCH ON INNER SPEECH

109

inhibiting internal articulation by simultaneously counting numbers aloud, reciting well-known poetry lines, or repeating extraneous syllables, such as ba-ba-ba; (b) reinforcing articulation by means of obligatory audible enunciation of the experimental material; (c) mechanical retardation of articulation by clamping the tongue or the lips between the teeth or opening the mouth as wide as possible; and (c) other tampering with articulation, such as tapping the hand at a certain steady rhythm (see Sokolov, 1972, for a thorough review of these procedures). Frequently, speech interference experiments are complemented by immediate retrospection. Immediately after the task is completed, subjects are asked to verbalize their mental efforts in carrying out the task under the articulation condition. The general effect of these speech interference techniques is a disruption of the inner speech processes of covert speech articulation and a consequent retardation or impairment of the mental operations being simultaneously conducted. Early Mechanical Devices Early twentieth century attempts at studying covert verbal behavior relied on mechanical sensing devices (see McGuigan 1970, 1978, for a full description). Most of these techniques consisted of placing some type of sensor in the vicinity of the vocal organs: a tambour on the larynx, a flattened glass piece at the end of the tongue, a small rubber balloon on the top of the tongue, or a block in the mouth attached to a rubber tube directing air into a tambour. The purpose of these devices was to register the extent of covert articulation during performance of mental tasks: reading or writing silently; solving arithmetic problems; silently reciting, etc. Covert responses were usually recorded on a kymograph, a device in which a stylus graphically traced changes in motion or pressure during articulation. In this way, responses in a variety of body parts associated with speech production, such as lips, mouth, larynx, chest, etc., were recorded. Contrived and primitive as they may seem, these early sensing contraptions gave the first indications of increased inner speech behavior during cognitive activity. Electrophysiological Techniques More recent psychophysiological techniques employed in the study of inner speech activity involve the electrical recording of some physiological event. Following are the main techniques and what they measure: electroencephalography (EEG): brain waves electrocardiography (EKG): heart or cardiovascular activity electromyography (EMG): muscle potentials and contractions electrooculography (EEO): eye movement and ocular activity galvanic skin response (GSR): skin resistance to electrical activity EMG, that is, the measurement of electrical activity a muscle generates when it contracts, has been by far the most useful and productive of all psychophysiological techniques. During this procedure, electrodes (surface, needle, or hooked wire) are positioned or inserted in the relevant muscles. Signals from the electrodes are

110

CHAPTER 4

graphically recorded allowing the observation of response amplitudes. Silent tasks the subject is engaged in as the electromyogram is being performed include counting silently, imagining telling someone the date, recalling a poem, multiplying certain numbers, taking dictation, and reading to oneself. Frequently, simultaneous EMG recordings of various body parts (chin, tongue, preferred arm) as well as EEG measurements in motor areas of the brain are performed. EMG has been widely used to study the covert articulatory processes associated with verbal tasks such as reading, word recall, word formation, language comprehension, and others (Korba, 1989, p. 229). According to Korba (1989), the research provides mounting "evidence that inner speech, or covert linguistic behavior, manifests itself most graphically through the implicit movements of the peripheral effector speech organs, the tongue, and the lips" (p. 231). Sokolov (1972) reported extensive research conducted by himself and others exploring the physiological correlates of concealed articulation-inner speech-by means of EMG. Electromyograms of concealed articulation were obtained during mental arithmetic, reading to oneself, listening to speech, mental reproduction and recollection of verbal material, and manipulation of graphic-visual material. In many of his experiments, Sokolov complemented EMG with verbal reporting and other physiological indices such as GSR and EEG. Figure 4-1 illustrates results of EMG recordings in one of Sokolov's studies of inner speech during silent reading. The graphic clearly indicates the increment in concealed articulation that occurred as difficulty increased when subjects read an FL text.

Phrasqs

1

Phrases ^ C

^

^

f ^

^

if

tf

^

Iff

g

-

„

fr

^

*

-

W

^

Figure 4-1. Electromyographic recordings of tongue movements during silent reading of (a) a text in Russian, the subject's native language; (b) an adapted text in English, the subject's foreign language; and (c)a difficult English text. From Sokolov (1972), p. 177. Copyright 1972 by Plenum Press. Reprinted with permission.

Covert linguistic behavior has been widely studied through electromyography (see McGuigan, 1978, for a review of early studies). Williams (1987), for example,

METHODOLOGY OF RESEARCH ON INNER SPEECH

111

explored electromyographic differences in subvocal speech between above-average and below-average writers as they were involved in a writing task. Korba (1990) utilized EMG to ascertain the rate of inner speech. In his study, subjects provided self-reports of the elliptical inner speech they had used in solving verbal problems (elliptical word count) and then expanded their self-reports into fully extended speech (extended word count). When extended, inner speech was found to have an equivalent rate of over 4,000 words per minute. Electromyographic measurements of the subjects' subvocalizations during problem solving were then correlated with elliptical word count. More recently, Livesay, Liebke, Samaras, and Stanley (1996) tested the prediction that covert speech behavior measured electromyographically from the lips is significantly more prominent during a brief silent-language recitation task than during a brief nonlanguage visualization task. Advantages and Limitations of Mechanical, Speech Interference, and Electrophysiological Techniques and Their Applicability in the Study ofInner Speech By focusing on the covert, but observable and recordable, processes of inner speech at the stages of articulation or motor activity, the psychophysiological techniques described above offer what might be considered "objective" measurements of inner speech. In particular, recordings of the electrical activity of the speech and other skeletal musculature registered by psychophysiological procedures constitute clear, almost tangible, indicators of hidden speech processes. Psychophysiological recordings provide information about inner speech activity as it occurs in real time and in association with specific cognitive operations. One advantage then is the possibility of obtaining precise details of some inner speech features such as speed, intensity, and duration as well as clear evidence of correlations with stimuli provoking covert linguistic activity, type of mental task, task difficulty, and performance. Perhaps the most important advantage, from a theoretical point of view, is that psychophysiological and verbal interference techniques tend to blur the boundaries between mind and body as they reveal the material, bodily underpinnings of thought (Korba, 1989). The application of psychophysiological techniques, however, requires high expertise and costly electronic equipment. As Behnke (1989) warns, researchers interested in applying psychophysiological techniques to inner speech research must be cognizant of (a) the nature of physiological responses, (b) the appropriate instrumentation to measure such responses, and (c) the methodology of measurement, quantification, and analysis of physiological responses. Laboratory tools, in addition, involve devices and procedures that are quite obtrusive and "unnatural"(extremely so in some cases, as has been seen!) The methodology that goes with these tools is, furthermore, almost always experimental. The elicitation of inner speech through these methods may thus be considered non-spontaneous or lacking "ecological validity." The effects of artificiality in the laboratory study of inner speech could be ameliorated, however, by complementation with more naturalistic methods of observation. Eliciting verbal reports, for example, to go along with psychophysiological methods, is not only desirable but necessary in some cases. As Korba (1990) put it, "it appears inevitable

112

CHAPTER 4

that any corroboration of physiological measures would require subjective introspection" (p. 233). His study of inner speech rate certainly called for the collaboration of EMG measures with subjects' verbalizations. Also, in some of Sokolov' s experiments, verbal reports from subjects were essential to ascertain the type of subjective experience (whether visual, auditory, verbal, or mixed) they were having at the time of the speech interference or psychophysiological condition and to find out effects on understanding or remembering. Neuroim aging Recent advances in technology have made it possible for researchers to investigate the neurophysiology of intellectual activity beyond measurements at the scalp level (EEG). With the advent of computed tomography (CT) in the 1970's, a new era of brain research was born, one that would lead to extraordinary findings about the neural underpinnings of human behavior (Posner & Raichle, 1994). Two developments in brain scanning techniques, in particular, would become instrumental in providing images of inner speech activity in the brain: positron emission tomography (PET) and magnetic resonance imaging (MRI). These two techniques, complemented with other technological tools, such as the measurement of blood flow in the brain and EEG, allow for the study of the anatomical as well as the functional aspects of inner speech.50 Positron Emission Tomography PET scans are readings of the subject's blood flow in the brain.51 Subjects undergoing a PET scan are injected a safe, radioactive material (tracer) that, when flowing in the brain, shows areas of greater activity. During a PET scan, subjects are required to perform a specific cognitive task. The assumption underlying PET is that blood flow will be greater in those regions that are more active during task performance. The radioactive substance emits positrons that can be detected by a scanner, a ring-shaped apparatus positioned around the subject's head. Positron emissions are converted into multicolored, two or three-dimensional computer images (tomographs). Because the scanner is equipped with multiple detectors, several images of the brain-"slices"-can be obtained simultaneously during task performance. Control images are also obtained of the subject in a "resting" or baseline condition that is assumed not to engage the same neural areas of the task under investigation. Control images are then subtracted from the images in the task condition in order to identify those regions of the brain more actively engaged in performing the mental operation. Subsequently, and to eliminate idiosyncratic differences in images from individual subjects, multiple

50 The following have been the main sources consulted for the discussion on brain scanning and imaging techniques: Frahm, Fransson, & Kriiger, 1999; Mathias, 1996; Mazziotta, 1994; Posner & Raichle, 1994; Raichle, 1994, 1997. 51 PET has also been used to measure blood volume, oxygen and glucose metabolism, drug concentrations, and other physiological and chemical responses (Frahm et al., 1999; Joseph, Noble, & Eden, 2001; Mathias, 1996; Raichle, 1994).

METHODOLOGY OF RESEARCH ON INNER SPEECH

113

subtraction images from various subjects are averaged. This procedure is known as image averaging. In addition, because head shape and size vary among individuals and to make comparison across studies easier, subtraction images are fitted into standard "atlas" brain shapes based on images from normal subjects. PET is a safe technique that can offer valuable structural, physiological, and neurochemical information on the working brain. PET studies are useful not only in locating those regions of the brain that are related to specific mental operations but also in determining the way the brain functions when engaged in mental activity. A large number of studies have used PET to measure anatomical structure and brain functioning during the processing of language tasks, such as in word perception (listening and reading), word rehearsal, speech production, and semantic access. Some PET studies have been conducted showing impressive effects of practice on brain functioning (Posner & Raichle, 1994). The more automatic a task becomes as a result of practice, the less active involvement of the brain in performing the task. One of the automatic functions of the brain that has been investigated is reading. PET studies have confirmed the hypothesis of the existence of two pathways in the brain, one for automatic responses and another one for the processing of novel tasks.

Figure 4-2. Image shows inner speech areas (in the left brain hemisphere) associated with the articulatory loop. PET scan was performed as subjects silently rehearsed visually presented letters. Lighter inner spot indicates increased cerebral blood flow in the left supramarginal gyrus (related to the phonological store) and Broca 's area (related to subvocal rehearsal). From Paulesu et al. (1993,p. 342). Copyright 1993 by Nature Publishing Group. Reprinted with permission from Nature Publishing Group and Dr. Eraldo Paulesu.

PET has been used to study inner speech in the Llin various conditions, e. g.: silent reading (as compared to oral reading) among stuttering and nonstuttering adults (De Nil, Kroll, & Kapur, 2000), auditory word presentation and verb generation (Fiez, Raichle, Balota, Tallal, & Petersen, 1996), auditory verbal imagery (McGuire et al., 1996), counting silently (versus aloud) Ryding et al., (1996), and subvocal rehearsal (Paulesu et al., 1993, see Figure 4-2). In the L2, PET has been utilized, for instance, to reveal areas of activation during translation and language switching (Price et al,

114

CHAPTER 4

1999), L2 word generation (Klein et al, 1995), and L2 listening (Perani et al., 1998). Magnetic Resonance Imaging Unlike PET, which utilizes injected radioactive tracers, MRI measures changes in a magnetic field created by a large cylindrical magnet around the subject's head. Radio waves are sent through the magnetic field and return signals from the brain are read and computed into an image. MRI produces high-quality images that can identify areas of activation in the brain with great precision. Lately, an important development in this technology, functional MRI (fMRI), has made it possible to detect changes in the brain as it functions. fMRI exploits the magnetic properties of blood to create images of online blood flow in the brain. This enables the measuring and observation of minute changes in brain activity while subjects are performing various mental tasks. In addition to being less invasive than PET, fMRI has the advantage of being faster in producing images. Thus, it is possible for fMRI to determine "when brain regions become active and how long they remain active. As a result, [researchers] can see whether brain activity occurs simultaneously or sequentially in different brain regions as a patient thinks, feels, or reacts to experimental conditions" (Mathias, 1996). In other words, through fMRI, one can visualize the processes the brain undergoes as it performs various functions. (See Figure 4-3.) Functional MRI has been widely used in studies of inner speech from an LI perspective, e.g., to explore hemisphere language dominance during noun generation (Baciu, Rubin, Decorps, & Segebarth, 1999), areas of activation during silent word generation (Friedman et al., 1998) as well as word encoding and retrieval (McDermott et al., 1999), and the functional anatomy of inner speech and different forms of auditory verbal imagery (Shergill, et al., 2001). From an L2 perspective, various studies have also used fMRI, e.g., to investigate how bilingual subjects can turn off successfully one language while processing another one (Rodriguez-Fornells et al., 2002), to determine areas of activation of LI and L2 sentence generation (Kim et al., 1997), and to find out areas of activation during LI and L2 processing among early bilinguals (Chee, Caplan, etal., 1999; Chee etal.(2003). A new development that makes possible the study of the brain in real time is magneto-encephalography (MEG). MEG combines MRI with EEG. As brain neurons become active during task performance, they produce not only electricity that is detected by EEG but also tiny magnetic signals that can be externally captured by MRI technology. MEG produces high-quality functional imaging as well as real time recordings of activity through EEG. MEG is ideally suited to examine the anatomical sequencing and patterning of brain functioning; in other words, how mental operations proceed and what paths they follow over time. Advantages and Limitations ofNeuroim aging and Its Applicability in the Study of Inner Speech Brain scanning and imaging techniques provide, to date, the most "direct" means of observation of inner speech processes. One of the greatest assets of neuroimaging is

METHODOLOGY OF RESEARCH ON INNER SPEECH

115

the rendering of remarkable visual records of inner speech activity in the brain. Through these images it is possible to observe not only what brain structures are implicated in verbal thinking but how the brain functions as subjects engage in inner speech activity. Particularly valuable among the neuroimaging techniques are those that allow for the observation of inner speech stages that are inaccessible through introspection: Where does inner speech arise? What course does it follow? What are the regions involved in inner speech as it becomes externalized speech? Studies in this area, still in its infancy, will throw much light on the hypothesized phases of speech reception and production. The application of findings from brain imaging studies to infant development (Posner & Raichle, 1994, p. 232) might also yield insights into inner speech development among children.

Figure 4-3. Images (obtained through fMRI) of selected axial slices showing brain areas activated during inner speech. Dark spots are areas significantly activated (left side of figure corresponds to right side of subjects) . Top row shows inner speech during silent articulation of sentences in the form "I like . . . ." Bottom row shows inner speech while subjects imagined the same type of sentence as if spoken in their own voice (firstperson inner speech). From Shergill et al. (2001), p. 245. Copyright 2001 by Cambridge University Press. Reprinted with permission from Cambridge University Press and Dr. Sukhi Shergill.

For all its advantages, neuroimaging is not without its limitations. Foremost among these is the elevated cost of the technology involved and the high degree of expertise in utilizing it. This fact alone makes the exploration of inner speech through neuroimaging restricted to only a few specialized and properly equipped laboratories. Obviously, then, the fact that neuroimaging research can only be conducted in highly mechanized and controlled laboratory conditions precludes any claims to ecological validity. Other limitations pertain to current problems in the implementation of these

116

CHAPTER 4

procedures, no doubt to be corrected with future refinement. Bogen (2002), for example, has raised concern about some of the usual practices of neuroimaging, such as fitting brain images into standard atlases, intra-subject averaging to reduce "noise", and inter-subject averaging to eliminate idiosyncratic differences. These practices, in Bogen's view, render images that are idealized, rather than real, pictures of brain anatomy. A third area of limitation concerns the internal validity of neuroimaging research: Do the images obtained really reflect the internal operation the study purports to investigate? Images of the brain show areas of activation in "presumed" conditions of verbal cognitive operations. It is very difficult, however, to ascertain beyond reasonable doubt that what the person is mentally doing while his/her brain is being scanned is precisely the cognitive operation called forth in the experiment. Although there are solutions that could attenuate this problem (having the subject exhibit some external sign of inward mental activity, such as pressing a Yes or No button, etc.), the possibility always exists of making unfounded claims on the relationship between what is seen on the image and the underlying inner speech processes. Despite its limitations, neuroimaging emerges as an extremely useful tool in corroborating observations gained through other means, clarifying areas of contention, raising further hypotheses, and discovering unknown aspects of inner speech. Specifically, brain imaging may help understand normal and abnormal functioning of inner speech, providing fine-grained information of inner speech processes in cases of deafness, blindness, dyslexia, aphasia, autism, and schizophrenia. Connections can be made between neurological and anatomical brain functioning and theoretical models of inner speech activity. Some working memory notions such as inner ear, inner voice, and phonological loop can be put to the test (see discussion in Chapter 2). And, as Luria and Sokolov did in the past with less sophisticated instruments, certain sociocultural theory constructs related to inner speech, such as internalization, externalization, and inner stages of speech production, can also be inspected under the light of the new technologies. Finally, as some of the research has begun to show (see Chapter 3), brain areas and pathways of activation can be observed during inner speech activity involving LI and L2 tasks, monolingual and bilingual/multilingual speakers, early and late bilinguals, and the like. THE PROBLEM OF METHOD IN THE STUDY OF INNER SPEECH WHAT TO DO? As has been seen in this chapter, the fact that inner speech constitutes one of the most difficult problems to investigate has not refrained researchers from making a try at bringing it down to the level of systematic and empirical perusal. Boldly and creatively, researchers have risen to the challenge of dealing with a phenomenon that is not only extremely complex and multifaceted but also slippery and evanescent. To date, various methodological approaches reflecting diverse theoretical orientations have been attempted, each with its own sets of goals, premises, tools, and procedures, each with its own strengths and weaknesses.

METHODOLOGY OF RESEARCH ON INNER SPEECH

117

No doubt the scientific approach to inner speech owes much to Vygotsky, who was the first in stressing the necessity of dealing with the problem of inner speech for a true understanding the relationship between thinking and speech and in proposing a principled and systematic methodology for its exploration. Lacking the instruments that today enable a more "direct" look at inner speech, Vygotsky resorted to the best objective solution he could avail himself of at that time: the genetic method, a method that was thoroughly in agreement with his sociohistorical approach to mental development. Turned into the study of private speech (or private writing), Vygotsky's genetic method has yielded great fruits and continues to inspire much investigation.52 But Vygotsky's method was not to remain the only one. Cognitive psychology, with its focus on internal mental processes and awareness of the strengths and pitfalls of memory, would bring a renewed and legitimized use of verbal reports to inner speech research. At the same time, sensitiveness to social and cultural factors that might affect the "ecology" of inner speech would call for methods that take the subjects' insider perspective into account. Finally, extraordinary advances in technology, specifically in psychophysiology and the neurosciences, would contribute with sophisticated tools to make the problem of inner speech more accessible to direct observation and investigation. As with most complex phenomena, inner speech has many "faces"-the philosophical, the biological, the psychological, the linguistic-and cannot be fully appreciated or understood from only one methodological perspective. For that reason, the best approach to the comprehensive study of inner speech is one that takes advantage of the tools and procedures made available by the different methodologies. A combination of methods (methodological triangulation) could exploit the best features and tools of each methodology. While a multimodal approach to the problem may be the best solution, it is a daunting task for single researchers competent in one methodological modality but untrained in and lacking the instruments of the other fields. Collaboration among researchers from different areas of expertise, however, could make the enterprise a lot easier. Data could then be pooled together for a composite picture. In short, triangulation and team work (Dance, 1994; Stacks & Sellers, 1994) could lead to great advances in the exploration of inner speech. A multimodal approach could still rest on general sociocultural theory principles. If inner speech is a psycho-social phenomenon with both a micro- and a macro-genesis and both internal and external manifestations, different methods can be introduced to explore the various facets of inner speech, in the LI as well as in the L2. The study of private speech, for example, can provide important insights into the ontogenesis and microgenesis of inner speech. It can go deep into the issues of internalization and extemalization as well as suggest hypotheses concerning the structure and function of inner speech. Verbal reports, in turn, can provide the subjective experience of inner

52

Unfortunately, although private speech has been extensively studied, it has mostly remained an object of research in itself and only a few connections between private and inner speech have been made. It is hoped that a recent focus on the internalization aspects of private speech in the L2 will contribute to illuminate the topic of inner speech.

118

CHAPTER 4

speech. Verbal reports offer rich data on people's preferences and tendencies towards inner speech and about the role that inner speech plays in their lives as thinking, affective beings immersed in particular linguistic, social, and cultural environments. Macro-development of inner speech can be traced through verbal reports in the form of questionnaires administered cross-sectionally to individuals, as diaries assigned longitudinally, or by way of autobiographical narratives from individuals who have crossed linguistic boundaries in their lives. Verbal reports, however, are severely limited in that they do not have access to unattended inner speech or to a host of nonconscious microprocesses associated with the phenomenon. Subjects' verbalizations therefore should be approached with caution and used judiciously when appropriate. Lastly, a comprehensive approach to inner speech should not dispense with the remarkable tools and techniques of psychophysiology and neuroimaging. These are essential to dispel any illusion about the immateriality of inner speech. They are also fundamental in revealing, graphically and visually, the crucial role inner speech plays as mediator of human intellectual functions. The psychophysiology of inner speech plays a critical role in demonstrating the extent to which the mind is an embodied phenomenon. Neuroimaging, on its part, offers the most exquisite evidence to date of the anatomical and functional presence of inner speech in the brain. Both psychophysiology and neuroimaging lend themselves to the inspection of microgenetic processes of inner speech that are not accessible to introspection or external observation. Brain scanning and imaging techniques, in particular, may one day be unobtrusive enough to allow for the study of inner speech in ontogenesis among children. In the enthusiasm that such big scientific steps might bring, however, one should not lose sight that these techniques are simply tools for an understanding of a very complex, human, socio-cognitive phenomenon, which has its roots in social activity and is, in and of itself, the most effective mechanism for the construction of the uniquely human mind. Unless properly framed within a larger humanistic and sociocultural perspective, these tools run the risk of reducing inner speech to a collection of electrical and chemical reactions in our brains and other body parts or to mere glowing bursts of energy on a glossy picture. CONCLUSION This chapter has delved into one of the most intricate and perplexing issues related to inner speech: how to deal with it methodologically. A review of the most prominent techniques and approaches has been produced, starting with Vygotsky's genetic method and continuing with its application in the study of private speech (and private writing). The implementation of verbal reports and the employment of various laboratory tools, such as psychophysiological measures and neuroimaging, was then described. Each method has been appraised for its advantages and limitations and its appropriateness in examining different aspects of the phenomenon. A final call for a multimodal methodological approach to a comprehensive exploration of inner speech has been made, one that fruitfully exploits the strengths and values of each technique.

CHAPTER 5 L2 INNER SPEECH: WHAT LEARNERS SAY

In this chapter, the aim is to go in depth into the author's data on L2 inner speech. Because these data are learners' verbal reports, the chapter opens a window into how L2 inner speech is perceived from the learners' perspective. The chapter is divided into three sections. The first pools together the results of two related studies on the nature of the inner speech involved in mental rehearsal of the L2 among learners at various levels of proficiency, the second presents the results of a separate study on the early stages of L2 inner speech development, and the third focuses on the strengths and weaknesses of the verbal report methodology utilized in these studies. The following questions are probed: (a) To what extent do L2 learners on various proficiency levels experience inner speech in the L2? (b) What are the structural characteristics of the L2 inner speech experienced by L2 learners on various proficiency levels? (c) What functions does inner speech play as L2 learners mentally rehearse in the L2, and how do these functions compare among the various proficiency levels? (d) To what extent is L2 inner speech class-related among L2 learners? (e) To what extent do L2 learners rely on their Lias they engage in L2 inner speech? (f) How do L2 learners begin to develop their ability to "think words" in the L2? (g) What are the pros and cons of a verbal report methodology in the study of L2 inner speech? Questions (a) through (e) will be dealt with in Section 1 of this chapter, question (f) will be the focus of Section 2, and question (g) will be discussed in Section 3. SECTION 1. INNER SPEECH AND MENTAL REHEARSAL OF THE L2 As mentioned in previous chapters, one of the functions that inner speech serves is that of rehearsal. Vygotsky (1986) had made reference to this function as he noted the role played by inner speech in preparing for external speech or planning what to say (pp. 88, 234). Psychologists acknowledge mental rehearsal as an important inner speech behavior leading to memorization, speech production, learning, and the monitoring of consciousness or metacognition (see, for example, Flavell, 1977; Gathercole &

120

CHAPTER 5

Baddeley, 1993; Houston, 1986; Clark, 1998, Smith, 1983). In the L2 literature, mental rehearsal has been identified as one of many successful language learning strategies under the categories of repetition (Chamot, 1987; O'Malley et al., 1985), advance preparation (Chamot, 19S7), production (TSLYOYIQ, 1983), andpractice (Rubin, 1987). Mental rehearsal has also been associated with such L2 phenomena as the Din (Guerrero, 1987; Krashen, 1983), spontaneous playback of the L2 (Bedford, 1985), and language play (Broner & Tarone, 2001; Lantolf, 1997). Mental rehearsal of the L2, in short, represents covert language behavior that learners engage in with the purpose of practicing or learning the L2. Although in this chapter mental rehearsal will be taken as a manifestation of inner speech because of its internal, self-directed nature, it should be kept in mind that, developmentally, mental rehearsal of the L2 is only a rudimentary form of inner speech, which learners engage in before they can operate on the basis of fully internalized L2 social speech. Two studies were conducted by this author to investigate inner speech as mental rehearsal of the L2. One was the 1990/1991 dissertation later published as Guerrero (1994) and the other was the study published as Guerrero (1999). In these two studies, the researcher's purpose was not only to explore in depth the nature of the inner speech used for rehearsal purposes by L2 students but also to cast light on the question of whether mental rehearsal disappears or wanes with proficiency, a question on which there was somewhat contradictory evidence in the literature (Bedford, 1985; Guerrero, 1987; Lantolf, 1997; Parr & Krashen, 1986). Guerrero's studies narrowed the investigation of L2 inner speech to its relationship with mental rehearsal; they did not, for example, deal with L2 inner speech as a problem-solving tool or as a mechanism involved in listening or reading comprehension. Rather, the studies sought to investigate the nature, that is, the structural characteristics-in terms of syntax, phonology, and meaning-and functions of inner speech during mental rehearsal of the L2. Following is a synthesis of the two studies. Participants The data for the studies were provided by 472 Spanish-speaking ESL students at a large private university in Puerto Rico. The students were selected on the basis of their scores on the ESLAT (the College Board's English as a Second Language Achievement Test), required for admission to the university. Within a 200-800 range of possible scores, the students fell under four different levels:53 Proficiency level: ESLAT scores: Number of participants:

Low 192 73 46

53 For comparison purposes, the four levels described here-low, intermediate, high, and advanced-would roughly correspond to the four proficiency levels established by the American Council on the Teaching of Foreign Languages (ACTFL): novice, intermediate, advanced, and superior (Omaggio-Hadley, 2001).

L2 INNER SPEECH: WHAT LEARNERS SAY

121

The low proficiency students entered the university having little command of English vocabulary and structures and only minimal ability to communicate in the English language. Most of these students, it should be noted, had been exposed to some English instruction in their elementary and secondary school and therefore could not be considered real beginners. These students were placed in an introductory ESL course that would provide them with intensive practice in the essential elements of the English language, including grammar, vocabulary, and communicative skills. Students on the intermediate level had some knowledge of basic English structures and vocabulary, could follow and engage in simple conversations, and could read and write in English in a limited way. The high proficiency students were considerably fluent in speaking and had a well developed vocabulary but still needed refinement in terms of grammar, reading, and writing skills. Students on the intermediate and high proficiency levels were assigned courses that aimed at the progressive development of their communicative skills, grammar knowledge, and vocabulary. The students on the advanced level were highly competent in all of these areas despite their background as L2 learners. They were therefore placed in English courses where essay writing, doing research projects, and literature were taught. Because of their high command of both Spanish and English, most students at this level could be considered fully bilingual. The participants belonging to the low, intermediate, and high proficiency levels were selected following the proportionally stratified sampling technique. This procedure consisted of calculating the sample that was needed to represent proportionally the population of students enrolled in each level. Students on the advanced level were selected on an availability basis. To ensure consistency in the sampling process, all students had to comply with two essential requirements to be considered participants in the studies: (a) having a verifiable ESLAT score and (b) declaring English their L2 and Spanish their LI. Data Collection and Analysis A questionnaire was constructed to measure students' responses on a variety of aspects related to their inner speech and mental rehearsal of the L2. This questionnaire was administered during the first two weeks of class to proficiency groups low, intermediate, and high (Guerrero, 1994). A revised version of the questionnaire, where a few questions were added and others deleted, was administered to the advanced group (Guerrero, 1999). Both versions contained space for the students to write comments. The data presented here draw from those items in the questionnaire that were common to both the original and revised versions. A subset of questions on language play and the affective function of inner speech that was added to the revised questionnaire in Guerrero (1999) will be discussed only as it pertains to the advanced level. To ensure understanding of the phenomenon under investigation, the introduction to the questionnaire broadly defined inner speech as "any type of language in English that occurs in your mind and that is not spoken. Inner speech may include sounds, words, phrases, sentences, dialogues, and even conversations in English." The

122

CHAPTER 5

introduction also offered a composite definition of mental rehearsal based on various descriptions in the literature: "Mental rehearsal is a voluntary or involuntary activity by means of which students practice in their minds the language they have learned, heard, or read, or the language they will have to use in a future oral or written activity. When mentally rehearsing, the students may simply be recalling, repeating, or imitating words in the second language. Sometimes, mental rehearsal is more creative, as, for example, when the students imagine dialogues, plan what they are going to say or write, mentally self-correct, evaluate other students' language, or engage in conversations with themselves." Students on the low, intermediate, and high levels received a Spanish version of the questionnaire whereas students on the advanced level were given the choice of taking the questionnaire in Spanish or English.54 In addition to the questionnaire, interviews were conducted with 9 randomly selected "high rehearsers," that is, participants who had shown a tendency to mentally rehearse in English as evidenced by their responses to certain key items in the questionnaire (see Guerrero, 1994). The students were interviewed on two occasions, each one a day after the students had participated in a communicative activity in the classroom for which they had to prepare with a week in advance. The interviews followed a semi-structured format that included some fixed questions on the nature of the inner speech associated with the activities but that allowed for the insertion of other probes as was deemed necessary (see Guerrero, 1994). Interviews were conducted in Spanish, except on occasions when students gave examples of their inner speech in English. Interview and questionnaire comments have been translated here to English, with words used in English indicated in boldface. Responses to the questionnaire were given on a 5-point Likert scale based on the following frequencies: never, almost never, sometimes, often, always. For the quantitative analysis, responses were given numerical values ranging from 0 {never) to 4 {always) and converted to a No/Yes distribution, where No included all the never and almost never responses and Yes all the sometimes, often, and always responses. Descriptive statistics (mean, median, mode) were applied and the Chi-square test was used to identify significant differences among proficiency levels. Content analysis was performed on the qualitative data supplied by the interviews and questionnaire comments. Results and Discussion To what extent do L2 learners on various proficiency levels experience inner speech in theL2? The first item in the questionnaire-"Have you had inner speech in English?"-examined the extent to which the participants had experienced inner speech in the L2. A significant majority of learners in each group answered this item affirmatively (for chi54 Questions from the questionnaire will be quoted here as they are discussed. For complete versions of the questionnaire, see Guerrero (1990/1991, 1994, 1999).

L2 INNER SPEECH: WHAT LEARNERS SAY

123

square values, see Guerrero 1994, 1999). Figure 5-1 provides the percentages of Yes {sometimes to always) responses in each group. As can be seen, the frequency of reported inner speech increased with the proficiency level, from 75% in the lowest level to 98% in the highest. These figures suggest that inner speech in the L2 was a widespread phenomenon for the participants and that it increased with proficiency. Empirical evidence in the form of verbal report data was thus established for the occurrence of L2 inner speech among ESL college learners and for the notion that L2 inner speech is a developmental phenomenon, much as Vygotsky (1986) conceived of LI inner speech. LOW

INTERMEDIATE

HIGH

ADVANCED

75%

89%

90%

98%

Figure 5-1. Percentages of affirmative answers to the question "Have you had inner speech in English? " by learners on various proficiency levels.

What are the structural characteristics (as regards syntactic complexity, phonology, and meaning) of the L2 inner speech experienced by L2 learners on various proficiency levels? Four items in the questionnaire tested the complexity of inner speech in terms of the extent to which it consisted of words, phrases, sentences, and conversations/dialogues (see Table 5-1). An analysis of these items revealed a tendency towards syntactic abbreviation and simplification, as had been proposed for L1 inner speech by a number of researchers (Korba, 1989; Sokolov, 1972; Vygotsky, 1986). This finding is suggested by the fact that there were fewer affirmative responses as the structural complexity of inner speech increased. Table 5-1. Percentages of Yes responses to items on structural complexity ofL2 inner speech by proficiency level Item

Low

Intermediate

High

Advanced

Is your inner speech made up of words?

84

88

85

89

. . . of phrases?

71

83

88

87

. . . of sentences?

58

75

81

85

. . . of conversations or dialogues?

35

52

59

72

124

CHAPTER 5

When the four proficiency levels were compared, important differences among the samples were found. In general, the data indicated a tendency for more complex structures to be more frequent as proficiency increased. The frequency of the wordtype structure, however, showed no significant difference among the samples, indicating that the most abbreviated form of inner speech was common to all the participants regardless of their proficiency level. The data thus indicate that, although more proficient learners are capable of more complex and elaborate inner speech structures than lower learners, words and phrases continue to be more frequent than more complex structures. This finding supports Vygotsky's (1986) hypothesis that the predominant structural characteristic of inner speech is its tendency toward reduction or abbreviation. One item, "Can you 'hear' the sounds of English in your mind?," provided data for the analysis of the phonological aspect of inner speech. Eighty-eight percent of the participants answered this question affirmatively, with figures ranging from 80% in the low proficiency level to 91% in the advanced. All the subjects in the post-activity interviews also reported "hearing" their own voice or that of another learner as they mentally recapitulated after the activities. Inner speech in the L2 thus, like LI inner speech, appears to be sonorous, rather than soundless. Vygotsky (1986) had described inner speech as being soundless or silent speech. Sokolov (1972), however, had pointed out that inner speech is soundless from the point of view of an outsider but not for the person experiencing it, an effect that is possibly due to the fact that in inner speech auditory memory is activated while motor speech production is inhibited (Hardyck & Petrinovich, 1970; MacKay, 1992; Shergill et al., 2001). A discrepancy, however, was frequently reported between the L2 that was heard in the mind and the language that was externalized. As one student said in an interview, "In my mind [the L2] is heard more beautiful. It's perfect. It's clearer than when I pronounce it. I feel there is a difference." The possibility that L2 inner speech could be externalized as private speech, that is, as loud speech for the self, was probed through the item "Do you repeat aloud any of the words of that inner speech when you are alone?" Seventy-three percent of the participants in the low, intermediate, and high groups answered this question affirmatively. The frequency with which participants in the advanced level reportedly vocalized their L2 inner speech, however, dropped to 59%. The data thus indicate that whereas overt vocalization or externalization of L2 inner speech does occur, this phenomenon tends to decrease with proficiency. Why would particularly lower level learners students loudly vocalize some of their inner speech? One plausible reason is that students may sometimes need to engage in unfolded phonemic articulation of their inner speech, especially when rehearsing the pronunciation of difficult words. Another reason may be that by outwardly vocalizing certain words, learners can regulate their inner speech, "freeze" (Clark, 1998, p. 178) it, so to say, so that it can become the object of attention and manipulation. (More on the issue private speech is discussed in Chapter 6.) A third item, "Do you hear in your mind voices of other people in English?," contributed to throw light on the phonological aspect of inner speech. This item,

L2 INNER SPEECH: WHAT LEARNERS SAY

125

answered affirmatively by 59% of the total sample, confirmed the possibility that other people's voices in the L2 may be heard in the mind. In fact, both the comments to the questionnaire and the interview reports contained references to the students being able to hear in their minds the voice of their professor, an interlocutor, or someone heard on TV. This has been referred to as the "polyphonic" (Trimbur, 1987) or "heteroglossic" (Bakhtin, 1981) nature of inner speech, which results from the internalization of other people's voices. The meaningfulness of L2 inner speech was tested through the item "When you mentally rehearse, do your thoughts in English make sense?" An extremely large number of participants answered this question affirmatively, with figures ranging from 91% in the low level to 98% in the advanced. The median for this item was often, a finding that suggests that for the majority of the participants their inner speech frequently made sense. Furthermore, this tendency seemed to increase with proficiency. The overall meaningfulness of L2 inner speech, as reported by the participants, should not be surprising given the fact that conscious inner speech activity is usually patterned as a dialogic interaction between the self as "knowing speaker" and the self as "understanding listener." The inner speech of an individual, if it could be somehow recorded, might be incomprehensible to an outside hearer, but for the inner speaker it is transparent because he/she knows what he/she is talking about. The aspect of meaning was complemented by analysis of the item "When you mentally rehearse, do words with meanings you do not know well come to your mind?" Seventy-eight percent of the participants answered this item affirmatively, the median being sometimes. However, the frequency of perceiving words with unfamiliar meanings in the participants' inner speech decreased with proficiency: from an average of 85% in the low, intermediate, and high level groups to 72% in the advanced. The combined results of the two items on meaning indicate that although L2 inner speech may contain words that are strange or incomprehensible, this phenomenon tends to decrease as L2 learners become more proficient in the language. Processing the meaning of difficult, unknown, or polysemous words, however, seems to be an inevitable and necessary aspect of learning of new language at all levels of proficiency. Experiments in reading and translating FL texts have shown that inner speech plays a crucial role in understanding difficult language and that semantic hypotheses about L2 are formed at the level of inner speech (Sokolov, 1972). Similarly, research on the rehearsal mechanism known as phonological loop indicates that its main function is the learning of new words, a process that is essential in FL acquisition (Baddeley et al., 1998; Gathercole & Thorn, 1998). The questionnaire findings related to meaning support the notion that inner speech has a critical role in the process of understanding and making meaning in the L2. Whatfunctions does inner speech play as L2 learners mentally rehearse in the L2 and how do these functions compare among the various proficiency levels? Several specific functions of inner speech, as it occurs during mental rehearsal of the L2, were tested through the questionnaire. These were the mnemonic, instructional,

126

CHAPTER 5

evaluative, preparatory, dialogic, play, and affective functions. Many of these functions had emerged in the process of reviewing the literature and in the pilot study conducted for Guerrero (1990/1991). Additional evidence for these functions was obtained through the interviews and comments to the questionnaire. The Mnemonic Function, Two items in the questionnaire measured the use of inner speech as an aid to memory. The item "When you mentally rehearse, do you repeat words you want to learn?" tested the learners' use of inward repetition to retain language in memory, whereas the item "When you mentally rehearse, do you try to recall words you have learned?" referred to the role of inner speech in retrieving words from memory. The use of repetition as a mnemonic strategy had a very high percentage of Yes responses among the low, intermediate, and high proficiency groups: 96% (and no significant difference among the groups). This function, however, dropped to 83% among the advanced students. Something similar occurred with the use of inner speech as an aid to recall: 98% of affirmative responses for the three lower groups (and no inter-group significant difference) dropping to 80% for the advanced group. These findings agree with the decrease in frequency mentioned earlier in terms of words with unfamiliar meanings coming to the learners' minds as proficiency increased. The considerable reduction in frequency that was observed in terms of the mnemonic role of inner speech among the advanced participants suggests that, with increased knowledge of the language, the need to rehearse new or difficult vocabulary is smaller as fewer words represent a semantic or phonological challenge to the learner. In addition to the questionnaire data, information on the mnemonic function of inner speech was obtained through the interviews. Several mnemonic techniques used to rehearse material for the communicative activities were mentioned by the participants: silent repetition, spontaneous recall, deliberate recall, repeating the text aloud, relating words to a visual image, jotting down key words, and a circular pattern of memorization consisting of iterative attempts to retrieve, repeat, store, and retrieve again language segments. One of the intermediate students, for example, commented: "When I am interested in a word, I think and think about it for five, ten minutes, so that I won't forget it. Within about two hours, the word comes back, and I repeat it until I learn it. That way I will always remember it." Another learner-in the high proficiency level-reported waking up in the middle of the night one day before the communicative activity would take place and spontaneously recalling phrases she could use. These results are consistent with the information-processing view of rehearsal as a crucial mechanism of memory (see Chapter 2). Studies within this paradigm have identified rehearsal as a short-term mechanism aiding retention in long-term memory. As discussed before, this mechanism is instrumental in language learning and particularly important in the acquisition of FL vocabulary (Baddeley et al., 1998; Gathercole & Baddeley, 1993; Gathercole & Thorn, 1998). Along with this type of research, Ellis and Sinclair (1996) found that "short-term repetition of FL utterances allows consolidation of long-term representation of words and sequences" (p. 246).

L2 INNER SPEECH: WHAT LEARNERS SAY

127

Service (1992) argued that words that sound unfamiliar are more difficult to keep in short-term memory and that "a number of rehearsal cycles might be necessary to establish an association between form and meaning, or just to strengthen the distinctiveness of the form" (p. 45), a hypothesis that is in line with this study's findings of a circular or recursive memorization pattern among L2 learners. The Instructional Function. Three items in the questionnaire focused on what may be called the instructional function of L2 inner speech. Questions on this particular function were included in the questionnaire on the basis of responses obtained in the pilot study conducted for Guerrero (1990/1991). In this study, learners had reported trying to learn the new language by silently imitating the pronunciation of their teachers, mentally constructing sentences, and applying the grammar rules they had learned. These efforts were considered to play the role of self-instruction; that is, they were efforts the students made to self-teach the language. The first item in this category, "When you mentally rehearse, do you try to imitate the pronunciation of words you have learned?," which explored the imitative aspect of mental rehearsal, obtained an average of 96% affirmative responses from the low, intermediate, and high proficiency groups. Advanced students reported an 80% of engagement in this practice. The second item "When you mentally rehearse, do you try to make sentences with certain words?" had a combined 81% of Yes responses from the low, intermediate, and high groups and 85% of responses from the advanced group. The third item "Do you try to apply the grammar rules you have learned to your inner speech in English?" had an average of 76% Yes responses from the low, intermediate, and high groups and 67% from the advanced learners. The difference in frequencies among the low, intermediate, and high groups was statistically non significant. Except for a slight increase between the lower level groups and the advanced group in the second item (making sentences with certain words), use of the L2 as a selfinstructional tool among the participants decreased with proficiency, at least in the imitative and grammar practicing aspects. This is not an unexpected finding considering that very advanced L2 students are presumed to have mastered most of the pronunciation and grammar aspects of the language. What is interesting to note is that, regardless of the proficiency level, the majority of the students seemed to engage in these self-instructional activities to some degree. The imitative function had a particularly high frequency (96% and a median response of often) among all groups except the most advanced (80%). Imitation seems to play a critical role in the internalization of a new language. Lantolf and Yanez (2003) believe that "internalization occurs as the person attempts to imitate privately" (p. 99). The type of imitation that L2 students engage in with the purpose of learning or self-teaching the language differs from other types of mimetic behavior in that it appears to be intentionally or deliberately deployed as a metalinguistic strategy. According to Lantolf and Yanez, "imitation, as a uniquely human activity, entails a transformative potential, and hence implies agency and

128

CHAPTER 5

intentionality" (p. 99). Furthermore, it appears that rehearsing the language for selfinstructional purposes can only take place at a level of inner speech where it is under conscious access, that is, under self-awareness (Vocate, 1994b). It is at this point that learners can intervene and deliberately exploit their ability to think words silently in order to teach themselves new features of the language. In the process, L2 learners become active agents of their own L2 internalization. The Evaluative Function. L2 learners are known to self- and other-correct language in spoken discourse (see, for example, Schwartz, 1980). In the studies reviewed here, the possibility that self- and other-evaluation of language might occur silently was explored. The evaluative function refers to the learners' use of L2 inner speech to assess what they know about the L2 and how others employ it. Like the instructional function, the evaluative is a metalinguistic function in that language-the L2 in this case-is used to focus upon itself. Evidence of this function was found in one of the students' comments to the questionnaire: "I use my inner speech to correct myself and to be sure that what I'm going to say is correct." The quantitative analysis of the questionnaire yielded a significant number of affirmative responses to all the items involving the evaluative function (see Table 5-2). Learners reported using their inner speech to (a) correct the pronunciation of words, (b) correct grammar errors, (c) evaluate the language of others, and (d) assess their own knowledge by silently answering questions asked in class. Interestingly, students seemed to be more concerned with correction of pronunciation than of grammar errors. When figures for the combined low, intermediate, and high groups were compared to those of the advanced group, it was evident that the evaluative function underwent a decrease. Apparently, as students gain more confidence in their language knowledge, they become less concerned with correcting and monitoring their own language and that of others; nevertheless, an advanced student and prospective English teacher made the following comment: "I constantly evaluate the use of the language in other people and in myself. I try to self-correct because there is nothing so horrible as a future teacher who does not master her field and mine is English." The frequency of responses for the item on answering questions mentally remained quite high despite the small reduction. Mentally answering questions posed in class is a frequently mentioned language learner behavior (Gillette, 1987; Ohta, 2001; Reiss, 1985). According to Gillette (1987), answering questions in the mind is a symptom of "active thinking during class periods" (p. 275), which may occur-as Reiss (1985) points out-whether the learner is called to answer or not. Ohta (2001) has labeled this behavior vicarious response and has defined it as "a process by which learners in a classroom setting formulate their own answers to a questions the teacher has addressed to another" (p. 17). Ohta found several instances of vicarious response in the private (audible) speech of her L2 learners. Because answering questions in the mind is a much more complex behavior than repetition or imitation, she speculates that "vicarious response evidences the learner's growing ability to produce thoughts in the L2" (p. 71). The high percentage of affirmative responses given by the advanced

L2 INNER SPEECH: WHAT LEARNERS SAY

129

learners in the studies reviewed here lends support to Ohta's hypothesis. Table 5-2. Percentages of Yes responses to items on evaluative function by proficiency level Item

Low, intermediate, and high*

Advanced

(a) Do you try to correct the pronunciation of words in your mind?

94

83

(b) Do you try to correct grammar errors when you mentally rehearse in English?

82

72

(c) When you hear other people speaking English, do you mentally evaluate how those people use the language?

90

85

(d) When your English teacher asks a 97 94 question in class, do you answer it in your mind even though you are not called to answer? *Note: Mean percentage of Yes responses at the low, intermediate, and high levels. Through the students' interview reports it was possible to ascertain the use of inner speech in its evaluative role as students mentally rehearsed in preparation for the communicative activities and after they had taken place. In some cases, the students reported correcting and assessing their English before the communicative task. In other cases, the evaluative role emerged after the activity as students reviewed their own performance or that of their interlocutors. One student, for example, reported evaluating her performance a posteriori: "As I was going home on the bus, I was thinking what I should have said and how I should have said it." The student also reported other-evaluating how her partner in the communicative task had used the language and her difficulty in understanding what this person had said. The student was very upset because she had performed less well than she had expected; therefore, she reported, "on the bus . . . I did it all over again to see if I was so stupid that I would forget everything. And I gave myself a 50 [the highest grade]." The Preparatory Function. Mentally preparing or planning L2 texts was another function of inner speech observed through the questionnaire.55 As Table 5-3 shows, inner speech had a major role in the planning of oral and written discourse. Analysis of item (a) in Table 5-3 revealed that the planning of oral texts was a major inner 55 This was referred to as the textual function in Guerrero (1994, 1999). The term preparatory is used here because it seems to describe more accurately the function involved.

130

CHAPTER 5

speech activity for the participants. Oral text rehearsal, or the planning and practice of spoken discourse, has been mentioned as a function of LI inner speech (Honeycutt etal., 1989,Smith, 1983;Sokolov, 1972; Vygotsky, 1986). Smith (1983), for example, believes that talking to oneself leads to "constructing and practicing . . . scenarios or scripts . . . that we employ in order to behave appropriately in new situations" (p. 92). Similarly, the role of inner speech in the creation of written texts was very important for the participants, as item (b) in Table 5-3 shows. This finding confirms the relevance of preparatory inner speech in the process of writing. Table 5-3. Percentages of Yes responses to items on preparatory function by proficiency level Item

Low, intermediate, and high*

Advanced

(a) If you have to talk to someone in English or you have an oral presentation, do you mentally rehearse what you are going to say?

92

85

(b) If you have to write something in English, do you rehearse first in your mind what you are going to write?

93

83

(c) Do you ever think how you would say or 83 76 write something in English, even if you are not going to use it? *Note: Mean percentage of Yes responses at the low, intermediate, and high levels. Interview reports confirmed the presence of preparatory inner speech among the participants. As they prepared for the communicative activity that had been announced, they planned what to say and how they would say it. Three specific roles of the preparatory function emerged: to give structure to discourse, to organize the sequence of words or ideas, and to experiment with language options. One student, for example, reported searching for alternative words as she tried to give some structure to her presentation: "I imagined a question I would ask. Then I looked for words to fit in to make it better." The preparatory function was not confined to planning text for particular speech activities in the future; there was also preparation of text for non-specific use, as item (c) indicates. It appears then that an important function of L2 inner speech is to serve as a sort of testing ground for the L2. This trying out of the L2 trough inner speech may be done with the purpose of learning the language or of strengthening knowledge of it. Lantolf (2000b) speculates that this type of rehearsal may lead to appropriation of the L2, which he describes as "the process through which the individual takes in particular features of the language through privately practicing and experimenting with these features" (p. 88).

L2 INNER SPEECH: WHAT LEARNERS SAY

131

The preparatory function, however, seemed to lose some of its importance as proficiency increased, as can be seen in Table 5-3, where all the items displayed a decrease at the advanced level. In contrast to the often or always medians obtained for the low, intermediate, and high levels for all these items, the median for the advanced level was sometimes. On the basis of these data it may be concluded that the preparatory function of inner speech, that is, its use in mentally practicing a text-oral or written-for future production is highly important for L2 learners but, as they become more confident of their language abilities, they tend to practice less ahead of time. The Dialogic Function. Imagining dialogues with oneself or others is one of the most frequently reported functions of inner speech (Vocate, 1994b; Cunningham, 1989; Honeycutt et al.; Smith, 1983). Vocate (1994b) referred to the dialogic function of inner speech as self-talk, a level of inner speech that is characterized by selfconsciousness and sufficiently elaborated syntax and semantics to sustain a dialogue with the self. Vocate believes self-talk is essential for the creation and development of the self. In self-talk there is an interaction between two sides of the self, the / or individual self based on the subjects' own intuitions and the Me or internalized social self based on other people's perspectives of the world and the individual. The interaction between the / and the Me makes possible self-reflection, a uniquely human process mediated by inner speech by which the individual critically considers choices and aims at the resolution of tensions between alternatives imposed by society and personal inclinations. An important dimension of self-talk pointed out by Honeycutt et al. (1989) is that inner communication may involve dialogue not only with the self, or with different versions of the self, but also with other interlocutors, real or imaginary. To reflect this fact, researchers refer to self-talk as "imagined interactions." This double role of the dialogic function of inner speech was well-expressed in one of the students' comments to the questionnaire: "When I'm alone at home or some other place, I make dialogues in my mind. I always imagine myself talking to someone, discussing various topics in English, or else I talk to myself." Two items in the questionnaire tested the dialogic function of inner speech. Dialogue with the self, or the intra-self function of inner speech, was measured by the item "Do you talk to yourself in English?" Dialogue with imagined others, or the interpersonal function of inner speech, was observed through the item "Do you imagine dialogues or conversations with other people in English?"56 Affirmative responses to the intra-self function of inner speech increased significantly with proficiency: 52% at the low level, 66% at the intermediate, 69% at the high, and 80% 56 In Guerrero (1990/1991, 1994, 1999), these two functions of inner speech were respectively called the intrapersonal and the interpersonal. To avoid confusion of the term intrapersonal with the overall concept of intrapersonal communication, as discussed in Chapters 1 and 2, an umbrella category has been created-the dialogic function-with two subcategories: the intra-self'role, which stands for silent dialogue with the self, or between the I and the Me, and the interpersonal role, which refers to imagined dialogue with others.

132

CHAPTER 5

at the advanced level. A similar pattern occurred with the interpersonal function of inner speech: 62% at the low level, 73% at the intermediate and high levels, and 76% at the advanced level. These findings offer strong evidence of the increasing role of L2 inner speech as a medium for talking to oneself. The Play Function. Several items on the questionnaire pursued what might be termed the play function of inner speech. L2 play has been hypothesized to facilitate, if not constitute in itself (Lantolf & Yanez, 2003, p. 100), acquisition of the L2 (Broner & Tarone, 2001; Lantolf, 1997). Language play has been described as an allencompassing form of rehearsal including talking to oneself, imitation, repetition, making up sentences, and spontaneous recall in the L2, an activity that occurs through both audible private speech or covert self-directed speech (Lantolf, 1997). Because some of these specific uses of inner speech have been contemplated here under other functional categories (namely, talking to oneself is considered to perform a dialogic function, imitation an instructional function, spontaneous recall a mnemonic function, and so on), the function of language play that was instantiated in the questionnaire is more in accordance with Belz's (2002) definition of language play as "the conscious repetition or modification of linguistic forms such as lexemes or syntactic patterns" (p. 16). Language play, in Belz's conception, is form-focused experimentation with language that entails both creativity and awareness of linguistic conventions and as such may serve not just a language learning function but a self-assertion role, through which L2 users exert and manifest their mastery of other languages and their linguistically multicompetent self. Belz's data of written language play show instances of morphological and syntactic experimentation with two or more languages, such as invention of new language names, modification of words based on melodic or rhythmic patterns, creation of funny or weird words, and experimenting with word order. Questions on language play including similar behaviors to those mentioned by Belz (2002) appeared in Guerrero's 1999 version of the questionnaire. Thus, only data from advanced learners are available. These data, however, are highly revelatory (see Table 5-4). On the one hand, the data show that advanced learners do engage in playful manipulation of the L2. On the other hand, it is plain that percentages of affirmative responses for these items are lower than for all the other inner speech functions. Actually, frequencies tend to cluster around the lower categories (never, almost never, sometimes), as indicated by the corresponding modes and medians. This might suggest that advanced learners do not engage in the play function of inner speech as frequently as in the other functions. Belz's (2002) data on language play emerged as part of an experiment in which learners were asked to write a multilingual text, a task that might have encouraged them to engage in playful and highly skilled manipulation of multiple languages. It is possible that unless they are prompted to do so, advanced L2 learners do not spontaneously resort to covert language play as much as they do to other types of inner speech.

L2 INNER SPEECH: WHAT LEARNERS SAY

133

Table 5-4 Responses to items on the play function of inner speech among advanced students

Item

Yes %

Mode

Median

Do you "play" with your inner speech in English, for example, . . . do you make up rhymes?

50

1

1.5

. . . do you invent funny or original combinations?

52

2

2

. . . do you invent your own words?

39

0,1

1

. . . do you experiment with the order of words?

52

1

2

The Affective Function. This function of inner speech has been widely documented. Researchers in the field of intrapersonal communication have investigated the role of self-talk or imagined interactions in controlling emotions and in creating an image of the self (Aitken, 1997; Apple, 1989; Hamilton, 1997; Honeycutt et al., 1989). Honeycutt et al. point out that positive or negative emotions may accompany imagined interactions. Self-talk may be used to vent anger or frustration, control fear or anxiety, chastise or encourage oneself, and provide self-comfort. Because inner speech is covert language behavior, it is an optimal setting for the discharge of private emotions that may not be conveniently expressed in public. Some examples of using self-talk for emotional purposes were elicited by Aitken (1997) in her study of (LI) inner speech patterns among college students: / Sometimes I scold myself or try to motivate myself with self talk. / Talking to myself helps me vent frustration, so when I do confront someone I am more calm and controlled. / I often blame myself for things that go wrong. I replay things in my mind. I think "if only" I had . . . . I always find a fault in myself to blame for the problem, (pp. 136-137)

Intrapersonal communication is particularly important as a release mechanism for people who suffer from communicative apprehension or anxiety, that is, fear or unwillingness to speak publicly (Richmond & McCroskey, 1989). This phenomenon is not uncommon among classroom L2 learners, who oftentimes feel afraid of communicating orally but may find freedom of expression in inner speech. The role of L2 inner speech in expressing or controlling emotions has been little explored. Cook (1998) is one of the few who has paid attention to the matter. In his survey study of internal and external uses of language (LI and L2), Cook included the use of internal language for emotional effects, such as when feeling happy or sad, being in pain or feeling ill, feeling tired, or when something goes wrong. Cook found greater use of the LI than of the L2 when having pain, feeling sad, or being tired and balanced use of LI and L2 when feeling happy or when things go wrong.

134

CHAPTER 5

Inner speech seems to perform an important affective function. Although this function was not tested through any of the items of the original version of the questionnaire (Guerrero, 1990/1991, 1994), it did emerge in the students' comments to the questionnaire and in the interview reports. Several students commented that they recurred to L2 inner speech because they felt shy or afraid to externalize their speech, for example: "I think I know enough English inwardly, but I don't express it because of shyness" and "I talk to myself because I want to speak English but I'm afraid to do it in public." An advanced student said she used L2 inner speech to say things that would be socially inappropriate: I usually do this [use L2 inner speech] when I'm alone because obviously I'm not gonna just sit and talk to myself out loud so whatever I'm thinking I just say it in my mind. Also, if I can't say what I'm thinking in front of someone in particular I do this a lot.

Planning what to say through inner speech appeared to help reduce the students' nervousness and increase their self-confidence. "Inner speech helps one to feel more confident about what one is going to say. I, for example, do not have many opportunities to practice English orally, so practicing it mentally helps me." In a postactivity interview, a participant said: "Unlike other students who had practiced only the night before or the same day, I was not nervous. I didn't think 'Oh, what am I going to say?' because I had everything planned." Inner speech was also used to boost the students' self-image, as in the case (mentioned in reference to the evaluative function) in which the student rehearsed her participation in the oral activity "all over again" as she was going home in order not to feel "so stupid" for having obtained a lower grade than she expected. In short, through the students' self-reports it was found that inner speech performed an affective role when used for the following purposes: to produce self-satisfaction, to reduce nervousness and acquire self-confidence, to provide selfdiversion, and to improve one's self-image. As a result of this evidence, several questions to measure these roles were introduced in the 1999 version of the questionnaire (see Table 5-5), which was administered to the advanced students. Analysis of the questionnaire items indicated that inner speech as mental rehearsal was an exceedingly positive affective experience rather than negative for the advanced students: Whereas 96% of them reported L2 inner speech made them feel good (item a), only 9% of them said it made them feel bad (item b). L2 inner speech also emerged as a powerful instrument to gain self-confidence (item d) and to derive self-diversion in the L2 (item e). Relatively high frequencies were reported for the use of inner speech to reduce nervousness, anxiety, or apprehension (item c) and to increase selfesteem (item f). Inner speech in English was found to be little used by the learners to criticize or punish themselves (item g). These findings thus confirm in a quantitative way the students' qualitative self-reports about the existence of an important affective dimension of inner speech during mental rehearsal of the L2.

L2 INNER SPEECH: W H A T LEARNERS SAY

13 5

Table 5-5. Responses to items on the affective function of inner speech among advanced students Item

Yes %

(a) Does your inner speech in English make you feel good?

96

(b) Does your inner speech in English make you feel bad?

9

(c) Does your inner speech in English reduce your nervousness, anxiety, or apprehension?

74

(d) Does your inner speech in English give you self-confidence?

93

(e) Does your inner speech in English entertain you and help you pass the time?

87

(f) Do you use your inner speech in English to increase your selfesteem?

54

(g) Do you use your inner speech in English to criticize or punish yourself?

35

To what extent is L2 inner speech class-related among L2 learners? One of the aspects of L2 inner speech measured through the questionnaire was the extent to which L2 inner speech is related or not to the L2 class among instructed language learners. Overall, it was found that affirmative responses to the item "Do you catch yourself thinking in English about things not related to your English class?" yielded a significant increase with proficiency, ranging from 53 % affirmative responses at the low level to 89 % at the advanced level. Conversely, there was a decrease of Yes responses to the question "Is your inner speech related to your English class?," from 77% among the lower levels to 70% at the advanced level. It seems then that as learners reach high levels of L2 proficiency and make more natural use of it, L2 inner speech becomes a very important alternative mediating tool for thinking about all kinds of things, not just topics associated with their L2 class. The interviews with the students confirmed the presence of class-unrelated rehearsal, that is, inner speech not related to any specific task or aspect of the L2 class but for various self-related purposes, for example, for self-fulfillment, selfdevelopment, and self-diversion. Several students said they engaged in L2 inner speech because it made them feel good about themselves or they enjoyed practicing it. One participant reported she had found herself using the L2 to mentally explain to her boyfriend why she wanted to study biology and not something else. The process had helped her clarify her point of view. L2 inner speech was also used for self-diversion, that is, with the purpose of entertaining oneself or helping pass the time. One learner

136

CHAPTER 5

mentioned that she usually rehearsed in English when waiting for someone or before falling asleep because "it's fun." To what extent do L2 learners rely on their LI as they engage in L2 inner speech? The role of the LI in L2 inner speech is one of the least understood and most arguable issues in the inner speech literature. One theoretical position holds that the LI always exerts a strong semantic influence on an L2 at the level of inner speech (Ushakova, 1994; see Chapter 3). In this view, the LI is the main medium for verbal thought, and the L2 is treated as a tool to be used in translating from the LI (in speech production) and into the LI (in speech reception). Another position maintains that the influence of the LI tends to decrease as learners develop a unified conceptual system where the LI and the L2 merge or coexist and translation back and forth between the LI and the L2 is no longer necessary (John-Steiner, 1985b; Kecskes & Papp, 2000). Still another argument is that, regardless of the proficiency level, the LI remains as an internalized cognitive tool to be used strategically in making meaning (Huh, 2002; Upton & LeeThompson, 2001). LOW

INTERMEDIATE

HIGH

ADVANCED

70%

67%

70%

76%

Figure 5-2. Percentages of affirmative answers to the question "Is your inner speech in English mixed with Spanish? " by learners on various proficiency levels.

In the data analyzed here, the item "Is your inner speech in English mixed with Spanish?" showed a slight decrease of Yes responses for the intermediate level and a moderate increase for the advanced (see Figure 5-2 above). Percentages among the levels, however, remained within a small range (67% to 76%) with a common median of sometimes. These figures are high enough to suggest that the LI (Spanish, in this case) does have a presence in the learners' L2 inner speech, but its extent is not as robust as to indicate that L2 inner speech necessarily has to rely on the LI all the time. On the other hand, the fact that even the most advanced students-whose level of the L2 put them in the category of fully bilingual-reported moderate frequencies of LI use in their inner speech does not allow one to rule out the possibility that L2 users will always count with their LI as a tool to be deployed strategically, either at the deep conceptual and semantic levels or at the precommunicative coding stage. Comments to the questionnaire or during the interviews provided more specific details on the role of the LI in L2 inner speech. Many of the students in the low and intermediate levels said that, when they rehearsed in English (the L2), if they could not think of a word in English or "got stuck," they usually lapsed into Spanish. Some of them said that usually their inner speech came first in Spanish and then they translated it into English word for word. One student commented that she used to think

L2 INNER SPEECH: WHAT LEARNERS SAY

137

something in English and then translate it into Spanish to see if it made sense. Comments from advanced students revealed a balanced use of both languages in the mind: / I tend to mix Spanish and English in my inner as well as in my external speech. / I often mix both Spanish and English in my inner speech although I'm quite capable of thinking separately in both languages and switching back and forth. / In Spanish or English I'm always having conversations in my mind that help me fix my thoughts and say things right.

Summary of Results The two studies reviewed above (Guerrero, 1994, 1999) investigated the nature of the inner speech that occurs as learners mentally rehearse in the L2. Data in the form of questionnaire responses and comments from 472 ESL learners and interview reports from 9 participants provided evidence of the widespread use of L2 inner speech for a variety of rehearsal-related functions. One of the most important findings of this research was that, overall, the presence of L2 inner speech in the learners' minds increased with proficiency in the L2. The studies revealed several structural and functional characteristics of L2 inner speech. Syntactically, L2 inner speech tended to be abbreviated although it could be expanded at times, for example, when holding mental conversations. Phonologically, L2 inner speech was experienced as sonorous, rather than silent, sometimes adopting a heteroglossic quality and often sounding "better" in the mind than when outwardly articulated. Semantically, L2 inner speech was usually perceived as meaningful although sometimes it appeared to be the venue for the elucidation of unfamiliar words. Inner speech as mental rehearsal of the L2 was found to perform several specific functions: Mnemonic: the use of inner speech as an aid to memory; inner speech as the mechanism for the storage and retrieval of L2 words Instructional: the use of L2 inner speech for self-instruction; trying to self-teach the L2 through deliberate imitation of L2 models, applying grammar rules, and attempting to formulate sentences in the L2 Evaluative: the use of L2 inner speech to assess, correct, and evaluate the learner's own knowledge of the L2 and that of others Preparatory: the mental preparation or planning of speech production, oral or written, in the L2 Dialogic: imagining dialogues in the L2; use of L2 inner speech to hold dialogues with oneself or different versions of the self (intraself function) or to mentally engage in conversations with other interlocutors (interpersonal function) Play: the playful or creative manipulation of and experimentation with the L2 Affective: the use of L2 inner speech to control or express emotions, derive selfsatisfaction, reduce anxiety, acquire self-confidence, entertain oneself, and improve self-image

138

CHAPTER 5

Despite a general increase in L2 inner speech with greater proficiency in the language, many of the rehearsal-related functions appeared to be less frequent as proficiency increased. Specifically, more advanced learners reported engaging less than lower level learners in the use of L2 inner speech for planning speech production, self- and other-evaluation, memory storage and retrieval, and self-teaching the language. Conversely, more advanced learners evidenced greater use than lower level learners of L2 inner speech for dialogic purposes. The play function of L2 inner speech was found to be rather limited among the advanced participants when compared to the other inner speech functions. L2 inner speech, however, seemed to play a very important affective role among learners at all levels, entailing positive feelings about oneself as well as providing self-confidence and self-diversion. The intensification of class-unrelated L2 inner speech and the concomitant reduction in class-related L2 inner speech as proficiency in the other language increased constitutes major evidence of the growing role of the L2 as an ideational tool, useful for diverse types of cognitive and communicative purposes and not just for classroom language tasks. The fact that moderate frequencies of L2 mixed with LI in the students' inner speech were reported at all levels of proficiency suggests that the presence of the native language cannot be ruled out, either as a fall-back aid when knowledge of the L2 is inadequate, as an important meaning-making tool, or-particularly among the most advanced learners-as an alternative cognitive and linguistic resource. To conclude this section, it seems worthwhile to briefly cite the replication study that Gutierrez (2000) conducted with a population of adult EFL learners at the Universitat Autonoma de Barcelona. Gutierrez's findings confirmed Guerrero's (1994, 1999) results regarding the widespread use of inner speech during mental rehearsal of the L2 among the FL learners and the fact that L2 inner speech increases with proficiency in the L2. Gutierrez's study corroborated the linguistic characteristics and functions of L2 inner speech found by Guerrero and the way they change in relation to proficiency. A new finding by Gutierrez was that L2 inner speech may have a contextualizing function; that is, inner speech may be instrumental in staging the learners' representation of a future communicative situation. Through this function, the learners in Gutierrez's study imagined the situation in which they would have to participate, their role in it, and that of other interlocutors. According to the researcher, the contextualizing function helped the participants rehearse for the communicative activity. SECTION 2. EARLY STAGES OF L2 INNER SPEECH DEVELOPMENT Although a few studies have specifically addressed the question of intemalization of social speech and its transformation into inner speech among LI children (Flavell, et al.,1997; Goudena, 1992; Liva et al., 1994; Otte, 2001), little is known about this process among L2 learners, children or adult. Some data in the form of audible private speech have begun to throw light on the process of intemalization or privatization of the L2 within instructed learning environments (Lantolf & Yanez, 2003; Ohta 2001;

L2 INNER SPEECH: WHAT LEARNERS SAY

139

Saville-Troike, 1988). It has been shown, for example, that learners engage in selfaddressed repetition or imitation of the language heard in the environment, vicarious response, preparatory rehearsal, and experimentation with language forms, all behaviors that are highly suggestive of how an L2 might begin to be internalized. However, these are instances of loud, fully articulated self-directed speech. What happens silently at the internal level in the early stages of L2 learning is still much of a mystery. Vygotsky (1986) claimed there is a transformation as private speech gives way to inner speech, but how does this transformation occur in the learning of an L2? In other words, what do learners do to internalize the L2 and how do they begin to develop their ability to "think words" in the L2? In this section, results of a study (Guerrero, 2004) pursuing the early manifestations L2 inner speech will be presented. Participants and Data Collection The study was based on data obtained through learner diaries, a source of verbal reports on inner speech that had not been previously exploited systematically for that purpose. The participants were 16 Spanish-speaking ESL learners whose low scores on the ESLAT (below 400 points; see description of levels in this chapter on p. 120) had placed them on the low proficiency level at the university. These students were taking an introductory ESL course designed to provide them with intensive practice in the essential elements of the English language, including grammar, vocabulary, and communicative skills. As part of the course requirements, the students were asked to keep a diary where they would reflect on the inner speech they experienced during the class and after class. Because the aim of this activity was to collect information on the students' L2 inner speech, specific instructions were given for keeping the diary (see Appendix). The instructions included a definition of inner speech (the same one used in Guerrero's 1994 questionnaire study) and a list of suggested questions to help students think about their inner speech. To ensure ease of expression, the diary was to be written in Spanish, the students' LI. To obtain fresh retrospective reports on the students' "in-class" inner speech, once a week during the last ten minutes of the class the students were asked to write in their diaries about the inner speech they had experienced during the class. Outside the classroom, students wrote their entries at their own convenience. Sometimes students indicated entering their comments immediately after experiencing an inner speech event. The diary was collected four times over a period of four months. After each collection, the teacher/researcher read the students' entries marking those that required clarification or further explanation. As the diaries were returned the next class, the teacher called those students whose entries had been marked and, as a form of stimulated recall (Gass & Mackey, 2000), asked them to clarify or explain the entries. As students did this, the teacher took notes of their responses or asked the students to add their comments to the entries. The stimulated recall sessions were conducted in Spanish, the students' LI.

140

CHAPTER 5

Content Analysis A content analysis of the 16 diaries collected from the participants was performed. A recursive process of reading and rereading the entries was undertaken, first, to identify the themes or responses that were relevant to the universe of content (Wenden, 1987), in this case, learners' comments about inner speech in the L2, and, second, to produce a categorization of inner speech occurrences. A useful resource that helped develop categories was Sokolov's (1972) extended definition of inner speech (see Chapter 2 in this book, pp. 43-44). Four major categories and various subcategories emerged in this process. This categorization was used to code the previously identified themes. Reliability of the coding was established with the aid of an external rater (96% agreement for major categories and 86% for the subcategories). The four categories (identified by the numbers 1,2,3, and 4) and their corresponding subordinate varieties (identified as a, b, c, etc.) are exemplified below. 7. Concurrent Processing of Language Being Heard or Read One major type of covert speech behavior reported by the participants was the concurrent processing of language being heard or read; that is, as students were listening or reading material in the L2, they inwardly worked upon selected segments of this material. Sokolov (1972)'s description of (LI) inner speech as "concealed verbalization . . . in the logical processing of sensory data" and as "the internal projection" of external speech, "arising first as a repetition (echo) of the speech being uttered and heard" (p. 1) helped identify the first category of L2 inner speech among the participants. Students mentioned doing several things with their inner speech as they were reading or listening to language: (a) repeating silently or sub vocally, (b) spelling the words being heard, (c) trying to understand the language, or (d) associating the words being heard or read. Subvocal or silent repetition was sometimes done in order to practice the pronunciation of words or to imitate an accent or pronunciation. Subvocal or silent repetition was also undertaken in order to memorize language. Students often focused on comprehending the language they were reading or listening to rather than merely repeating it; that is, they tried to understand the language by isolating troublesome or unfamiliar words, translating to the LI, analyzing words, looking at the written captions on TV to figure out what was heard, or even interrupting the listening or reading process to look up words in the dictionary or to ask someone. Associating words implied relating the language being heard or read to language heard or read before, to an event or concrete referent, or to a visual image. The following are examples of each of the subcategories in the concurrent processing of language being heard or read. (Please note that the entries were originally written in Spanish but have been translated to English here. Boldface is used for words written in English by the students.)

L2 INNER SPEECH: WHAT LEARNERS SAY

141

(a) Subvocal or silent repetition of language being heard or read - (to practice pronunciation) Today during class as I heard the word yawning I started repeating it several times in my mind and in a low voice so as to pronounce it better. (b) Spelling word in mind or subvocally - I imagined how some words were written according to their pronunciation. They were words that the teacher mentioned and I thought how they were spelled. (c) Trying to understand language being heard or read - (by translating into the LI) I was in my apartment watching a movie and it was in English. To understand it a little I tried to translate it mentally. - (by analyzing words) I was watching a TV program in English and when they mentioned the word myself I tried to imagine its meaning: my, "yo" [Spanish word], self, "mismo" [Spanish word]. I don't know if that is correct. - (by asking someone about the meaning) I thought of the word turn [Comment added as stimulated recall response: While we were reading the story of the woman with the cat]. I asked the teacher what that verb meant. (d) Associating language being heard or read - (by relating language to an event) Today in class when I heard the word robbers I remembered what had happened to my friend yesterday. Some robbers broke her car window and stole her CD player, so when I heard the word robbers I remembered that. - (by relating language to a visual image) While the picture in the story was being described, I visualized in my mind each description that was done in English. 2. Recall of Language Hear d, Read, or Used Previously Sokolov's (1972) experiments demonstrated that covert inner speech movements are present in the recollection of phrases heard or past events (p. 180). This led him to the assertion that we "remember with the aid of words," for example, as "we recall books read or conversations heard" (p. 1). Inner speech as recall was the second category of L2 inner speech found among the participants. In their reports, the learners mentioned L2 inner speech occurring as they remembered language they had heard, read, or used previously. In this case, the learners' inner speech was engaged in delayed or reactivated processing of language, rather than concurrent. The difference between this category and the first is that during recall the person is not trying to process language he/she is hearing or reading at the moment. Rather, the person is remembering, and in so doing he/she is reprocessing the language heard or read before. Mnemonic processes are very important in this type of inner speech activity. One of these was (a) playback, that is, the insistent hearing or replay of words in the mind.

142

CHAPTER 5

The students often mentioned that words would "get stuck" in their mind. There was also (b) spontaneous recall of words, for example, when a word suddenly popped into the students' minds. Sometimes the students reported (c) making a deliberate effort to recall or remember some word in English. They also engaged in (d) subvocal or silent repetition of language they had heard or read before. The recall category also involved (e) reflecting on language heard, read, or used before. This frequently implied recapitulating language knowledge in terms of grammar rules, vocabulary, or meaning. Sometimes, the students went beyond recapitulating and (f) tried out their knowledge of the L2 by applying it in novel constructions. As part of recall, the students engaged not only in delayed repetition of a word but also in (g) trying to spell it in the mind or subvocally. The following are examples of recall (delayed or reactivated processing) of language heard, read, or used previously. (a) Playback - I saw the phrase the new war on the [TV] news. Besides being scared, I was all day repeating that phrase in my mind. (b) Spontaneous recall of words (words suddenly pop into the mind) - 1 was sleeping but then I woke up. In my mind I said Go back to sleep. [Comment added as stimulated recall response: Those were words I had seen in the story and I remembered.] (c) Deliberate recall, trying to remember - At home, I was trying to remember how to say handbag in English. (d) Subvocal or silent delayed repetition of language heard or read - After class, I kept repeating all the words I had learned, for example: expensive, cold, clean, etc. (e) Reflecting on language heard, read, or used; recapitulating language knowledge -1 woke up at 1 am. I had been dreaming that I was in the English class and in my internal language I started remembering a part of the exam that gave the answer and one had to ask the question, He's eighteen years old. I was thinking internally about the question I put in the exam How old are you?, but this is when you ask a person face to face and I understood that my answer was wrong and should have been How old is he? This type of internal language helped me to learn something by thinking over the exercise in the exam on which I had doubt. (f) Trying out knowledge of the L2 (Applying knowledge) -1 was analyzing in my mind the possessive nouns and I realized that I can use them in a conversation, that I can say "el lapiz de Omar" [Spanish words], Omar's pencil. I understand I need to practice to make it part of my daily living so that I do not forget it. (g) Spelling a word in the mind or subvocally - A few days ago, someone mentioned the word laser. The word has remained in my memory and I have been trying to spell it.

L2 INNER SPEECH: WHAT LEARNERS SAY

143

3. Preparation before Writing or Speaking Inner speech in the LI has an important preparatory function, as pointed out by Sokolov (1972, pp. 65-66), Vygotsky (1986, pp. 88, 243), and others. In this study, students reported using L2 inner speech to plan speech production. Through inner speech, students prepared what to say (or write) or how to say (or write) something. Planning speech production is a far more complex cognitive task than simply repeating or trying to understand language that is heard or suddenly recalling language. In preparation before writing or speaking, the students must have heard or read the language before and done something with it, and they must retrieve it from memory in order to plan speech production. Students reported (a) planning what to say or how to say something, to the teacher or someone else, orally or in written form. Preparation sometimes involved (b) planning an exercise response or (c) planning a conversation the students knew they were going to have, that is, before holding a real conversation. As students planned their language production, several processes were evident, for example, translating from the LI to the L2, retrieving words from memory, rehearsing pronunciation or spelling, activating grammar knowledge, and writing their planned responses. Preparation also involved (d) mentally answering a teacher's question addressed to the class and (e) mentally repeating some language material in order to memorize it or learn it so as to be able to use it on a predetermined future occasion, for example, a test or oral report. Each subcategory of preparation before writing or speaking is exemplified below. (a) Planning what to say or how to say something - (by retrieving words from memory) In the English Lab class I was waiting for the teacher to come to me to ask her a question and I was thinking how to call her in English so that she would come to me. After a few moments of searching for words in English in my mind, I could say, Profesor [sic], scused [sic] me, can you help me? The teacher heard me and came to me. (b) Planning an exercise response - During class today I had a quiz in which I had to think what words to write in English to ask a question, for example, the word which. (c) Planning a conversation (before holding it) -1 had to call my brother who lives in the States, but because my sisterin-law might answer the phone and she doesn't speak Spanish, I thought of the words I could use, like How you doing? [sic] and Where my brother? [sic]. But then I didn't call her. (d) Answering a teacher's question in the mind - When the teacher asked "and [what happened to] the man?," I answered in my mind is gone. (e) Repetition to memorize (learn) language for future use (as in preparation for a test or oral report) -1 was at home studying the story for the oral test, and it was hard for

144

CHAPTER 5

me to pronounce feel it. At night, when I was in bed, I kept repeating it in my mind. 4. Silent Verbalization of Thoughts for Private Purposes According to Sokolov (1972), inner speech arises "at the instant we think about something . . . . The elements of inner speech are found in all our conscious perceptions, actions, and emotional experiences" (p. 1). Evidence of an incipient form of this type of inner speech-labeled here the "silent verbalization of thoughts for private purposes"-was found to occur through L2 mediation among the participants. This category refers to the use of L2 inner speech to verbalize feelings, ideas, or intentions, without actual speech production being involved. A subcategory was (a) when a thought, expressed in English, suddenly came into the students' minds. Some of these occurrences included insults and formulaic expressions like "I'm sorry" or "Thank you." Another subcategory was (b) imagining conversations for purely private purposes and not in preparation for a future real conversation. Here are some examples of the silent verbalization of thoughts for private purposes, not related to actual speech production. (a) Spontaneously thinking in the L2 (without an audience in mind) -1 was feeling cold, and I thought I'm cold. (b) Imagining conversations (not in preparation for a real conversation) - I was having a conversation in English with myself asking myself things. [Comment added as stimulated recall response: I was half asleep, trying to go to sleep, as if in a dream. It was my voice and like that of another person]. Quantitative Analysis The content analysis of the entries yielded a total of 291 themes (units of analysis) for the 16 participants. Once the relevant themes had been identified, the coding and quantification of themes according to the four categories described above was undertaken. The purpose of this analysis was to obtain an estimate of predominant and less predominant modes of inner speech in the L2 among learners at the early stages of acquisition. Results can be seen in Table 5-6. Overall, a predominance (45%) of concurrent L2 inner speech while processing language being heard or read was found. Also important was the presence of inner speech (38%) in the recall of language previously heard, read or used. Some moderate inner speech activity (13%) was observed in the preparation of speech production, and only a small role (4%) for the silent verbalization of private thoughts was obtained. Following is a more detailed look at the quantitative results.

L2 INNER SPEECH: WHAT LEARNERS SAY

145

Table 5-6. Categories of inner speech by number (no.) of participants and frequency of themes

Categories of inner speech

No. of participants who reported each subcategory

Frequency of themes

1. Concurrent processing of language being heard or read a. Subvocal or silent repetition

13

63

b. Spelling word

4

7

c. Trying to understand

15

46

d. Associating language

7

16 132(45%)

Subtotal

2. Recall (delayed or reactivated processing) of language heard, read, or used a. Playback

6

9

b. Spontaneous recall of words

8

28

c. Deliberate recall

8

11

d. Subvocal or silent repetition

5

11

e. Reflecting, recapitulating

7

35

f. Trying out L2 knowledge

8

14

g. Spelling words

2

2 110(38%)

Subtotal 3. Preparation before writing or speaking a. Planning what (how) to say

9

14

b. Planning exercise response

3

7

c. Planning a conversation

2

2

d. Answering questions in mind

5

7

e. Repetition to memorize

6

8

Subtotal

38 (13%) (continued)

146

CHAPTER 5

Table 5-6 (continued). Categories of inner speech by number (no.) of participants and frequency of themes Categories of inner speech

No. of participants who reported each subcategory

Frequency of themes

4. Silent verbalization of thoughts for private purposes, not related to speech a. Spontaneously thinking in the L2

4

4

b. Imagining conversations

4

7

Subtotal Total

11 (4%) 291 (100%)

According to Allwright and Bailey (1991), in analyzing learner diaries, three features need to be observed: (a) frequency of mention, (b) distribution of mention across participants, and (3) saliency (the strength with which a topic is reported). When applied to the 291 themes obtained in this study, Allwright and Bailey's criteria reveal that some subcategories were not only frequently mentioned but also widely distributed among the participants (see Table 5-6). For example, subcategory la, sub vocal or silent repetition of language being heard or read, with a total of 63 themes, was not only the most frequently reported subcategory but also mentioned by a large number of participants (13 out of 16). The combined effects of frequency of mention and distribution of mention makes this subcategory one of the most salient types of inner speech reported by the participants. Another frequently and widely mentioned subcategory was lc, trying to understand language that is being heard or read. Some subcategories, such as the spontaneous recall of words (2b), reflecting and recapitulating (2e), trying out knowledge (2f), and planning what to say or how to say something (3a), had considerable frequency of mention and a fairly wide distribution across the participants. Conversely, other subcategories, such as trying to spell a word as it was recalled (2g) and planning a conversation to be held (3c) were rarely reported and then only by two individuals. Discussion The diary data obtained from 16 low proficiency ESL students provided evidence of how learners begin to internalize L2 social speech and turn it into L2 inner speech. Four early manifestations of inner speech in the L2 were identified: (1) concurrent processing of language during listening and reading, (2) recall of language heard, read, or used previously, (3) preparation before writing or speaking, and (4) silent verbalization of thoughts for private purposes. Although it is difficult to generalize from diary self-reports, a tendency observed among these low proficiency learners was

L2 INNER SPEECH: WHAT LEARNERS SAY

147

to engage predominantly in the concurrent and delayed processing of the L2 and less predominantly in the planning of speech production and mental verbalization of private thoughts. Various specific behaviors were detected among these categories. For example, one way of processing the L2 was associating the words being heard to known referents. Another way was to recall a word heard previously and mentally try out its use in a sentence. There was considerable variability in the extent to which the 16 learners engaged in these behaviors. Some students reported high numbers of one particular inner speech type; others reported a variety of inner speech behaviors. Some students produced a high number of relevant themes (maximum 35); others, only a few (minimum 5). According to Frawley and Lantolf (1986), wide variability in individuals' inner speech production is to be expected since it is a natural consequence of the process "by which humans individuate through socialization" (p. 708). An important finding in this study was the existence of L2 inner processes that were concurrent with listening or reading, especially subvocal or silent repetition and a selective focus on difficult or unfamiliar words, trying to understand them, relating them to known material, or even struggling with their spelling. This finding is compatible with the notion of a working memory mechanism known as the "phonological loop" that relies on subvocal rehearsal and auditory imagery to help learn new words (see Chapters 1 and 2). Results were also consistent with the notion of elaborative mental rehearsal, a complex form of subvocal repetition crucial for longterm retention that involves subjecting verbal material to various levels of semantic-and possibly formal in the case of L2 learners-analysis (see Chapter 1). Subvocal repetition is a salient private speech behavior (Ohta, 2001); it is therefore not surprising to find it as a form of internal, covert speech. Sokolov (1972) believed that the need to articulate words subvocally or aloud is a natural phase in the automatization and mastery of foreign languages that is gradually substituted by rapid, reduced mental verbalization through inner speech. In Sokolov's view, it is necessary to engage in "maximal" articulatory subvocalization of foreign language material during listening or reading in order to develop the ability to rapidly process and understand language with "minimal" speech articulation (p. 263). The participants' efforts at silently repeating, associating, spelling, and trying to understand the L2 in this study might have been signs of the maximal concealed articulation that is essential for the internalization of the L2 and its eventual development as a tool for thought. Similar effects might have the students' second most frequent type of covert speech behavior, that is, the spontaneous or deliberate retrieval of L2 words or phrases and the subsequent re-processing of this material in the form of further repetition, reflecting upon its form or meaning, and trying it out in novel constructions. The use of L2 inner speech for purposes of speech production or externalization, or simply for verbalizing ideas privately to oneself, was plainly reduced in comparison with the two other categories (concurrent and delayed processing). It is important to note, however, that low proficient students do make efforts, even if minimal, to deploy the L2 intramentally in preparation for overt oral or written tasks, such as holding a conversation with a customer or responding to a teacher's question, as well as for the expression of some of their feelings through formulaic expressions in the L2. At this

148

CHAPTER 5

level of proficiency, however, students seem to be much more immersed in the process of turning L2 social speech into L2 inner speech than in the process of verbalizing their thoughts for intra- or interpersonal communication. Furthermore, as the students' reports indicate, there is still a strong dependence on the LI as the semantic base from which students translate meanings into the L2. In summation, this study threw light on some of the inner speech processes that characterize the transformation that L2 social speech undergoes in the early stages of L2 internalization. Foremost among these processes is the inward reproduction of the L2 in the form of concealed verbalizations, concurrently as language is being heard or read or in a delayed fashion, as it is spontaneously or deliberately recalled for further examination, reflection, manipulation, and memorization. At the early stages of L2 development there seems to be little use of L2 inner speech for complex mental operations. L2 learners have simply not internalized enough of the L2 to carry out sustained inner speech activity in the L2 to mediate complex cognitive tasks. Rather, L2 inner speech in its beginnings seems to function largely as an echo of external language. In form, this L2 inner speech appears to be syntactically expanded as learners engage in maximal inner articulation of speech heard or read or as they rehearse for future speech production. Inner speech, however, begins to show signs of increasing abbreviation and condensation as learners selectively single out key words they hear or read and make them the focus of their attention or as they verbalize their answers in fragmentary, compressed semantic units.57 A striking conclusion that might be derived from this study is that learners do not appear to be passively parroting in their minds the L2 in their surroundings-although this may occur at times-but seem to be primarily occupied in working with the language, subjecting it to mental handling that may eventually lead to comprehension, retention, and ultimately production. Also impressively evident is the fact that, as far as their inner language processes go, learners do not stop being learners once they step out of the classroom. Internalization of an L2 does not exclusively take place in the classroom, not just because learners may be exposed to multiple forms of the L2 in the social world, speech that they appropriate too, but also because they continue to ruminate over the language used in class once this is over. To van Lier (1996), this between-classes "process of inner speech, being mentally 'busy' with the language, reflecting on language-related phenomena, and noticing things that are relevant to progress" (p. 43) is essential for language development to take place. It is hypothesized that these initial efforts at appropriating the L2 through internal activity are indispensable for the eventual development of the L2 as a tool for thought. SECTION 3. PROS AND CONS OF A VERBAL REPORT METHODOLOGY IN THE STUDY OF L2 INNER SPEECH This chapter has reviewed research on inner speech involving a variety of verbal report 57 Typical of this tendency to semantic abbreviation is the mental answer a student reported giving to a teacher's question. As the teacher asked "How's the cat?," she responded in her mind "afraid."

L2 INNER SPEECH: WHAT LEARNERS SAY

149

methods of data collection: a questionnaire, interviews, and a diary combined with the stimulated recall technique. Several insights were gained in terms of the pros and cons of applying these methods to the study of inner speech processes. The Questionnaire As applied in Guerrero (1994, 1999), the questionnaire had the advantage of yielding abundant quantitative data from a large number of participants. In this way, certain general tendencies among individuals concerning the extent to which they experienced L2 inner speech, its form and functions could be ascertained. Importantly, the questionnaire made possible the measuring and comparison of responses from numerous participants at various levels of L2 proficiency and across various populations (as mentioned, the same questionnaire was utilized with an FL group of learners by Gutierrez, 2000). The inclusion of an open-ended section in which subjects could voluntarily write down their comments was useful as a means of corroborating information already contemplated in the questionnaire items and collecting additional insights into the phenomenon. Interestingly, the comments were instrumental in showing the effect that taking the questionnaire had on the learners' metacognitive awareness of the phenomenon. Many mentioned that through the questionnaire they had become aware of their inner speech in the L2 and how this could benefit them in learning the language. One student commented: "This questionnaire helps to raise consciousness among people about how they can help themselves in practicing English." The major weakness of a questionnaire on inner speech is that it invites participants to retrospect and generalize about a mental phenomenon that is greatly impermanent and likely to be forgotten or unheeded at the time of its occurrence. Several problems related to memory may affect responses to a questionnaire. One is the irretrievability of the inner speech event; in other words, the event has been forgotten. Another problem is incompleteness; the subject remembers but only partially. The third is inaccuracy; the subject does not remember exactly as the event occurred. The main consequence of these problems is that the subject may arrive at unfounded generalizations and inaccurately increase or decrease the reported frequency of events. For these reasons, responses to a questionnaire based on delayed retrospection should be taken with some reserve, keeping in mind the elusive nature of the phenomenon, the unreliability of memory, and the subjective nature of self-reports. If possible, other measures focusing on specific recent inner speech events less likely to have been influenced by loss of memory should be utilized to substantiate or complement questionnaire data. The Interviews The semi-structured interviews held with selected participants were useful in providing fresh retrospective reports about the inner speech the learners had experienced before and after two communicative activities in the classroom. The interviews confirmed

150

CHAPTER 5

many of the formal and functional characteristics and varieties of L2 inner speech that had been probed through the questionnaire, only this time the reports were not generalizations but specific comments about particular inner speech events. One important finding was that of the affective function of inner speech, a role that had not been tested through the original version of the questionnaire (Guerrero, 1994) but that appeared consistently in the students' interview reports58 Another finding, which was made possible by the specificity of the interview reports, was that the nature of inner speech depends a lot on the type of the cognitive operation it mediates. For example, because one of the activities involved talking about a picture, the students reported that in preparing for the activity their inner speech was highly dependent on the visual image and therefore of a mixed verbal-pictorial character, an observation that lends weight to Sokolov's (1972) claim that sometimes the language of inner speech "is complemented by graphic images" (p. 31). The interviews also made patently clear the "review" function of inner speech (Honeycutt et al., 1989; Smith, 1983) as students reported going over what they and their interlocutors had said and how they had said it after the activities took place. The Diary Because the aim in Guerrero (2004) was to collect information on the students' L2 inner speech, the diary used for that purpose was not open-ended but rather tightly focused on the topic of interest. This type of diary is not infrequent in language learning research (McDonough & McDonough, 1997, p. 124), especially when the universe of content is pre-established by the researcher (see Wenden, 1987, p. 116). This procedure ensures that at least a large portion of the entries will deal with the topic at hand. Content analysis, however, is still imperative to sift relevant from irrelevant responses as well as general from specific comments about the phenomenon. One advantage of the diary, as it was implemented, was that many of the students' entries were written during the last ten minutes of the class or immediately after the inner speech event had taken place and therefore had much more recency of recall than those one could get through a questionnaire. In addition, because the diary was conducted outside the classroom as well as inside, it was possible to obtain reports on a wide variety of inner speech experiences and not just those associated to one activity in particular. Also, because the diary was carried out during a fairly long period of time and assigned to a whole class of learners, a great deal of information was gathered on a variety of inner speech occurrences from multiple participants. Finally, complementing the diary with the stimulated recall technique was a valuable procedure. Holding brief interviews with the students after their entries were read and marked for clarification and expansion facilitated and enriched the interpretation of the data. Collecting data on inner speech through a diary, however, has its limitations. A diary by definition is a self-report, and self-reports are self-manipulated. The diarist 58 This finding led to the incorporation of several items on the affective function of inner speech in the 1999 version of the questionnaire.

L2 INNER SPEECH: WHAT LEARNERS SAY

151

is the only person who has control over what is reported, the frequency of the reporting, and the veracity of what is written. As noted, some learners reported exceedingly numerous occurrences of certain types of inner speech. At the same time, some students wrote a lot of entries whereas others wrote just a handful. The quantitative analysis of diary data from multiple participants should thus be approached with care, keeping in mind that beneath some total frequencies and percentages there are individual human beings with particular ways of experiencing and reporting inner speech behavior. The problems associated with awareness and recognition of such an elusive event as inner speech should also be taken into account when analyzing diaries. It is possible that students in this study might not have paid attention to some of their L2 inner speech or they might not have recognized what was going on in their minds as instances of inner speech. Therefore, one should bear in mind that what a participant reports in a diary in terms of his or her inner speech is always a fraction and a reproduction of what went on in his or her mind. In review, however, the diary remains, despite its limitations, a very fruitful way of gaining access into covert language behavior and a rich source of insights into the nature of hidden speech processes, in particular those that are available for conscious awareness and examination. A final limitation that should be mentioned in regards to the three types of verbal report data discussed in this chapter is that they can only yield information on inner speech processes that are available in consciousness. Verbal report data, whether introspective or retrospective, cannot go into unconscious, inaccessible inner speech processes, such as how meaning is vested on thoughts and to what extent the LI or the L2 is implicated at the deep conceptual stage. Thus, when looking at the students' comments and responses to the verbal report instruments, it should be borne in mind that they refer only to those processes and experiences that were under the focus of the subjects' attention. CONCLUSION This chapter has examined "what learners say" about the inner speech they experience in the L2. Two research projects based on verbal report data and conducted by the author were reviewed. The first project, founded on questionnaire and interview reports, consisted of two separate studies (Guerrero, 1994,1999) on the nature of inner speech as mental rehearsal of the L2 at various levels of ESL proficiency. The studies determined that inner speech in the L2 is a widespread phenomenon among L2 learners and that it increases with proficiency in the language. In addition to shedding light on several linguistic characteristics of L2 inner speech, the studies explored seven rehearsal-related functions: mnemonic, instructional, evaluative, preparatory, dialogic, play, and affective. Although much of the learners' L2 inner speech appeared to be related to their English classes, class-unrelated inner speech in the L2 augmented with proficiency in the language. The coexistence of the LI and the L2 in the learners' inner speech, as reported by the participants, suggests the native language remains a

152

CHAPTER 5

powerful cognitive and linguistic tool even among the highest levels of L2 development. The second project described in this chapter was a study of the early stages of L2 inner speech based on diary data and the stimulated recall technique. The study revealed four main types of covert speech activity among low-proficiency L2 learners: (1) concurrent processing of language being heard or read, (2) recall of language heard, read, or used previously, (3) preparation before writing or speaking, and (4) silent verbalization of thoughts for private purposes. These early attempts at internalizing external social L2 speech seem to be indispensable for the eventual development of L2 inner speech as a tool for thought. The last section of this chapter took a critical look at the methodology employed in the studies, namely, the questionnaire, interviews, and diary combined with the stimulated recall technique. The strengths and weaknesses of each of these types of verbal report methods, particularly as related to the covert and elusive nature of inner speech, awareness of the phenomenon, and retrievability of the inner speech event, were considered.

CHAPTER 6 AN INTEGRATED VIEW OF THE ORIGIN, NATURE, AND DEVELOPMENT OF L2 INNER SPEECH

Based on the preceding review of research in both LI and L2 inner speech and anchored in sociocultural theory principles, this chapter offers an integrated view of the origin, nature, and development of inner speech in the L2. The ontogenesis of L2 inner speech as a process of internalization of L2 social speech is examined in the light of research on L2 private speech and on the transformations that external L2 speech undergoes as it is internalized. The microgenetic process of externalizing thought through L2 inner speech is discussed, and a model of L2 speech production including an L2 inner speech phase is proposed. Attention is also given in this chapter to aspects of form and functions of L2 inner speech: its syntactic, semantic, and phonological structure as well as its thinking and rehearsal macro functions. The role of inner speech as mediator of verbal tasks in the L2 and changes in the nature of L2 inner speech related to growing proficiency are noted. Finally, the focus is turned to the impact that development of an L2 inner voice has on the creation of an L2 identity. INNER SPEECH AS INTERNALIZATION OF THE L2 One of the processes that Vygotsky signaled as crucial for the development of higher psychological functions was that of the internalization of social speech and its transformation into inner speech. Vygotsky regarded the internalization of social speech as a paradigm of the transformation that cultural, external processes and artifacts undergo as they are turned into psychological processes and tools. The internalization of social speech, Vygotsky stressed, is a long and gradual transformative process entailing changes in function as well as structure. As social speech turns into speech for the self, first in the form of egocentric (or private) speech and subsequently inner speech, there is a gradual shedding of formal, social traits: Expanded syntax becomes predicative and abbreviated, socially explicit and conventionalized meaning evolves into condensed personal idiosyncratic sense, and audible, fully articulated vocalization turns into soundless, elliptical enunciation. Among monolinguals, the internalization of social speech entails the "ingrowing" (Frawley, 1997, p. 95) of a specific system of signs and communicative practices

154

CHAPTER 6

known as the LI and its application as mediator of higher intellectual functioning. The internalization of the LI communicative code results in LI inner speech, or inner speech mediated by the LI. The development of LI inner speech has been thoroughly documented, particularly through the observation of the private speech of monolingual pre-school children (Berk, 1992; Goudena, 1992; Kohlberg et al., 1968; Pellegrini & DeStefano, 1979; Weir, 1962). Research has shown that LI private speech tends to decline in frequency at around school age, an event that signals the capacity to "think" words through covert, inner speech (Kohlberg et al., 1968; Patrick & Abravanel, 2000; Vygotsky, 1986). What happens, however, when an additional language is introduced in the social environment of the individual? Does inner speech develop in the additional language, just as it develops in the LI? What does research say about the inner speech that might result among children growing up bilingually from birth? What types of inner speech develop among children as they begin to learn an L2 in school? What happens when an L2 or FL is introduced post puberty or in adulthood? Can bilingual or multilingual inner speech take place? These are some of the questions that come to mind once the possibility of L2 inner speech is contemplated. In this chapter, an attempt is made to address these issues on the basis of relevant available information. Private Speech as a Transitional Phase in the Internalization ofL2 Social Speech Studies of children raised in bilingual environments indicate that internalization of two languages can occur before school age. Early evidence of this can be found in Leopold's (1949) diary of his daughter Hildegard' s acquisition of English and German from birth. During her childhood, Hildegard was exposed to the two languages, with periods in which, because of family travels between Germany and the United States, one language was more dominant than the other in the environment. Hildegard's use of the two languages in social speech tended to shift accordingly. Interestingly, so did the private speech she engaged in while playing alone. After a prolonged stay in Germany in which German took over English as Hildegard's main form of social communication, Hildegard is reported as having a bit of preparatory private speech in English: "Before she went over to call for her friend, she asked to have the English sentence which she was to use said for her, and practiced it in several repetitions" (p. 124). By age 5:7, however, an entry in Leopold's diary indicates the girl's use of both languages in private speech: "The great preponderance of practice in English has pushed back the German. It is still intact subconsciously. The other day I heard her speak German in a dream.... When she plays alone she speaks English because her solitary games continue her games with other children" (p. 127). This private use of both languages, in dreams and solitary play, suggests that the two languages were on their way to becoming or had already been internalized as inner speech. Children who start learning an L2 after their LI is well developed also display use of private speech in the L2. Such is the case, documented by Hakuta (1986), of Uguisu, a Japanese girl who started acquiring English after moving to the United States at the age of five. Uguisu picked up her English mostly through interaction with other

AN INTEGRATED VIEW OF L2 INNER SPEECH

15 5

children at kindergarten. Her first use of English phrases were "usually imitations of what her friends had just said" (p. 107). It was after seven months of exposure to and interaction in English that Uguisu began to use the L2 in a variety of social and private contexts, including "when she was playing on her own, such as in the bathtub with her toys" (p. 108). Uguisu's private use of the L2 during solitary play suggests that interiorization of the L2 was on its way. Chen (1988) also observed use of both the LI and the L2 in the private speech of a Chinese child learning English as an L2 in nursery and kindergarten. The boy in this case used the two languages privately for all the functions in the initial stages of L2 acquisition, yet shifted gradually to the L2 in most functions with greater competence of the L2. Attempts to internalize the L2 may also be seen in the private speech of children learning an L2 in school contexts. Evidence of this has been documented in SavilleTroike (1988), de Courcy (1995), and Broner and Tarone (2001). In Saville-Troike's (1988) study, 6 of 9 children of Chinese, Japanese, and Korean background who were enrolled in nursery and elementary school classes displayed intrapersonal, rather than interpersonal, language learning strategies involving the use of private speech in English (the L2) while ostensively undergoing a "silent" period. The children's intrapersonal learning strategies included (a) selective repetition of others' utterances, sometimes evincing a focus on form, sometimes on meaning; (b) recall of words or phrases heard previously and practice; (c) creation of new linguistic forms, sometimes combining the L2 with the LI, sometimes inventing new words in the L2; (d) substitution of words and expansion in sentences; and (e) rehearsal before speaking aloud to others. Following are examples of (a) and (e): (a) Repetition of others' utterances Child 1: Pooty. S: Pooty. Child 2: Pooty? S: Pooty? Child 3: Hey, look. S: Hey, look. Child 2: What are you doing? S: What are you doing? (p. 579)

(e) Rehearsal before performance Child to self: Triangle, please. Triangle. Teacher to child: Yoshi, what shape do you want? Child to teacher: Triangle. (p. 576)

Interestingly, in the example for (a) above, S (the subject) had his back turned to the other children. As Lantolf (2003) points out, this is a form of internalization by eavesdropping on others' interpersonal discourse. Saville-Tro ike (1988) characterized the children who made most use of private speech as being inner-directed, that is, as learners who "approach the learning task as an intrapersonal task, with a predominant focus on the language code" (p. 568) and who are likely to be the most successful in learning the L2. In contrast, other directed children "approach language learning as an interpersonal, social task, with a predominant focus on the message they wish to convey (p. 568). Saville-Troike's study confirmed earlier observation that during the so-called "silent period" inner-directed learners "were not merely passively assimilating second language input, but were using private speech in an active process of engagement with input data" (p. 568).

156

CHAPTER 6

Evidence of the internalizing function of private speech in the L2 within school settings is also provided by de Courcy (1995). Two studies conducted by the author yielded data on the use of private speech by students in Year 9, the 2nd year of a 3-year late French immersion program in Australia. Interviews with the students indicated they were using French in their private and inner speech, which the author takes as a sign of their emergent bilingualism. Two forms of the students' L2 private and inner speech were answering teacher's questions to themselves and rehearsing silently what they were going to say. Internalized speech in French took place inside and outside the immersion setting. De Courcy reports the students underwent four developmental phases in the process of understanding the language they were immersed in: First, they relied on translation as a receptive strategy; second, they focused on recognizing key words in the language around them; third, they were able to relax as they listened and read for the main idea; and fourth, they could comprehend without using any conscious strategy. These four phases suggest that there are different stages in the progressive internalization of the L2 with different manifestations of L2 inner speech along the way (as noted by Sokolov, 1972): from intentional focus on isolated words and laborious translation into the LI to automatic processing (with comprehension) of the L2. In their study of language play among fifth-grade learners in a Spanish immersion classroom, Broner and Tarone (2001) found instances of L2 private speech in the form of what they call language play forfun and language play for rehearsal. According to the researchers, both types of language play "occur as a process in which learners appropriate the L2 speech of others in interaction and internalize it" (p. 366). In the private speech used for rehearsal purposes, the learners were using the L2 for the "serious" purpose of practicing target language forms and new vocabulary. In private speech for fun, the children were engaging in language manipulation mainly for selfamusement. This type of private speech was usually accompanied by smiles and laughter as well as changes in voice quality and pitch and consisted predominantly of language forms already known by the learners. In some cases, private speech seemed to function both for rehearsal and fun. The authors contend that both types of language play, for fun and for rehearsal, are important for language acquisition. The internalizing function of private speech has also been observed among adult L2 or FL learners (Centeno-Cortes, 2003; Lantolf, 2003; Lantolf & Yanez, 2003; Ohta, 2001). Centeno-Cortes (2003) investigated the private speech of 3 Spanish L2 learners as they were learning the language both in North America and Spain. Centeno-Cortes was able to identify five functions of private speech: internalization: to work on some language feature the learner wishes to acquire private rehearsal: to practice for future public production participation: to participate covertly in public production task-related: to maintain or regain self-regulation during a task affective: to indicate understanding, disappointment, or enthusiasm Some instances of private speech were produced only for the purpose of internalization, others had not only the internalizing function but one or more of the other four functions as well. Internalization, therefore, emerged as the most important

AN INTEGRATED VIEW OF L2 INNER SPEECH

157

function in the data. A remarkable contribution to the study of private speech as intemalization made by Centeno-Cortes was her being able to show links between intemalization and social production. In other words, she was able to show that intemalization had indeed taken place by demonstrating that what was internalized through private speech mediation was later used as social speech. A student, for example, is shown pronouncing the Spanish word acento with English phonology, as a[ks]ento, twice before she is shown privately imitating the teacher's pronunciation of the word as a[6]ento (that is, using the dialectical variety of Spanish of c as /9/). On a later episode, the learner is recorded spontaneously pronouncing the word as afOJento. Ohta (2001) found three main categories of L2 private speech as intemalization or, as she put it, for "the classroom acquisition of foreign language" (p. 36), among her English-speaking college students of Japanese: vicarious response, repetition, and manipulation. In vicarious response a learner produces to himself or herself an answer to a question or prompt addressed to another person or to the whole class. Why is it argued that such behavior is indeed a sign of intemalization and that it may contribute to acquisition rather than simply being a private display of knowledge? Ohta (2001) suggests vicarious response works as hypothesis testing, that is, as a means to privately try out language forms learners are not yet sure of. In receiving confirmation or disconflrmation of their hypotheses, learners may be strengthening knowledge of and thereby internalizing correct language forms. Another concept invoked by Ohta (see also Centeno-Cortes, 2003, and Lantolf, 1997) in explaining the internalizing effect of vicarious response is the ZPD, the Vygotskyan notion that it is with social assistance from others or socially constructed mediational means that learners may advance in those areas that are in the process of being developed. During vicarious response the learner creates his/her own internal ZPD by enacting a private dialogue or internal dialectic through which he/she can compare his/her own provisional L2 forms with those of others in the classroom. The following example from Lantolf and Yanez (2003) pointedly illustrates the dialectical nature of vicarious response. Teacher: Mas autosfueron vendidos el ano pasado. Student 1: Se vende mas autos. Teacher: Se what ? Student 2: Se vendieron. Learner: [Quietly repeating to herself] Vendieron ... I knew it. (p. 106)

While the teacher and Students 1 and 2 publicly negotiate the correct form of the Spanish verb "vender," the learner is probably thinking to herself through inner speech what the answer could be. The learner's private utterance of the correct choice and her almost triumphant, though sotto voce, exclamation "I knew it" are evidence of the internal dialectical process of confirmation and possibly L2 intemalization that she went through. Repetition or imitation of others' utterances has emerged in Ohta's (2001) and others' research as an important internalizing device in L2 private speech. Repetition was, in fact, the most common form of the private speech generated in Ohta's study of classroom learning of Japanese. Two crucial observations are made by Ohta regarding

158

CHAPTER 6

private repetition. First, learners "do not repeat everything they hear" (p. 54). Rather, learners tend to repeat selectively focusing on those items that are possibly new or difficult to them and that they "are working to acquire" (p. 54). Second, private repetition occurs not only immediately after a model has been produced but also later, even after the public focus of attention has changed (p. 59). A student in Ohta's data, for example, dutifully repeats aloud a newly introduced Japanese word by the teacher: "kutsushita" (socks). The learner, however, continues to repeat the word to herself even after the teacher and the other students have turned their attention to other clothing accessories such as shoes ("kutsu"), sneakers, sandals, and bag. What is also interesting to note is that the learner does not just parrotlike reproduce the word each time she returns to it but seems to be working with it at a semantic level, building on her understanding that the literal meaning of the word is "under-shoes." This can be seen in two of her delayed repetitions: "Kutsu ohh ano: shita" (shoes o::h uh: under) and "Kutsu kutsushita" (shoes socks) (p. 60). Repetition or imitation in language learning, as in the learner case just noted and as Lantolf (2003; Lantolf & Yanez, 2003) maintains, should not be construed as mindless mimicking of model behavior but as a constructive and transformative process denoting agency and intentionality on the learner's part and leading to the reconstruction, and not just the replication, of external language on the internal plane. Manipulation of language forms is another frequent private speech strategy for internalization, employed by adults (Lantolf, 2003; Ohta, 2001) as well as children (Broner & Tarone, 2001; Saville-Troike, 1988). Manipulation entails experimenting with language forms at the morphological, syntactical, lexical, or phonological level. Multiple manifestations of language manipulation have been observed: breaking up words, word compounding, morpheme substitution, sentence expansion, practice with sounds, and others. In Ohta, for example, a learner experiments with the adjective warui (bad): [in a very soft voice] warn- waruku (.) waru : : ku (.) ku warui::. As Ohta explains, the learner "breaks the adjective warui down into its stem waru-, builds the adverbial form waruku, then returns to the nonpast form of the adjective waruf (p. 64). Importantly, Ohta points out that manipulation, repetition, and vicarious response do not usually occur as discrete categories. As learners repeat or answer vicariously, they often experiment creatively with language. The creative shuffling and reshuffling to which learners subject the L2 in their private speech demonstrate the transformative and agentive nature of L2 internalization. Among private speech researchers, Lantolf is the one who has most forcefully emphasized the role of private speech in the internalization of the L2. Lantolf s (2003) thesis is that private speech as internalization represents language learning "in flight", borrowing Vygotsky's (1978) metaphor for describing the microgenetic processes that can be observed in the development of higher intellectual functions. In their data, Lantolf and Yanez (2003) found, like Ohta (2001), private speech in the form of selective repetition of L2 forms and vicarious response. In addition, they produced evidence of private use of the LI to make sense of the L2 and use of metalanguage to understand the L2. In the following example, the learner whispers the grammatical label "imperfecto" as she searches for the correct verb form:

A N INTEGRATED VIEW OF L2 INNER SPEECH

15 9

I need to change, I need to change fue . . . seria [20-second pause] Imperfecto [whispered] (p. 106)

The use of metalanguage in private speech by university students, according to the authors, is not surprising given the emphasis granted to metadiscourse in FL classes. A distinctive feature of internalization of the L2 among adult classroom learners, which may not be present among pre-school children or people learning the L2 in noninstructional settings may thus be conscious and deliberate objectivization of the language to be internalized. Whether such level of metalinguistic awareness in private or inner speech is beneficial for learning the L2 remains to be seen although there is evidence that the introduction of literacy, with its deliberate focus on language as an object of study, has a positive impact on inner speech development (Campbell, 1992; Flavell, et al., 1997). To sum up, then, given the significant mediating role that private speech apparently has in the learning of an L2, is private speech in the new language a necessary transitional phase in the internalization of social speech as inner speech? It would be tempting to think so-such an idea would be congruent with Vygotsky's hypothesized trajectory of speech forms in ontogenesis. But the fact is that learning an L2, unless it is learned together with an LI from birth or is introduced at an early childhood, implies developing the ability to think through another language after this ability has been developed in the LI. Adult language learners as well as older children and adolescents, unlike small children, have already differentiated between social and inner speech functions in the LI. There comes a time when people realize, in Vygotsky's (1986) words, that "inner speech is speech for oneself [and] external speech is for others" (p. 225). This time coincides with the decline of private speech and the emergence of inner speech in children. In the LI, private functions-verbal thought, self-regulation, self-communication-are normally carried out through inner speech. Of course, grown-ups do not lose their ability to access earlier ontogenetic forms of behavior, and they do, in fact, by the principle of "continuous access" (Frawley & Lantolf, 1985) revert to vocalized private speech when cognitively challenged. The key question, however, is whether adults or older learners who have already learned the crucial distinction between social and inner speech need in fact to go through a period of overt private speech in the L2, just as they do in LI acquisition, in order to internalize the new language. Ohta (2001) has pointed out that oral rehearsal involving articulatory activity may be necessary in the development of inner speech. She states: "For the foreign language learner, the physical utterance of the language being learned may be a necessary component. Involvement of the articulators is likely to be necessary for efficient development of phonological analysis and processing" (p. 67). The few L2 "private speech as internalization" studies that exist show not only great variability in the frequency of private speech among learners-some display very little use, others show considerable use-but also that some learners do not engage in private speech of the L2. In Saville-Troike's (1988, p. 586) study 3 of 9 child subjects did not produce any audible private speech, in Centeno-Cortes (2003)'s research lof the original participants was dropped because he did not produce any private speech, and in Ohta's (2001, p. 38) study some students engaged in only minimal amounts of

160

CHAPTER 6

private speech.59 Saville-Troike attributes the absence of private speech in the L2 among 3 of her subjects to their other-directed or socially-oriented learning style, in other words, to their preference for social interaction and their interpersonal, communicative approach to language learning. Another possibility, that Ohta suggests (p. 68) and that seems to be a plausible one especially in the case of adult language learners, is that some learners may refrain from using private speech out of fear that it might be considered socially inappropriate and may opt instead to conduct their oral practice in other ways, for example, in speaking activities during class. Yet the most likely cause, also mentioned by Ohta (p. 68), for the apparent absence of private speech rehearsal appears to be that learners are instead engaged in inner rehearsal, inaudible to observers. Covert verbalization, with minimal activity of speech articulators, and activation of auditory or phonological imagery might be sufficient for some learners to obtain the oral practice that is necessary for internalization to take place. If one strictly defines private speech as audible, vocalized speech for the self, then the answer to the question of whether audible private speech is an indispensable phase in the internalization of social L2 speech appears to be "no."60 The internalizing function of such private speech as has been observed among some learners might well be taken over among others, particularly adults, by rehearsal-type inner speech. Galperin (1967), in developing the notion of the internalization of external actions, suggests that the first type of "action in the mind" is "external speech to oneself (p. 30). External speech to oneself, according to Galperin, is soundless speech or "ordinary speech without the volume" (p. 30), a transitional phase that temporarily serves as mediator of mental action while the more permanent stage of inner speech is being formed. When applying these notions to the internalization of an L2, external speech to oneself may be understood as an incipient form of inner speech or, as Swain (2000) puts it, "the beginning of inner speech" (p. 114) in the L2. For many adult learners, soundless speech for oneself in the form of mental rehearsal might be the only socially acceptable, socially "mature," or even socially safe way of practicing the language, aside from social interaction. Notwithstanding what has been said, it is also important to point out that some L2 features might indeed require some form of external, unfolded enunciation. For instance, as MacKay (1992, p. 142) observes, some FL sounds, such as a German trilled /r/, are internally unrehearsable for someone who does not know the language and thus need to be practiced through fully articulated, oral speech. At any rate, the scarcity of studies on L2 and FL private speech as internalization among children and adults and, particularly, the lack of such research in non-instructional settings prevent one from definitively concluding one way or another on the need for an explicit private speech phase in the internalization of the L2.

59 Of course, one hastens to point out that it is very difficult to generalize from these findings since the absence of private speech during the taping sessions does not rule out the possibility that the learners might have engaged in L2 private speech when not observed. 60 Private speech in the L2 may, however, be a widespread and perhaps necessary phase among preschool bilingual children or L2 learners who have not yet distinguished between social and inner speech.

AN INTEGRATED VIEW OF L2 INNER SPEECH

161

Early Inner Speech Manifestations in the Inter nalization of an L2 In sociocultural theory, social forms of activity and socially mediating cultural artifacts are thought to undergo a process of transformation as they are recreated on an internal plane. How does this transformation occur in the case of L2 external speech activity? More simply put, how does L2 social speech become L2 inner speech? Aside from the phase of private speech in the L2, which may or may not occur (as discussed above), there are several early manifestations of inner speech, or soundless self-directed speech, which suggest how an L2 and its speech activity are transformed as they are internalized. The diary study conducted by Guerrero (2004) and summarized in Chapter 5 documented some of these early signs of L2 inner speech among adult L2 learners. Three important ways in which L2 social speech re-emerges on an internal plane in the early stages of learning an L2 are (a) the inward recreation of the L2 that takes place concurrently with ongoing L2 affordances,61 (b) the recall and delayed reprocessing of L2 speech (not concurrent with ongoing external L2 speech activity), and (c) preparatory use of the L2. Internalizing Processes Concurrent with Ongoing L2 Affordances One of the most salient processes denoting the learners' efforts at internalizing L2 social speech is the inward reproduction of the L2 that takes place as learners are listening to or reading L2 material. Sokolov (1972) described this stage in the genesis of inner speech as the "psychological transformation of external speech, its 'internal projection,' arising at first as a repetition (echo) of the speech being uttered and heard" (p. 1). Inward repetition was very frequently reported in Guerrero (1994,1999,2004). Repetition was also the most common form of private speech among the adult learners in Ohta (2001) and one of the first manifestations of private speech among the children in Saville-Troike (1988). Based on his experiments involving the reading and listening of difficult, and sometimes FL texts, Sokolov (1972) argued that echoic repetition is a necessary phase in the development of inner speech. The role of inward repetition is fixation in memory, a process that takes place as attention is focused selectively on certain words and comprehension is allowed by virtue of the semantic analysis and generalization that readers or listeners impose upon those words. Fixation in memory or remembering will not occur if repetition is done word for word or if it is prevented by artificial impairment of subvocal articulation (see speech interference techniques described in Chapter 4). As texts or oral speech become easier to understand the need to repeat subvocally diminishes and inner speech becomes more abbreviated: Words are reduced phonologically to mere hints of sounds or letters and text is compressed to key words and generalized ideas. In the following example of private speech from Saville-Troike (1988), a child repeats only the topic of a sentence (notice also the

61 As a substitute for the mechanistic term "input," associated to the information-processing paradigm in the SLA field, van Lier (1996) has suggested the concept of "affordances," meaning the linguistic opportunities offered by the environment and perceived by the learner in a process of interaction and engagement.

162

CHAPTER 6

omission of the initial Is/ sound): Teacher: Squirrels like to eat these, don't they? Child (to himself): Quirrel (p. 579)

When affordances are difficult to understand, as in the case of beginning L2 or FL learners, fragmentary inner speech is not enough. In this case, the speech that is heard or the text being read has to be inwardly reproduced in an unfolded manner, that is, with maximal subvocal articulation (Sokolov, 1972, p. 263). The importance of subvocal repetition in acquiring new words has been highlighted in studies of the phonological loop and elaborative rehearsal, as discussed in previous chapters. Repetition of L2 material frequently involves imitation, especially during listening. Students in Guerrero (2004) indicated that they tried to "imitate" the words they heard and sometimes even the "accent" of speakers they heard on TV. Vygotsky (1986) thought that "imitation is indispensable" in learning to speak a language as well as in other forms of activity (p. 188). He distinguished, however, between "drill imitation," which is nothing but "automatic copying," and "intelligent, conscious imitation" (p. 188). The latter is characterized by understanding what is being imitated and involves deliberate, instant reproduction of behavior without need for repetition. Drill imitation, on the other hand, is repetitive and unintentional and does not imply comprehension. Vygotsky assigns conscious, intelligent imitation a crucial role in learning within the ZPD. Through imitation of others' behavior, the learner performs vicariously at a more advanced level of development. Eventually, the learner internalizes the imitated behavior and is capable of performing it alone. Conscious imitation with understanding is transformative because it leads the individual to new developmental levels. Vygotsky, however, notes that imitation can only take place in the ZPD. Learners cannot imitate everything. They can imitate only what is within the limits of theirZPDs(p. 187). It is plausible that the two forms of imitation signaled by Vygotsky may occur covertly in L2 learning: both repetitive, drill imitation without understanding and the more active, intentional imitation of L2 material with a focus on comprehension, as Sokolov (see above) pointed out and research on elaborative rehearsal suggests (see Chapter 1). Drill, mindless imitation, does take place in L2 classroom "learning" (although the example that follows can hardly be called learning). As Murphey (1990) points out, there is what he calls "the song stuck in the head phenomenon," an event that frequently consists of repeating in the head songs sung in a language that one hardly understands. The conscious, intelligent imitation of L2 material, on the other hand, may have an internalizing function; that is, it may help students operate at an intramental level of phonological or lexical ability that is currently beyond their level of independent performance. It is possible, however, that imitation of certain phonological features may require external vocalization or rehearsal, as MacKay (1992) contends, and thus intramental imitation may not be enough. This may explain why some learners report that in their inner speech everything "sounds" perfect but when they externalize that speech it does not "come out" as good. The important fact about conscious, intelligent imitation, whether covert or overt, is that it leads to the

AN INTEGRATED VIEW OF L2 INNER SPEECH

163

creation of a new plane of L2 performance, the plane of internal, mental operations in the L2. According to Baddeley et al. (1998),"imitation of novel phonological forms may indeed serve to promote the long-term phonological learning of new words, possibly by increasing the period over which they are held in the phonological loop" (p. 163). The inward reproduction that occurs as learners hear or read in the L2 may thus consist of simple repetition (or imitation) or of more complex forms of attending to language. One of these forms-in the case of alphabetic languages-is attention to the orthographic constitution of words by spelling words in the mind. This may be achieved by mentally visualizing words or trying to figure out their spelling by their pronunciation. According to MacKay (1992), the visual reconstruction of letter strings in the mind is not a simple phenomenon and is based on complex acoustic, phonological, and visual processes not yet fully understood (p. 141). Attention to the phonological and graphic form of words does not seem to occur for its own sake. Rather, the visual reconstruction of phonological material may be tightly related to semantic processes. In Guerrero (2004), students reported trying to read the written captions on TV (in the L2) to understand the on-going oral speech (in the L2). By "seeing" the way a word is spelled, students may get a better grasp of their meaning. For instance, the English word "hospital," which is a cognate in English and Spanish, is probably better understood in its written form by a Spanish-speaking beginning ESL learner than in its pronunciation. Internalizing language as a social artifact implies grasping the complex elements of words in all their sound, meaning, and graphic aspects as well as in their grammatical relationships. The students' concurrent covert verbalization of oral or written L2 is thus often accompanied by behaviors that signal the students' deliberate attempts at "cracking the code." In addition to spelling words, already mentioned, learners in Guerrero (2004) most frequently attempted to understand the meaning of L2 words by translating them into the LI, breaking them up in parts, associating them to known referents, looking them up in a dictionary, or asking someone about their meaning. The learners' focus on meaning as they attend to oral or written L2 was also evident in the immersion study conducted by de Courcy (1995), who found that the first strategy (over time) the students employed to make sense of the language around them was translation into the LI and the second one was identifying key words. Some of the learners' ongoing efforts at making sense of the L2 are externalized, for example, by looking words up in a dictionary or muttering the LI equivalent of a word through private speech. Some meaning making strategies, however, are covert and probably better suited to inward verbalization. Such is the case of associating L2 words that are heard or read to concrete objects, past events, or visual images. Mentally associating words is an activity that implies swift changes in cognitive domains, activation of multiple sensory modalities, and retrieval of past experiences, all behaviors that seem particularly well served by inner, rather than overt, speech. Because inner speech is exceedingly fast behavior that works on the basis of abbreviated syntax, inhibited articulation, and condensed semantics, it has the potential for evoking more than can be expressed in words. As Vygotsky (1986) would put it,

164

CHAPTER 6

to understand the meaning of a word, one must attempt to reach inside the microcosm that it represents (p. 256). This does not mean that associations cannot take place during private speech, but inner speech seems to be an ideal venue for rich and rapid associations to occur. MacKay (1992) believes that because inner speech occurs at a much faster rate than over speech, there might be "a possible benefit of internal rehearsal relative to overt rehearsal" (p. 131). Recall and Reprocessing in the Internalization of the L2 In addition to the inner speech that learners experience as they are listening to or reading in the L2, there is the inner speech that is activated as learners recall L2 speech seen, heard, or used by them before. Saville-Troike (1988) referred to this type of behavior in private speech as "recall and practice." It is a type of inner speech that involves retrieving language from memory and subjecting it to re-processing in various ways. The fact that students return to or revisit the language they have heard, read, or used before indicates that internalization is a prolonged, extended process rather than an instantaneous one. To internalize a language feature may mean having to submit the data to processing and reprocessing. Recall and reprocessing can occur spontaneously without conscious efforts, as suggested by the learners' reports in Guerrero (2004) indicating that L2 words suddenly popped into their heads. Also involuntary (and quite annoying to the person who experiences it) seems to be the occurrence of playback, that is, the insistent replay of L2 words in the mind that is often described as "getting words stuck in the mind." According to MacKay (1992), "why inner rehearsal is sometimes involuntary and difficult to control" (p. 122) is one of the unexplained phenomena of inner speech. Recall, however, can also occur deliberately as learners try to remember a word or silently repeat language they have come into contact with before. In some cases, the learners' recall may take on a heteroglossic nature as they remember L2 segments in the voice of others. One student in the diary study (Guerrero, 2004) reported: "At home I could remember the voice of the little girl who spoke very funny in the story that I heard in the lab." Students often take a metalinguistic stance as they recall L2 in their inner speech. In Guerrero (2004) it was not infrequent for students to report deliberately retrieving L2 forms and reflecting on them or recapitulating their knowledge about these forms. One student, for instance, reported: "I remembered the words shy and cheap and their different [initial consonant] pronunciation, which the teacher explained. I understood that you have to have a good ear to distinguish words in their pronunciation. I need to practice pronunciation more." (This comment was written in Spanish, the learner's LI, except for the words in boldface that were written in English.) Students sometimes used grammatical labels in their metacomments, referring to "adjectives," "possessives," "the plurals," and so on. The issue of having metalinguistic thoughts about the L2 has been raised by Cohen (1998), who points out that when learners are thinking about the L2 most of their metalinguistic thoughts are in the LI and that the L2 in these instances is used only in a fleeting or limited way (p. 167). Cohen's point

AN INTEGRATED VIEW OF L2 INNER SPEECH

165

is well illustrated in the example cited above, where only a few words actually came to the students' minds in the L2 and the rest is in the LI. It appears, then, that a large portion of the students' inner speech, at least in the beginning stages of L2 acquisition, consists of thinking about the L2 in the LI, rather than through the L2, as Lantolf contends (cited in Cohen, p. 173). In the reprocessing of L2 material, learners often put their knowledge of the L2 to the test by applying what they have learned in novel constructions or new situations. In the data for Guerrero (2004), a student wrote: "After class, I kept practicing how to tell the time. I even made sentences in my mind using different times." Learners will sometimes try to see whether they can "say" things in the L2 even without the need to communicate what they think, for example, "I wanted to go to the movies and I tried to think how I would say this in English: I will go to the movies this afternoon." Extending the internal practice of the L2 by using it in new contexts is one way in which students internalize the L2. Perhaps the notion of "appropriation," in the sense defined by Wertsch (1998; see Chapter 1 in this book), is the best way to conceptualize the internalization process that takes place as students try out their knowledge of the L2. In doing this, they seem to be deliberately appropriating the L2 (making it their own) and turning it into a tool for thought for their own cognitive purposes. Preparatory Inner Speech as Internalization Learners have been observed to engage in preparatory rehearsal of the L2, that is, mentally planning what they are going to say and how they are going to say something in the L2 at a future communicative event. Whereas these processes are obviously related to the externalization of speech, it can be argued that they can also be conducive to internalization. The question is whether this sort of mental "output" can help learners internalize the new language.62 Swain (1995, 2000) hypothesized that output may lead to learning in three ways: (a) It may encourage "noticing gaps" or differences between target language forms and learners' interlanguage forms and may also promote "noticing holes" or problems in not knowing how to say something when learners want to say it; (b) output may function as hypothesis testing, that is, as a way to test whether hypotheses held about language work in actual production; and (c) output may serve a metalinguistic function, as learners take a reflective stance toward their production. Swain argued that all these functions of output contribute to internalization for several reasons: Output requires a greater mental effort than engaging with L2 affordances; it alerts students on what they can and cannot do; and 62 The word "output" is not used here haphazardly. First, the word is deliberately employed to invoke Swain's "output hypothesis." Second, as with the term "input," one is aware of the conceptual shortcomings of using a word that is so closely associated to the narrow information-processing "input-output" view of L2 learning. Swain (2000) herself has indicated her discomfort with the word "output" and has proposed in turn, for lack of a single term, to talk about learners' production in terms of "speaking," "writing," "utterance," "verbalization," and "collaborative dialogue," (p. 103) or, as Alistair Cumming has suggested, "purposeful language production" (cited in Swain, p. 114). Of all these terms, "verbalization" and "purposeful language production" seem to be the most appropriate to refer to learners' output in inner speech.

166

CHAPTER 6

it encourages learners to fill in the "gaps" or "holes" with new knowledge. The hidden, intentional verbalizations that learners produce in preparation for communicative tasks may have, like overt production, an internalizing as well as externalizing function.63 (The externalization role will be developed further on in this chapter.) All forms of preparatory inner speech, such as those documented in Guerrero (2004), namely, planning what to say or how to say something to someone, planning an exercise response, planning for a conversation the learner knows she will have, and answering questions in the mind, may serve, like externalized output, to notice discrepancies between learner forms and target language norms or models, test hypotheses about the L2, and develop metalinguistic awareness about the L2. The following entry from a student's diary illustrates "noticing gaps" in interlanguage knowledge through mediation of inner speech: The teacher started a new chapter and taught us how to say the parts of the house and the names of appliances. I thought a couple of things I might say that I thought I knew, but I kept quiet because I was afraid they would laugh at me. I didn't dare. I thought I didn't know how to pronounce them. When the words were said aloud, one of the words I thought was correct but the other wasn't.

In the example above, the student not only notices a gap between his own internal production and the model presented in the class, but also tests hypotheses about pronunciation. He confirms one but disconfirms another one. The next example shows the metalinguistic function of preparatory inner speech: When I answered the questions the teacher gave us on the story, I had to look in my mind among the prepositions I knew for the right one to say "Ali is sitting behind Sofia."

The learner in this case strengthens knowledge of the L2 by taking a reflective stance towards his mental production and using metalanguage to describe it. This treatment of language as an object of study may be typical of the inner speech production that learners experience in classroom settings. Evidence of Inter nalization How can one tell that internalization of L2 speech has taken place? The obvious way is by observing L2 speech in external production. If the learner can externally produce language intentionally, meaningfully, and creatively, one must presume that internalization has taken place. But production can also take place covertly, as we have seen in the preceding section. Looking for evidence of internalization in inner language production is not easy. Learners' verbal reports, however, may offer some enlightenment on this issue. Some of the questionnaire and interview comments in Guerrero's (1994, 1999) studies give hints that internalization has occurred: [Comment from an advanced ESL student, written in English] Most of my oral practice in English [the L2] is with friends, but I do find myself talking alone in English and

63 The emphasis that has been put in this section on the internalizing role of L2 inner speech production does not mean that all L2 inner speech is conducive to internalization. Some forms of L2 inward production may simply be attempts at externalizing thought or regulatory inner speech guiding cognitive activity.

AN INTEGRATED VIEW OF L2 INNER SPEECH

167

thinking about things in English. It has become a useful practice and very normal and easy. [A high proficiency ESL student talking about her mental preparation for an oral activity] I was lost. But when I came to the university, the girls explained everything to me, and I could imagine what I was going to say. As soon as they told me I had to talk about colleges degrees, the sentences came very rapidly. [An intermediate ESL student talking about how he prepared to talk about a picture] As I tried to describe the picture, rapidly the words for clothes, the accessories, and the equipment came to mind.

In the above students' reports, terms such as "very normal and easy," "sentences came very rapidly," "rapidly the words . . . came to mind" suggest that the students have internalized much of the L2 and can operate internally quite fluently on the basis of the L2. The following example, by contrast, evidences that a functional internal plane mediated by the L2 has not yet been formed and the learner still relies on his LI to conduct inner speech. [The learner is talking about his preparation for an oral activity.] Sometimes words came to my mind in English, but most of the time I said them in Spanish [the LI], and little by little, I tried to translate them into English.

Further evidence that internalization has taken place and that some inner speech processes are indeed conducted in the L2 is offered by learners' reports about the thoughts they silently verbalize in the L2 for private purposes, that is, without intending to communicate them. (This manifestation of L2 inner speech has been discussed in some detail in Chapter 5). Silent verbalization of private thoughts is the use of the L2 to verbalize feelings, ideas, or intentions to oneself. This form of inner speech is quite common among advanced L2 learners (see later discussions of L2 inner speech and proficiency). But signs that internalization has occurred, albeit to a limited extent, can be observed among low proficiency L2 learners. Two forms of silent verbalizations were identified in Guerrero (2004): (a) spontaneously thinking in the L2 for purely private purposes and (b) imagining conversations. This type of inner speech was the least frequently reported by the participants. Spontaneous thoughts in the L2 usually appeared very suddenly in the mind and were very short and formulaic. Students reported, for example, mentally verbalizing expressions such as "Thank you," "I'm cold," "I'm sorry," and an insult (unmentionable here).64 Imagining conversations with oneself or other interlocutors also occurred, though infrequently, among the low

64

Luria (1981, p. 148) argues that very simple utterances, such as verbal responses to sudden stimuli, epithets, or affective exclamations, as well as echolalic responses to a question, do not represent higher intellectual thinking because they are involuntary, automatic, and do not require a complex motive. For an utterance to indicate true intellectual functioning it must originate in a motive. Luria substantiated this claim by noting that people with major lesions in those brain regions that are associated with forming the intention of an utterance suffer severe impairment of speech production but not of impulsive or abusive exclamations. It is hard, however, to deny intentionality or true motives in the learners' mental use of emotional or formulaic expressions. Although one may agree that those simple utterances do not represent very complex or rational thinking, they do indeed seem to spring from very real, and perhaps important, motives for the learners.

168

CHAPTER 6

proficiency students. Often, students reported having these conversations before falling asleep or immediately after awakening. The students were not very specific about the shape or contents of these conversations. It is possible that the students did not retain these in memory, but it is also possible that the conversations did not involve much substantial L2 production. One learner reported: I was half asleep, half awake, and I was having a conversation in mental English with a person. That person knew English and I didn't. I was trying to explain to this person that I didn't know. Finally, I ended up saying: Talk me slow [sic] and I can understand.

Overview of Inner Speech as Internalization of the L2 Internalization has been viewed in this section as the creation of an internal psychological plane mediated by the L2. Like internalization of an LI, internalization of an L2 proceeds from external social speech and is in part fulfilled by audible private speech. To a large extent, however, and particularly among adult L2 learners, it is through the covert use of the L2 that internalization takes place. Internalizing a new semiotic tool and turning it into a functional tool for thinking is an ontogenetic process that normally takes place over a long period of time. Transforming an external tool into an internal one is not a simple, instantaneous process; there has to be a lot of mental tinkering with the tool before it is fully internalized. Three main transformative processes have been identified in the early stages of internalization of an L2: concurrent inward reproduction of L2 oral speech and written texts, recall and reactivated processing of the L2, and preparatory L2 use. These processes are thought to be indispensable for the development of L2 inner speech and its role as mediator of complex intellectual activity. EXTERNALIZATION OF THOUGHT THROUGH L2 INNER SPEECH As discussed in Chapter 2, one of the processes mediated by inner speech described by Vygotsky was the externalization of thought. In the externalization of thought, inner speech represents the mechanism through which thought finds the words that give it shape. This process requires the presence of a motive as the propelling impulse to thought, an internalized semiotic code to concretize thought, and the transformation of thought, first into condensed inner speech and then into unfolded inner speech, before being externalized as social or private speech. Put succinctly by Wertsch and Stone (1985), externalization is "the transformation of internal activity into external activity" (p. 162). L2 learners have been observed to externalize their inner speech in two ways: as communicative L2 production and as L2 private speech. Communicative social production is one of the most thoroughly researched topics within the field of L2 acquisition. In other words, the focus on learners' discourse, their interlanguage, has occupied and continues to occupy a privileged place in the attention of researchers within mainstream SLA. The focus of these studies is, however, on the (grammatical, phonological, pragmatic, etc.) features of external learner production with little or no

AN INTEGRATED VIEW OF L2 INNER SPEECH

169

regard to the inner speech in which it originates. Bringing inner speech into a discussion of externalized L2 speech has been the task of research conducted within the sociocultural theory approach to L2 learning. One growing strand of research within sociocultural theory concerns the study of private speech as L2 internalization, already discussed in this chapter. Another line of research on L2 private speech focuses on it as a mechanism by which learners externalize their inner order in an attempt to gain and maintain control during challenging intellectual tasks. To explain why L2 students use private speech for selfregulation, researchers such as Appel and Lantolf (1994), Brooks & Donato (1994), Frawley and Lantolf (1985), Lantolf and Frawley (1984), Lantolf, DiCamilla, and Ahmed (1997), and McCafferty (1992, 1994a, 1994b, 1998), rely on the principle of "continuous access" (Frawley & Lantolf, 1985, p. 22; Galperin, 1967, p. 32), the principle in sociocultural theory that holds that ontogenetically prior forms of activity do not disappear but can be accessed for regulatory purposes in moments of cognitive difficulty. Because solving difficult cognitive tasks through the medium of L2 inner speech may be at times beyond the ability of learners, they may find it necessary to verbalize their thinking processes through L2 private speech. Private speech functions in this case as a form of external, rather than internal, form of speech for oneself, organizing and guiding mental L2 activity. The more expanded, explicit, and relatively stable nature of private speech, in contrast to that of inner speech, facilitates higher intellectual functioning by concretizing and stabilizing thought and helping keep attention focused during the solving of particularly difficult tasks. Several linguistic properties have been identified in the regulatory private speech of L2: peculiar uses of tense and aspect, a syntax based on focus, odd pronominalization, externalization of the macrostructure of the discourse, and display of affective markers (see Frawley & Lantolf, 1985, for seminal ideas on L2 private speech features; see also DiCamilla & Anton, 2004; Frawley, 1997; Roebuck, 1998). Private speech as the externalization of mental action is not restricted to oral discourse or adult L2 speakers; it can also be found in certain forms of writing (DiCamilla & Lantolf, 1995; John-Steiner, 1992; Roebuck, 1998), in the speech of bilingual children (Glaessner, 1995), and among adult LI speakers (John-Steiner, 1992). Private speech features have also been found in the gestures that L2 learners produce during oral discourse (McCafferty, 1992). In addition to being externalized as private speech, inner speech is involved in the externalization of thought during L2 production, oral or written. In Chapter 5 it was pointed out that learners often make use of preparatory inner speech before externalizing their thoughts. (It has also been argued above that preparatory inner speech may serve internalization purposes. Here the focus is on inner speech production as externalization.) Evidence from learners' verbal reports in the form of questionnaire responses, interviews, and diary entries was presented, showing that inner speech plays a crucial role in the planning of oral discourse or written texts. Oftentimes, preparation may occur in the face of a specific future communicative event that the learner is aware of, for example, an announced oral report, a scheduled oral interview, or an impending conversation with a teacher. An advanced ESL learner, for instance, commented: "Right now I am doing some tutoring in English and I have

170

CHAPTER 6

realized that sometimes I think how I am going to say something to a student before I am with the student" (author's data for Guerrero, 1999). Preparation may also take place as a rehearsal exercise for imaginary communicative experiences, in the form of "Let's see what I would say if I wanted to make a reservation" or "If I could talk to the author, I would ask her " An intermediate ESL student reported: "I use my inner speech to practice or imagine dialogues with other people. I do a lot of imitation of what I hear on Cable TV and sometimes I learn dialogues that I think are going to help me" (Guerrero, 1990/1991). The preparatory function of inner speech seems to be more widespread among learners of lower proficiency levels than of more advanced levels (Guerrero, 1999). A trilingual learner's comment presented in Cohen (1998) perfectly illustrates the use and decline of the preparatory inner speech function: I planned what I would say and prepared for various scenarios ahead of time in a language-thinking of what words I would use and how to express myself in a situation. It was very helpful and after a few months, I gave it up because I no longer needed to rehearse, (p. 166)

It has been seen that the externalization of thought in an LI is an extremely complex process of assigning meaning and structure to thought through inner speech. To exteriorize thought, the structure of the inner utterance must be in turn semantically unpacked, lexically stabilized, syntactically unfolded, and phonologically encoded. For all its complexity, the externalization of inner speech in an LI appears to be a much less complicated affair than that in an L2. It is known that bilinguals show longer reaction times than monolinguals in verbal responses (Kecskes & Papp, 2000, p. 47). In multilinguals,65 this situation is aggravated, as an entry in Wilga Rivers's (1979) diary of her experience in learning Spanish, her sixth L2, poignantly indicates: When I'm in France I think in French all the time. Here [in South America] I think in English and so have to switch consciously to Spanish when I need to speak it. This makes for a slow reaction, then a slow construction of the message, and occasional false starts like coming out with greetings in Italian or German or English, (p. 78)

In the externalization of an L2, both the LI and the L2 have to be accounted for as part of the inner speech process that mediates the transition between thought and external utterance. One speech production model that integrates the L2 as an inner speech component is A. A. Leontiev's (1981) schema of speech production in an FL, discussed and diagramed in Chapter 3. This schema locates the FL at a translation phase in the post-programming stage, that is, after encoding in the LI has taken place. Leontiev's model, it should be pointed out, is a hypothetical representation of the role of an FL in speech act production at "the early stages of mastering a foreign language" (p. 26). This representation, therefore, conceives FL externalization as basically a translation process, from LI to FL, occurring as an intermediate stage between LI encoding and motor implementation of the FL utterance. Leontiev notes that this 65 Very little is known about the inner speech of multilingual speakers. Further research in this area would not only be fascinating in itself but also help expand knowledge of the much neglected field of multilingualism and contribute to overcome the present bias toward one-language L2 or FL learning.

AN INTEGRATED VIEW OF L2 INNER SPEECH

171

translation process is not automatic and that "the learner will not immediately or without effort come up with the foreign equivalent to the utterance in the mother tongue, remember the rules, and successfully apply them" (p. 27). Rivers's report quoted above perfectly illustrates the intricate translation work that exteriorizing an L2 represents for a multilingual individual who starts learning a new language. Research on bilingual or multilingual competence is congruent with A. A. Leontiev's model of speech production at the early stages of language learning. According to Kecskes and Papp (2000), adult learners at an early phase of FL learning mediate their knowledge of FL words through the LI (p. 64). Although Kecskes and Papp do not explicitly address the issue of inner speech in FL speech production, their view of the bilingual/multilingual mental lexicon would imply that in the process of externalizing thought learners at an early phase of FL development access the LI first and then re-code LI words into the L2. At more advanced levels of FL competence, however, learners "rely more on concept mediation between the two languages . . . and can integrate meaning across languages" (p. 65). The researchers note that although new conceptual links for the FL may be formed, the associations between concepts and LI words will always remain and are usually stronger than those of the FL (p. 66). What does this mean for a conceptualization of L2 inner speech externalization? The implications are that, no matter how competent the FL or L2 learner may become, externalization in a language other than the LI will involve accessing a conceptual base with more or less pronounced links to both the LI and the FL/L2. At advanced stages of FL/L2 learning, inner speech externalization would thus entail, not so much translation (or word association) from LI to L2, but interactions between the LI and the L2 at the conceptual base (p. 83). This conceptualization is less radical than that held by Ushakova (1994), who does not allow for the building of direct associations between L2 words and the concepts they stand for and therefore does not conceive of L2 words as having any major impact at the conceptual level:66 The second language is incorporated into the classification system already available in the first language, relies on the previously developed semantic system, and actively deploys first language phonology.... To put it figuratively, second language is looking into the windows cut out by the first language, (p. 154)

Support for the notion that bilinguals or L2 learners at advanced levels of proficiency may develop common conceptual and semantic foundations for LI and L2 is provided by incoming evidence from neuroimaging research. As discussed in Chapter 3, it has been reported that bilinguals who learn their L2s in infancy develop common neural areas of semantic processing for the LI and L2 (Chee, Caplan, et al., 1999; Chee et al., 2003; Kim et al., 1997). Whereas Kim et al. (1997) found that, in contrast to early bilinguals, those who had learned their L2s in adulthood displayed different activation areas for LI and L2, other studies indicate that L2s may share 66 Ushakova's (1994) research involved the learning of decontextualized, isolated words in an artificial language. Because this was a new "language" for the participants, the study did not consider the effect of proficiency on word conceptualization. A critique of the methodology and the conclusions arrived at in the study is offered in Chapter 3.

172

CHAPTER 6

common brain areas with the LI regardless of onset age of acquisition (Chee, Tan, et al., 1999; Klein et al. 1995). Perani et al.'s (1998) research, in turn, proposes it is not so much age of acquisition but ultimate level of L2 proficiency that determines whether Lls and L2s will be assigned common or segregated brain areas. The possibility that at advanced stages of L2 learning learners may develop a solid conceptual foundation with links to L2 meanings and lexical or grammatical forms is suggested by Pavlenko (1999) in her thorough and insightful review of research on bilingual memory.67 Her position is that we need to account for language learning other than in tutored, institutionalized settings. Whereas in decontextualized classroom teaching the learning of new words in the L2 often leads to attaching word meanings to LI conceptual representations, in other learning environments an "astonishing variety and multiplicity of connections" arise when words are learned in context (p. 213). Research conducted among bilinguals learning an L2 in diverse contexts and varying in levels of proficiency, degrees of acculturation, language learning history, and language dominance (to name just a few of the factors that may affect the development and use of a bilingual conceptual store) suggests that concepts may interact with each other in multiple ways. Based on this research, Pavlenko proposes three main types of bilingual conceptual representation: coexistence of two conceptual domains (one for LI and one for L2), conceptual transfer from LI to L2, and conceptual change, in turn comprising one or more of the following processes: internalization of new concepts, conceptual shift from LI to L2, convergence into a unitary conceptual system, restructuring of concepts associated with the LI, and attrition, usually of concepts associated with LI. It is difficult to claim, in the face of such variety of bilingual conceptual scenarios, that the inner speech process that leads to the externalization of an L2 would only occur as a translation of one code into another at the post-programing stage or as the mere attachment of L2 labels to ready-made LI forms. A. A. Leontiev (1981) made room for the possibility of processing an L2 at deeper levels of semantic programing during speech production in his insistence that L2 teaching methods should "get rid" (p. 27) as soon as possible of the intermediate translation stage from LI to L2. If it is possible to do this-and Leontiev spells out some ways in which it can be achieved (to be discussed in Chapter 7)-then the influence of an L2 could be exerted much earlier (microgenetically) than at the post-programing stage. Vygotsky's belief that "while learning a foreign language, we use word meanings that are already well developed in the native language, and only translate them" (p. 159) would appear to be true at early stages of L2 learning or in certain instructional environments where language is learned through translation or in a decontextualized way and life experiences are not

67 Pavlenko (1999) stresses the need to differentiate between three levels of processing and representation for lexicalized concepts: (1) the lexical level, i.e. words forms; (2) the semantic level, i.e. information relating words to other words on the basis of meaning; and (3) the conceptual level, i.e. nonlinguistic information based on experience and knowledge of the world. In addition to lexicalized concepts, Pavlenko includes in her model grammaticized concepts, that is, grammatical concepts such as gender or aspect.

AN INTEGRATED VIEW OF L2 INNER SPEECH

173

culturally lived through the medium of the L2. It is hypothesized here that the extemalization of inner speech in L2 production in very advanced stages of L2 development, such as those resulting from bilingual/bicultural immersion, may involve drawing upon not just LI but L2 subjective word meanings (senses) at a condensed inner speech stage.68 This hypothesized L2 inner speech production model is displayed in Figure 6-1 and explained below. motive thought (based on non-linguistic conceptual component)

L2 inner speech

condensed inner speech (based on bi-lingual semantic component operating on LI and L2 word senses or subjective meanings with links to conceptual representations) application of L2 code synthesis of subjective and objective meanings, rudimentary syntax

L2 inner speech

transitional phase (may occur as expanded inner speech) lexicalization, grammatical elaboration, and phonological encoding in L2 L2 external speech (objective word meanings, lexical items, expanded syntax, and phonological form)

Figure 6-1. Schema of inner speech extemalization at advanced levels ofL2 development.

The extemalization of an L2, at advanced levels of L2 development, presupposes that learners will have developed a conceptual foundation based on representations linked to both the LI and the L2. Advanced L2 or FL learners or bilinguals, such as those described in Pavlenko and Lantolf (2000) or Larsen et al. (2002), will have at their disposal a rich dual-language69 conceptual store, a sort of "common underlying conceptual base" (Kecskes & Papp, 2000, p. xv) or "largely unified meaning system" (John-Steiner, 1985b, p. 366) comprised of both LI and L2 meanings, lexicalized as well as grammaticized, and built along the lines of conceptual bilingual development laid out by Pavlenko (1999): coexistence, transfer, and/or change. The process of extemalization of L2 speech starts, as all higher intellectual activity does, with a motive (see Figure 6-1). This motive will respond, as Luria (1981) explains, to a desire to 68 The "condensed inner speech" stage represented in Figure 6-1 would be roughly equivalent to the inner speech stages discussed in Chapter 2: Vygotsky's "condensed inner speech," Luria's "inner speech schema," Vocate's "inner speech coding" stage, and A. A. Leontiev's "inner programing" stage. 69 In the case of multilinguals, the conceptual store will be multilinguistic. The L2 label used in the model described here could stand for an L3, L4, and so on. However, the model may result too simplistic in these cases, as the transition from thought to word among multilinguals might be even more complex than among bilinguals.

174

CHAPTER 6

request something, communicate something, or formulate a thought (p. 147). Motive will set off thought, a level of activity that feeds on non-linguistic intellectual resources. Leontiev defines "authentic thought" as "based on conceptual components and thought-oriented auxiliary devices and tools, such as schemas, plans, visual images" (p. 103). Pavlenko (1999) also defines conceptual representations as "nonlinguistic multimodal information" but points out that, except for a few universally shared ones, they are "linguistically and sensorily acquired and are thus molded by a unique configuration of linguistic, cultural and sociohistoric factors at play at a particular time in a particular speech community" (p. 212). The process of turning thought into L2 words at advanced stages of L2 development will entail an inner speech phase mediated by linguistic resources in both the LI and the L2 (see Figure 6-1). Inner speech will be at the first stage highly condensed but rich in allusions to subjective word meanings, or word senses-to use the Vygotskyan term-and with links to LI, L2, unified, or bi-codal conceptual representations. Because of the presence of the two languages and the interaction between them as well as the shifting and unstable nature of the relationship between thought and word, this will be a highly dynamic phase, much more so perhaps than in a monolingual situation. In the transition from thought to utterance, condensed "bilingual" inner speech will progress to a more stable stage in which the L2 code will be applied. At this stage, there will be a synthesis of subjective and objective meanings, concretization through L2 word meanings, and rudimentary L2 syntax. For L2 inner speech to be externalized, it must undergo a transitional phase of lexicalization, grammatical elaboration, and phonological encoding resulting in the expanded syntax, word meanings, lexical items, and phonology of L2 external speech. Sometimes, the result of this phase may take place internally as expanded inner speech. For example, it may be necessary to engage in maximally unfolded inner speech in preparation for a lecture or in planning written texts (Vygotsky, 1986, pp. 88,243) and when talking to oneself (Sokolov, 1972, p. 121). The transition from thought to external speech, except in routinized, automatic speech activity, is not a smooth process. As Vygotsky (1986) remarked, it is characterized by "an infinite variety of movements to and fro" (p. 254). This should apply to LI as well as to L2 verbalization. In the latter case, when the aim is the production of a written text, the transition may even be more elaborate and agitated than when the aim is spontaneous oral speech. It might be fitting to close this section with a cautionary note about the model presented in Figure 6-1. Considering the immense variety of L2 cognitive and linguistic configurations that might affect inner speech processing and the vast terrain still to be covered by research in terms of elucidating the nature of bilingual and multilingual mental organization and functioning, it would be unwise to claim much defmitiveness or generalizability for such a model. Rather, such hypothesized schema should be taken as a first step toward specifying the stages that might be involved when the externalization of inner speech takes place in a language other than the LI. It is hoped that further research and theorizing will help revise such a model to reflect specific linguistic and contextual situations, such as the learning of conceptually

AN INTEGRATED VIEW OF L2 INNER SPEECH

175

similar or different languages, L1 - or L2-dominant settings, child versus adult learning, bilingualism and multilingualism, natural and classroom environments, and others. Overview ofL2 Inner Speech in the Externalization of Thought L2 inner speech has a crucial role in the microgenetic process of materialization and externalization of thought as L2 speech production. L2 learners may externalize their inner speech in two ways: as private speech or as social, communicative speech (in oral or written form). L2 inner speech may resurface as private speech during the performance of difficult cognitive or linguistic tasks. In this case, private speech will retain some of the structural and functional features of inner speech: It will be selfdirected, low in volume, elliptical, and sometimes unintelligible. The externalization of L2 inner speech as overt, communicative L2 production entails a much more complex transformation. There are two complementary hypotheses concerning the role of the L2 in the inner speech phase before externalization. One hypothesis postulates that, at early stages of L2 development and in decontextualized classroom settings, the L2 will be accessed only after thought has been semantically and syntactically configured by the LI. The inner speech process in this case will include a translation phase, from LI to L2, just before motor speech production is activated. The second hypothesis holds for cases of very advanced L2 development in which, as a result of internalization, conceptual representations of the L2 have been formed and have somehow affected the existing conceptual store of the LI. When this happens, inner speech will be influenced by the L2 to the extent that it has mingled with, restructured, or substituted the LI semantic foundation. Because these types of conceptual change do not usually occur all at once or across all cognitive domains in each individual, it should not be expected that the L2 will be equally implicated at the condensed semantic inner speech level in all communicative situations or at every stage of L2 development. FORM AND FUNCTIONS OF L2 INNER SPEECH The Structure ofL2 Inner Speech Verbal reports from L2 students indicate that L2 inner speech has a similar structure to that of LI inner speech. As reported in Guerrero (1994, 1999) and discussed in Chapter 5, inner speech in an L2 tends to be syntactically abbreviated, semantically condensed, and phonologically sonorous to the person experiencing it. Learners who have developed enough competence in the L2 do engage, however, in more elaborate syntactic and discourse structures, formulating sentences or constructing dialogues in their minds. The overall predominance of reduced inner speech, in the form of words or condensed semantic complexes, among L2 learners may be attributed to the fact that for certain cognitive operations inner speech functions more efficiently on the basis of short, abbreviated inner utterances. In the case of routinized or simple verbal tasks in

176

CHAPTER 6

the L2, for example, inner speech will have to be swift and elliptical so that the automatic activity mediated by inner speech may proceed smoothly and successfully. In certain circumstances, however, when rehearsing for public speaking or when engaged in contemplative self-talk, proficient learners may resort to more expanded, elaborate forms of L2 inner speech. The two forms of inner speech that Sokolov (1972) identified-abbreviated thinking in mere hints of words and expanded inner talking (p. 121)-may be present among proficient L2 learners. In fact, there may even be, as in LI inner speech, a constant "to and fro" between condensed and unfolded L2 inner speech, as Sokolov (p. 121) and Vygotsky (1986) suggested (p. 254). Studies of the private speech used by L2 learners while solving difficult cognitive or linguistic tasks have uncovered several linguistic features that account for the brevity of self-directed speech (Centeno-Cortes & Jimenez, 2004; DiCamilla & Anton, 2004; Roebuck, 1998). Some of these features are related to the fact that in selfdirected speech the speaker knows what he/she is thinking and does not need to verbalize all that is present in his/her consciousness. Speech-for-oneself has several mechanisms that allow known meaning to be omitted in an utterance. Some of these are, for example, the use of deictic elements such as this/that, here/there, and pronouns to signal known referents. Other features, such as unfinished utterances, semantically loaded key words, reduced morphology, and elliptical phonology, allow inner speech to proceed swiftly as the speaker shifts focus. Frawley (1997) has suggested that the crucial discourse feature of the "language for thought" is "focus" (p. 191), that is, the linguistic resource that allows the mind to keep both given (presupposed) and new information under the intentional scrutiny of the speaker as he/she speaks. Frawley's notion of focus is a revised proposal for what Vygotsky (1986) referred to as predicativeness, the tendency for inner speech to omit the psychological subject-that which is known-and to preserve the psychological predicate-that which is unknown-and for what Wertsch (1979b, 1985b) designated in terms of given and new. It is probable that many of the syntactic/semantic features identified for L2 private (audible) discourse may also be present in L2 inner speech. As regards the phonological nature of L2 inner speech, when learners say that they can hear the L2 in their minds may be indicative that sounds of the L2 have been internalized to some extent, in other words, that the social, vocal features of external L2 speech have been recreated on the internal plane. A crucial point in the genesis of internal phonological representations of the L2 is the concealed motor activity that takes place in the speech organs during listening or reading (Sokolov, 1972). The inward reproduction that occurs concurrently with L2 reading or listening contributes to the formation of acoustic representations and auditory images of the sounds of the L2. In turn, these contribute to the development of articulatory schemata to be put in motion in L2 speaking. Listening to LI speakers of the language, as well as engaging in the oral practice of L2 sounds, seems to be essential for the development of an "inner ear" as well as an "inner voice" for the L2. In listening or reading, and not just in speaking, the learners are reproducing the L2 in their own voice. Blonskii (cited in Sokolov, 1972) believed that the rudiments of inner speech emerge in children first from audible repetition and then from internal reproduction, echolalia, of the social

AN INTEGRATED VIEW OF L2 INNER SPEECH

177

speech heard around: "Listening to speech is not simply listening: To a certain extent we, as it were, talk together with the speaker" (p. 49). The Functions ofL2 Inner Speech L2 Inner Speech as a Tool for Thought: The Cognitive/Regulatory Macro Function People make plans, remember jokes, talk to God (some even claim God talks to them!), write messages, calculate, solve puzzles, make mental notes, dream, all through inner speech. It would seem that the functions enabled by inner speech are endless. However, if one had to identify the chief function of inner speech, it would be undoubtedly the thinking function or the cognitive/regulatory function. In inner speech thought and language come together, engendering verbal thought and enabling a higher plane of human activity-the plane of higher psychological processes. In monolinguals, verbal thought is thought mediated by the LI, but the mind of an L2 speaker, a bilingual or a multilingual person, is a mind constructed, or in the process of construction, on the basis of more than one language, the LI and one or more L2s. The question is then whether the vast range of higher mental functions-logical memory, mediated perception, voluntary attention, reasoning, remembering, planning, selfreflection, problem-solving-normally mediated by the LI can also be mediated by an L2. Though research in the area of inner speech functions from an L2 perspective is scant, the existing information suggests that the L2 can indeed serve the "macro" thinking function, operating alongside or instead of the LI. Only a few mental operations within the "higher thinking" category have been investigated from the L2 point of view. The function of problem solving was investigated by Cohen (1998) in his study (discussed in Chapter 3 of the language of choice (LI or L2) used by 5th and 6th graders in a Spanish immersion program to solve mathematical word problems. Only 3 of 15 participants reported total use of the L2 in solving the problems. The rest indicated use of the L2 to read the problem and translation or switching into the LI to understand instructions better or when the problem got difficult. A host of factors seemed to be at work in the students' choice of language of thought: general proficiency in the L2, knowledge of task-specific language, task difficulty, and academic achievement. Further evidence on the problemsolving function of L2 inner speech has been provided by Centeno-Cortes and Jimenez (2004) in their study of what they call "private verbal thinking," or externalized inner speech in the form of private speech that reflects reasoning during a problem-solving activity. Only the advanced L2 (college) learners in the study attempted to use the L2 for reasoning; the intermediate learners tended to use the LI. The researchers point out, however, that more than half of the L2 learners (N = 12) produced little or no private speech, a fact that suggests that some of them were able to reason out the problems through inner speech. A variety of "thinking" functions were tested in Cook's (1998) survey of the internal (as opposed to external) uses of the L2. Although Cook does not make a distinction between private and inner speech, it is clear that some of the surveyed functions could be conducted silently through inner speech. The functions included

178

CHAPTER 6

planning or self-organization (for example, making shopping lists); arithmetical mental tasks; remembering everyday information (such as phone numbers); unconscious uses (dreaming, talking to oneself );70 expressing emotional effects; non-communicative uses (such as talking to non-speakers-babies or animals); praying to oneself; and receptive uses (for instance, listening to the radio or reading the newspaper). The results, based on data from 59 proficient users of more than one language, showed that whereas both the LI and the L2 were used in general by all the participants in all the functions, there was a preference for the LI in arithmetical mental uses and praying and for the L2 in unconscious uses, memory tasks, and noncommunicative uses. Larsen et al. (2002) specifically pursued the role of L2 inner speech in remembering past events. Their study focused on the language used for autobiographical memories among a group of Polish immigrants in Denmark. When asked to recall a past experience associated with certain key words, the participants tended to recall in the LI those memories of events that had taken place before migration and in the L2 those that had occurred after migration. The participants also showed a tendency to use more L2 inner speech for other purposes, such as writing notes to themselves, thinking in words, expressing feelings, dreaming, and talking to themselves, if they had migrated to Denmark at a younger age and more LI inner speech if they had migrated at an older age According to the researchers, "older adults who immigrated to other countries provide a unique window onto the relation of memory and language, because these individuals have encoded the world of the homeland in one language and the 'new world' in another language" (p. 53). There is also evidence of use of L2 inner speech for introspection or selfdevelopment. In the interview data for Guerrero (1994), various students mentioned using inner speech in the L2 to clarify ideas, solve personal conflicts, have internal debates, and so on. One student reported she had found herself using the L2 to mentally explain to her boyfriend why she wanted to study biology and not something else. The process had helped her clarify her point of view. L2 inner speech was also used for self-diversion, that is, with the purpose of entertaining oneself or helping pass the time. One learner mentioned that she usually talked to herself in the L2 when waiting for someone or before falling asleep because it was "fun." In sum, it appears that it is possible for L2 learners to develop the ability to "think words" in the L2 for a variety of cognitive, regulatory functions. This use is contingent on a number of factors. Foremost among these is the extent to which the L2 has been internalized. As Cohen (1998) put it, "in order for inner speech to take place in an [L2], learners may need to attain a certain functional level with regard to vocabulary and structure" (p. 160). The key requirement in regards level of proficiency is the need for L2 learners to have developed a conceptual foundation based on lexicalized and grammaticized concepts (Pavlenko, 1999) in the L2 as well as in the LI. Only then can the crucial convergence of thought and the L2 take place at a condensed inner speech 70 It is not clear why talking to oneself has been categorized by Cook as an unconscious use of language though it is possible that he might be referring to the type of self-talk that occurs without much awareness or control by the speaker.

AN INTEGRATED VIEW OF L2 INNER SPEECH

179

level. Learners who are in the process of development of such a semantic foundation may find themselves using LI inner speech at the semantic level and translating into the L2 code during the externalization process (or vice versa, translating L2 affordances into the LI when listening or reading and conducting the "thinking" in the LI). Because L2 learners do not develop equal competence in the L2 across all cognitive domains, it is also possible that partial use of L2 inner speech may occur. In this case, L2 mediation of mental activity will only take place in those particular cognitive domains that have been encoded in the L2. The variety of scenarios where an L2 may serve as a mediating tool in the process of turning thought into words or vice versa is manifold. It is not only the level of proficiency or the linguistic configuration of cognitive domains that may impact use of L2 inner speech. As studies of competent L2 speakers have shown (Cook, 1998, John-Steiner, 1985b, Lantolf & Pavlenko, 2000; Larsen et al., 2002), the presence and interaction of multiple other factors-personal choice, contextual influence, level of difficulty of cognitive activity, and degree of acculturation-will determine whether the L2 will indeed function as a thinking tool in certain cognitive operations. The Rehearsal Macro Function ofL2 Inner Speech In addition to the cognitive/regulatory function, which on the basis of the existing evidence can be said to apply to L2 learners in certain cases, there is a very important function of inner speech that is typical-though not exclusive-of L2 learners, namely, the rehearsal function. Rehearsal, as has been seen in previous chapters, is a rather ubiquitous term in studies of memory, cognition, language learning strategies, and language play. In L2 learning, rehearsal is the covert practice of L2 forms or meaningful content in L2 form. The studies conducted by Guerrero (1994, 1999; described in detail in Chapter 5) point to rehearsal as an interim form of L2 inner speech, the overall purpose of which is learning or consolidating knowledge of the L2.71 Ultimately, rehearsal has a key role in the internalization of the L2 social speech and in the development of full blown L2 inner speech. The specific functions played by rehearsal-type L2 inner speech (mnemonic, instructional, evaluative, preparatory, dialogic, play, and affective) have been thoroughly discussed in Chapter 5; therefore, only a synthesis of the role played by these functions will be presented here. The first of these functions is the mnemonic one. As a mnemonic device, mental rehearsal facilitates retention of L2 words in short-term and long-term memory. Through rehearsal, learners can also retrieve or recall words they have learned. Rehearsal appears to be particularly important in the acquisition of new vocabulary and has been also speculated to have an impact on the acquisition of syntax (Baddeley et al., 1998; Ellis & Sinclair, 1996). Though mnemonic rehearsal can occur spontaneously, as for example when an L2 word or phrase suddenly appears and repeats itself in the mind, it can also be deliberately

71

As pointed out before, rehearsal-type inner speech does not represent fully developed L2 inner speech but rather a transitional phase in the interiorization of external L2 speech.

180

CHAPTER 6

deployed. The instructional function is the covert use of the L2 for self-instruction. It is trying to self-teach the L2 through deliberate imitation of L2 models, applying grammar rules, and attempting to formulate sentences in the L2. The self-instructional function of L2 inner speech appears to create an internal ZPD where learners can operate, on the basis of external models as the social "other," at levels beyond their current stage of L2 development. In its rehearsal function, L2 inner speech also plays an evaluative role. This is the use of the L2 to inwardly assess the extent of knowledge of the L2 (to see how much of the L2 one knows), to evaluate its use, in oneself and others, and to correct internally one's and others' production. Like the self-instructional function, classroom learners may have appropriated the evaluative function from teacher models and instructional contexts that place heavy emphasis on correction and evaluation. Rehearsal-type inner speech can also be preparatory, that is, focused on future or imagined communicative situations in the L2 for which the learner prepares. Through preparatory rehearsal, learners plan the content and form of what they are going to say or write; this includes planning exercise responses, oral reports, written text, conversations, and the like, as well as mentally answering questions posed to other students or the whole class. The preparatory function may serve internalizing purposes, that is, it may serve as a means to form enduring mental representations that can be put to use on a future occasion. The preparatory function may also be seen as revelatory mental output in that it may help learners notice gaps and holes in their own production, test hypotheses, and take a metalinguistic stance towards the L2. Preparatory rehearsal is not exclusive of L2 learners; LI speakers also prepare or plan mentally what they are going to say. The process of preparing for external speech in the L2, however, may involve a great deal more of inner rehearsal than for producing oral speech or written text in the LI. It is therefore a crucial function for L2 oral and written production. Another function of inner speech rehearsal is the dialogic one, that is, the use of L2 inner speech to imagine dialogues in the L2 with oneself or different versions of the self (intra-self function) or to mentally engage in conversations with other interlocutors (interpersonal function). Although the dialogic function of inner speech is extremely important and frequent in the LI as in any other language, it seems particularly useful for L2 learners as it allows, via the imagination, participation and engagement in L2 communicative encounters, thus providing additional practice in L2 conversational discourse. Because it can be intentionally activated and exploited (Vocate, 1994b), talking to oneself also appears as an ideal mechanism for L2 learners to engage in for purposes of self-reflection, self-monitoring, and self-development. As will be argued further down, the dialogic function may be essential in the creation of a new L2 identity. Inner speech rehearsal can sometimes perform ^play function, defined here as the playful or creative manipulation of and experimentation with the L2. Language learners play with the L2 when they make up rhymes, combine words in funny or original ways, invent words, or experiment with word order. This type of inner speech function requires rather high levels of competence in the L2. In order to engage in

AN INTEGRATED VIEW OF L2 INNER SPEECH

181

language play, learners must be aware, at least implicitly, of language rules and conventions in order to break them or use the language in uncommon ways. Through the play function, learners not only consolidate their knowledge of the L2 but also, as Belz (2002) suggests, assert their new identity as multicompetent language users. Because L2 rehearsal involves the cognitive as well as the emotional engagement of the individual, it also has an important affective function. In rehearsing through L2 inner speech, learners may shape or improve their self-image as L2 users; they may derive self-satisfaction and even fun in using the L2 internally; they may acquire selfconfidence and reduce communicative anxiety; they may also, unfortunately, notice shortcomings in their L2 use and find reasons to chastise or debase themselves.

INNER SPEECH AS MEDIATOR OF VERBAL TASKS IN THE L2 Reading, writing, listening, and speaking are, by definition, verbal tasks, that is cognitive tasks mediated by language in which the primary aim is the meaningful perception or production of oral or written speech. As mediator of verbal thought, inner speech activity is crucially implicated in verbal tasks, either in the processes of speech decoding and comprehension during reading or listening or in the processes of thought formulation, verbal encoding, and externalization in speaking or writing. Among monolinguals, the processes of speech perception and production are mediated by LI inner speech. Among L2 learners, bilinguals, or multilingual speakers, inner speech has more than one linguistic code to rely on. The specific role of the L2 in the processes of L2 speech production (speaking or writing) and speech perception (listening or reading) has been discussed in some detail above. To review, it has been argued that, depending on a variety of factors such as level of proficiency in the L2, type of instructional setting, teaching method, cognitive domain, context of the communicative situation, and others, the L2 may be implicated in two ways in inner speech activity related to verbal tasks. On one extreme, the L2 will be present as a code to translate from (in listening or reading) or to translate into (in speaking or writing) with the LI being the primary tool for assigning meaning to words or giving structure to meanings. On the other extreme, the L2 may be involved at deeper levels of semantic encoding (in speaking or writing) or decoding (in listening or reading), acting together with the LI as an alternative or additional tool for turning thoughts into language or vice versa. These two are diametrically opposed, hypothetical roles of the L2 as mediator of language tasks. Realistically speaking, however, the multiplicity of cultural, contextual, personal, linguistic, and psycholinguistic factors affecting communicative speech activity would render a not so neatly delimited but rather contingent use of the LI and L2 as tools for thought in speech perception and production, in the LI as well as in the L2. As Kecskes and Papp (2000) demonstrate, once a person develops competence in an L2/FL, verbal tasks in both the LI and the L2 may be affected by the presence of an L2/FL. The effects of multicompetence (Cook, 1992) on the inner speech related to speech production can be seen in a study conducted by Cohen and Olshtain (1993). The researchers interviewed 15 native and near-native speakers of Hebrew, all advanced

182

CHAPTER 6

EFL learners, to find out-among other things-the "language of thought" used in planning speech acts in the English (the FL). The researchers found three main patterns of Hebrew and FL use: (1) planning in the FL and responding in the FL, (2) planning in Hebrew, translating into the FL, and responding in the FL, and (3) planning in Hebrew and responding in the FL. Some of the non-native, but fluent, speakers of Hebrew also reported relying on their Lls during the planning stage, to the effect that in their cases there were three languages of thought used in FL production. The role of the LI in the inner speech processes related to L2 writing and reading has been emphasized in the empirical literature. The studies by Huh (2002) and Woodall (2002) (discussed in Chapters 3 and 4) confirm Cook's (1992) observation that "the L2 user does not effectively switch off the LI while processing the L2, but has it constantly available" (Cook, 1992, p. 571). In Huh's study, it was shown that the learners' main use of the LI in writing was at the deep semantic inner speech phase, after which the LI semanticized thoughts were rendered into L2 form. Woodall had similar findings with respect to the LI, additionally noticing that LI use in producing an L2 written text declined with proficiency but increased with task difficulty. Like Cook, Woodall contends that "in L2 writing, two languages can be at work at the same time" (p. 23). Similarly, Upton and Lee-Thompson (2001) found the LI to be a critical meaning-making tool in reading, not just used for translating L2 text into LI words but to wrestle with meaning, confirming comprehension, predicting text structure and content, and monitoring the reading task. In Upton and Lee-Thompson's study, use of the LI also declined with proficiency. These findings confirm Sokolov's (1972) early evidence that translation into the LI in reading is not necessary when FL texts are fairly simple or when the learner has good command of the FL. In those cases, immediate understanding can take place without translation (p. 84). The importance of inner speech in the processing of FL speech or text during listening or reading has been emphasized by Sokolov (1972). This was clearly demonstrated in Sokolov's (1972) experiments (discussed in detail in Chapters 2 and 3). Participants could comprehend and remember an FL text better when reading or listening was accompanied by reduced inner speech consisting of key words or semantic complexes: "The role of these inwardly reproduced words was enormous: by marking off the primary sense of the speech heard [or read], they became condensed expressions of large semantic groups" (p. 120). According to Sokolov, the ability to focus on a few semantically loaded words to get the general gist of what is heard or read is imperative for comprehension and retention in any language. When reading or listening difficult or highly abstract texts, however, individuals may need to resort to unfolded, even subvocally articulated, inner speech. Sokolov showed the importance of unfolded articulation through inner speech: The process of understanding an abstract text by repeating individual phrases and generalizing them to a single idea . . . required that words be articulated in an unfolded manner. And conversely, the simpler the text the more abbreviated would be its reproduction. Only a few generalizing words, a brief textual scheme, are fixed when a text is simple. This is not enough when a difficult text is to be understood; a more unfolded form of its reproduction is required here. (p. 115)

AN INTEGRATED VIEW OF L2 INNER SPEECH

183

L2 INNER SPEECH AND PROFICIENCY Proficiency has been consistently linked in the literature to aspects of L2 inner speech.72 In general, the tendency that has been shown is that the capacity to think words in an L2 increases with L2 proficiency. Strong evidence for this trend is provided by studies sampling various levels of L2 proficiency, from the lowest to the most advanced (Guerrero, 1994, 1999; Gutierrez, 2000), and including cases of very high attainment in the L2, such as those of immigrants who, in crossing cultural and linguistic borders, have developed a new self on the basis of the L2 (Larsen et al., 2002; Pavlenko & Lantolf, 2000). What does this correlation between L2 inner speech and proficiency mean? It means that with more of the L2 internalized as a result of exposure to L2 speech and interaction with other L2 speakers learners have a wider and stronger L2 foundation to draw upon in their inner speech activity. Low level learners may experience covert speech behavior in the L2, but this does not mean they are using L2 inner speech as a thinking tool. Learners who are in the process of developing L2 competence engage in a rudimentary form of L2 inner speech the main purpose of which appears to be rehearsing the language or thinking about the L2. The ability to think words in a language, however, is not developed uniformly by all speakers across all cognitive domains or communicative situations. There may thus be interim or even ultimate use of L2 inner speech for certain functions and not for others. It should also be noted that, whereas greater proficiency in a language progressively enables the utilization of that language as a tool for thought, the question of whether a person relies on the L2 or not for cognitive/regulatory purposes is largely a matter of personal choice. In terms of structure, L2 inner speech shows a tendency towards brevity and predicativeness across proficiency levels. With higher proficiency, L2 inner speech is likely to display much more complex levels of semantic elaboration and dialogicity as well as the capacity for more expanded syntactic forms during unfolded inner speech activity. In terms of functions, inner speech in the L2 varies greatly with proficiency. As noted, the use of L2 inner speech for cognitive/regulatory purposes increases with proficiency in the L2. As learners become more proficient in their L2, their ability to think through that language for a variety of cognitive and self-regulatory purposes (problem-solving, remembering, reasoning, introspection, planning) increases. Likewise, reliance on the L2 as a complementary thinking tool to the LI also increases with proficiency in the performance of L2 verbal tasks such as reading and writing. With greater proficiency, classroom learners come to use the L2 more and more for class-unrelated functions. The rehearsal function of inner speech, which is typical of L2 learners, however, 72 Proficiency is a problematic term when it comes to defining it. In this section, the term proficiency is understood as the dynamic, open-ended measure of competence and performance in any language that an individual demonstrates when participating in authentic, linguistically mediated activities, jointly with other individuals or independently (see Lantolf & Frawley, 1988, for a discussion of proficiency along these lines). Proficiency here is not seen as simply the result of the accumulation of knowledge of the L2 system but as an attribute that is developed on the basis of personal agency and commitment to learning the L2.

184

CHAPTER 6

decreases in several of its particular roles. The mnemonic, evaluative, and preparatory roles, all decline in frequency as proficiency increases. The instructional function, in its imitative and grammar practicing roles, also decreases. The dialogic use of L2 rehearsal, that is, the ability to engage in imaginary dialogues in the L2 with the self and others, increases, a phenomenon that is not surprising given the fact that talking to oneself is one of most common manifestations of inner speech in one's LI. The play function of inner speech, as defined here (the creative and playful manipulation of the L2), does not seem to be the most frequent of all rehearsal functions, even among the most advanced learners. But evidence of its use is found at all levels of L2 development (Belz, 2002; Broner & Tarone, 2001; Ohta, 2001; Saville-Troike, 1988). Finally, inner speech rehearsal in the L2 has an affective role, mostly beneficial, which seems to be important regardless of the learners' level of proficiency. A comparison of L2 inner speech manifestations among low proficiency L2 classroom learners-such as those in Guerrero (2004)-and very advanced L2 users-such as those in Guerrero (1999) or the skilled bilinguals in Pavlenko and Lantolf (2000)-suggests that L2 inner speech evolves as proficiency in the L2 develops. At the early stages of L2 development, the L2 inner speech of classroom learners appears to be largely echoic and imitative as they inwardly reproduce or recall the language they hear or read in their surrounding environments. Gradually, L2 learners attempt to engage in L2 inner speech in the form of rehearsal for a variety of purposes: preparation for future or imaginary L2 encounters, memorizing language for future use, testing how much of the L2 they know, and the like. These covert forms of L2 speech activity, which basically represent efforts to internalize the L2 through appropriation, are essential for the progressive development of a sound L2 foundation to mediate the articulation of thoughts in the L2. At early stages of L2 development, speech acts in the L2 will take place through mediation of LI inner speech at the conceptual level and translation into the L2 (in speaking or writing) or from the L2 (in listening or reading) at the coding stage (see A. A. Leontiev's schema of speech acts in Chapter 3). As L2 learners grow in proficiency and their conceptual foundation becomes reorganized and reshaped on the basis of an L2, learners should be able to conduct sustained inner speech activity in the L2 to mediate complex cognitive tasks. In most cases, with the exception of LI attrition or LI aphasia, the native language will continue to operate, to a greater or lesser extent, as a complementary instrument of verbal thought in relation to the L2. INNER SPEECH AND THE CREATION OF AN L2 IDENTITY In his review of Kozulin's 1986 edition of Vygotsky's Thought and Language, Frawley (1989) asks a series of questions on the repercussions of Vygotskyan ideas on the SLA field. Among these is the question "Is it possible to develop a person in a second language (L2)?" (p. 331). The question cuts right through a host of notions related to inner speech raised here and there in previous pages of this book, namely, notions about identity, self, self-awareness, self-reconstruction, agency, intentionality, and choice. Let us begin by defining "person" and "self." According to Harre (1987),

AN INTEGRATED VIEW OF L2 INNER SPEECH

18 5

"person" refers to "a human being as a social individual embodied and publicly identifiable." Self, on the other hand, is a discursively organized construct reflecting the "inner unity to which all personal experience belongs" (p. 42). In discussing the issue of L2 inner speech in relation to identity, it is the notion of self as a dynamic and ever-changing conglomerate of personal experiences that emerges as people talk about themselves that is going to be invoked. The concept of self is tightly related to the concept of personal identity. Having a personal identity is having the sense that one is not only an individual person, a unique identifiable human being, but also a certain "type" of person (Harre & Gillett, 1994, p. 102). It is largely through inner speech, that is, through the internalized activity of deploying language for thinking and intrapersonal communication, that a sense of self and a certain personal identity are construed. Inner speech-by means of the language mechanism of indexing first person through pronouns-makes possible self-reflection. In inner speech, the I is scrutinized or monitored by Me or Myself. Through inner speech, the self analyzes itself. For Vygotsky (1986), it is the inner word that makes possible verbalized introspection (p. 171) and leads to the development of selfconsciousness—the "consciousness of being conscious" (p. 170). Similarly, for Bakhtin "self-observation (introspection) is the understanding of one's own inner sign" (Volosinov, 1973, p. 36). Through inner speech, too, humans imagine "possible selves" for themselves, that is "ideas of what they might become, what they would like to become, and what they are afraid of becoming" (Markus & Nurius, 1987, p. 157). According to Vocate (1994), personal growth or change can only occur through self-talk, a dialogic modality of inner speech characterized by interaction between the I-the personal self-and the Me-the internalized social self-out of which selfconsciousness emerges. Interactions between the I and the Me are interpretive, intentional processes through which a self is constructed and reconstructed. The new self emerging from these interactions is a "synthesis" of the individual and society, of personal senses and societal meanings. When self-talk takes place through inner speech in the LI, the self that emerges is structured along the lines of the culturally determined semantic foundation of the LI. The question is what type of self results when self-talk draws upon the resources of an L2. Or, rather, can a new L2 self emerge from the selftalk process when inner speech is mediated by the conceptual and culturally specific basis of an L2? Because concepts are so tightly related to one's way of perceiving reality and one's own inner selves, changes in conceptual thinking as a result of the appropriation of another language and culture may lead to profound changes in self-concept and personal identity. Data on the cognitive and conceptual changes undergone by proficient learners of an L2, such as the late bilinguals in Pavlenko and Lantolf (2000), the sophisticated adult learners in John-Steiner (1985b), or the dual-language immigrators in Larsen et al. (2002), suggest that changes in the basic LI semantic foundation supporting inner speech are possible with the acquisition and use of an L2. Larsen et al. (2002) sum up the type of changes that may occur: As conceptual representations formed in one cultural context (the homeland) are challenged by second-language acquisition and acculturation in another country (the "new"

186

CHAPTER 6

country), they may interact in a number of ways. Older conceptions may co-exist with new conceptual representations, they may shift in scope and amplitude in the direction of new cultural dimensions, or they may converge towards wholly new representations different from both the first and second language/cultures, (p. 52)

The inner speech resulting from such interactions between the LI and L2 is aptly described by the Polish-born writer Eva Hoffman (1989): When I talk to myself now, I talk in English. English is the language in which I've become an adult, in which I've seen my favorite movies and read my favorite novels, and sung along with Janis Joplin records. In Polish, whole provinces of adult experience are missing. I don't know Polish words for 'microchips,' or 'pathetic fallacy,' or The Importance of Being Earnest. If I tried talking to myself in my native tongue, it would be a stumbling conversation indeed, interlaced with English expressions. So at those moments when I am alone, walking, or letting my thoughts meander before falling asleep, the internal dialogue proceeds in English. I no longer triangulate to Polish as to an authentic criterion, no longer refer back to it as to a point of origin. Still, underneath the relatively distinct monologue, there's an even more interior buzz, as of countless words compressed into an electric blur moving along a telephone wire. Occasionally, Polish words emerge unbidden from the buzz. They are usually words from the primary pallette of feeling: 'I'm so happy,' a voice says with bell-like clarity, or 'Why does he want to harm her?' The Polish phrases have roundness and a surprising certainty, as if they were announcing the simple truth, (p. 272)

In Hoffman's case, although English has become the basis on which her new conceptual framework has been built and the medium through which she talks to herself, Polish (her LI) is still represented in memory as the language in which primeval "truths" are engraved. Conceptual/semantic changes occasioned by the appropriation of a new language/culture are often accompanied by great disturbances in self-concept and personal identity. As discussed in Chapter 3, these changes may cause loss of one's identity as a particular LI speaker and painful reconstruction of a new L2 speaker self. Not all L2 learners develop an L2 persona that takes over and supplants the old L1 self. Some proficient L2 learners are able to shift between different linguistic identities depending on a variety of factors: cultural contexts in which they move, language in which the memories have been encoded, job-related demands, and the like. Writer Alice Kaplan (1993), for example, reports on the process of shifting to her L2 (French) persona on her first day of class as a professor of the L2: I haven't been in France for six months and I'm rusty, I'm out of shape for French. I'm in the middle of a sentence and I wonder how I'm going to get to the end, my intonation is off a little and I hear some air in my "p"s. I start correcting myself, I'm feeling double. I say the ritual words, a welcome to the course, and I settle down. I'm in my French persona, (p. 207)

Kaplan's craving for French causes her some anguished feelings, which she tries to appease by deliberately resorting to inner speech in French: Sometimes I don't want to need French so much. I want to be free of it. No more secret language, no more veering off, no more wanting in and never quite getting there. Because I can't imagine not having French. I think I would starve without it.

AN INTEGRATED VIEW OF L2 INNER SPEECH

187

I can't stand not to be in France in June . . . . So every year around the same time, I start speaking to myself in French and dreaming in French, and swearing in French when I'm driving my car. (pp. 207-208)

Inner speech seems to have an important role in the construction of a new identity or self-concept on the basis of an L2. Two ways in which inner speech impacts selfreconstruction are (1) through changes in the conceptual structures on the basis of which inner speech operates and (2) through self-talk, as the dialectic, intentional, and controllable mechanism that makes possible shifting between languages and adopting different linguistic personae in the mind. Reaching the high levels of appropriation of a new L2 language and culture that are required for self-reconstruction may be very difficult to achieve, one reason being that changes in conceptual structures require very high levels of attainment in the L2, another being the crucial matter of personal choice. Despite high levels of mastery over an L2, some learners will resist adopting an L2 speaker identity. As Wertsch (1998) has noted, appropriation of cultural mediational means will sometimes be accompanied by resistance. Resistance may very well lead learners to view the L2 as an alternative code that might be useful to know but not necessary for filtering and reconstituting their world view and personal life experiences. As Pavlenko (1998) points out, in these cases, "bilingualism will be mainly reduced to bi- or multicodalism (i.e., the speakers may speak more than one language, but the meanings will all be supplied by their native one)" (p. 18). Complex ideological and affective factors may drive a person to choose becoming-or not-a certain L2 speaker persona. Kaplan (1993) wonders why people want to adopt an L2 culture and language. Her very personal answer: "Because there's something in their own [language] they don't like, that doesn't name them" (p. 209). The L2 may help express some feelings and ideas more easily than the L1. In Kaplan's case, speaking a foreign language is "a chance for growth, for freedom, a liberation from the ugliness of our received ideas and mentalities" (p. 211). An ESL learner planning to become an ESL teacher views dominating the L2 as essential for preserving her self-esteem. She achieves this by relentlessly evaluating and correcting herself and others through inner speech: "I constantly evaluate the use of the language in other people and in myself. I try to self-correct because there is nothing so horrible as a future teacher who does not master her field and mine is English." An ESL learner in Puerto Rico (where learning English is sometimes viewed as a threat to having a Puerto Rican identity, see Schweers & Hudders, 1993) feels his "ego" is involved in learning the L2.73 In this case, it is not only a question of self-esteem but also of a sense of belonging to an L2 speaking group. The learner says he likes to engage in L2 inner speech because "every person has an ego. My ego tells me that I want to learn English no matter what. I would love to learn English. . . . I have relatives who speak English. If they can, why can't I?" The learner in this case uses inner speech to construct a possible self, the self he would like to be in the L2. 73 As Harre and Gillett (1994) explain, "it is very difficult to express the English idea of the self in Spanish" (p. 101). The Spanish-speaking subject who is quoted in this example is probably using the word "ego" in Spanish to refer to the idea of "self in English.

188

CHAPTER 6

Positive or negative attitudes toward the L2 and the LI may begin to be shaped early in life with strong effects on developing identities. When the dominant culture is that of the L2, such as in the case of immigrants in the United States, it may be very difficult for children acculturated in such environments to preserve a strong, favorable attitude toward their LI s. This situation is forcefully illustrated in a study by Orellana (1994). In the study, it is shown that one of the uses of English (the LI) that 3 Spanish-speaking preschool children engaged in was that of giving voice to the superheroes of the dominant culture: Superman, Supergirl, Spiderman, Peter Pan, Mickey Mouse, and the like. Often this employment of a dramatic voice took place as private speech during the children's solitary play, and it was always in English. At their young age, the children were appropriating the language of the power culture through the voice of the superheroes. The children's internalization of the L2 and the association they were making between English and "being powerful" had started molding unfavorable attitudes toward the LI and a rejection of a bilingualism that entailed being "Spanish-speaking." Carlos, one of the children, is reported announcing that when he grew up he would "only speak one language" and, predictably, that that language would be English (p. 185). The following conversation, originally held in Spanish, is highly revelatory: Carlos: "Mami, I'm not going to speak Spanish anymore." Carlos's mother: "Why not?!" Carlos: "Because I'm getting older. I'm growing up." Carlos's mother: "But lots of grownups speak Spanish!" Carlos: "But Batman doesn't speak Spanish. Superman doesn't speak Spanish. Peter Pan doesn't speak Spanish, either." (p. 185)

Even a child as young as Carlos was already making a conscious choice about the type of linguistic person he wanted to be. No doubt, powerful societal influences had contributed to shape Carlos's decision to opt out of his LI, yet, as much as that might be true, it is no less a fact that whatever type of linguistic self one is in private remains a matter of personal choice. As a student, cited in Cohen (1998), vividly expressed: You usually think that in an immersion school, it's totally in Spanish. If you get caught speaking English, you'll like be in trouble, but that's not really what it is. I mean, you are always thinking in English. I mean they can't really stop you from thinking in English. You can think Spanish, you can act Spanish, you can do everything in Spanish, but you're really not a Mexican, (p. 208)

CONCLUSION In this chapter, L2 inner speech has been viewed as the result of the internalization of social speech and discursive practices in the L2. It has been seen that children and adults may develop inner speech in the L2 and that in both cases there may be a transitional phase of audible private speech during which manipulation of the L2 will serve mainly an internalizing function. It has been argued, however, that the internalization of an L2 and the development of L2 inner speech occur largely, and particularly in the case of adult learners, as an internal process characterized by the

AN INTEGRATED VIEW OF L2 INNER SPEECH

189

covert reproduction of L2 speech, inward practice of the L2, and concealed verbalizations in the L2. Two different processes have been elucidated in the externalization of L2 inner speech. One process concerns the externalization of inner speech as private, regulatory, speech in the course of demanding intellectual tasks. Another process is related to the externalization of inner speech as social, communicative L2 speech. Externalization in this case involves the materialization of thought through semantic, syntactic, and phonological encoding. In this process, convergence of thought and speech will occur through LI mediation first and then translation into the L2 at early stages of L2 development or in decontextualized learning situations. Thought mediation through an L2 will occur at advanced stages of L2 development and when the individual's conceptual foundation has been reshaped and reorganized by the L2. In this case, the L2 will operate as an alternative or complementary mediator of verbal thought in relation to the LI. Inner speech in the L2 has been found to share similar structural properties with inner speech in the LI, that is, a tendency toward syntactic abbreviation, semantic condensation, and sonority in the mind. L2 inner speech may be structurally unfolded at times, for example, when reading or listening to difficult L2 material or when rehearsing for public speaking or written production. Two "macro" functions of L2 inner speech have been identified: the cognitive/regulatory one and the rehearsal one. In its cognitive/regulatory function, L2 inner speech functions, like LI inner speech, as a tool for thought in a variety of mental operations, such as remembering, planning, reasoning, and introspection. The rehearsal function, which is characteristic of a developmentally early stage of L2 inner speech, allows for the creation of a temporary internal plane of psychological activity in which the L2 is put to use on a sort of "trial" basis and its effectiveness as a tool for thinking is tested. Covert rehearsal of the L2 has been found to play several specific roles: mnemonic, instructional, evaluative, preparatory, dialogic, play, and affective. Whereas the importance of some of these rehearsal roles decreases with proficiency, L2 inner speech in general increases with proficiency. Because of its strong, genetic links to sociocultural activity, it has been argued that L2 inner speech is crucially implicated in the construction and development of a linguistic identity. A new L2 self may emerge as a result of the internalization of culturally shaped concepts associated with the L2 and the intentional activity of reshaping identity through internal self-talk. The internalization of a new tool of semiotic mediation, an L2, cannot be an inconsequential phenomenon for anyone's mind. It is a transformative experience that affects the way in which an individual discursively establishes an intimate relationship with the self and the external world. This chapter has attempted to demonstrate how an L2 can so impact the human mind in the fluid, dynamic, and on-going activity of developing and utilizing L2 inner speech.

CHAPTER 7 DEVELOPING L2 INNER SPEECH: A PEDAGOGICAL PERSPECTIVE

This chapter is aimed at providing professionals in the L2 teaching field with an understanding of the critical role teachers, learners, and instructional environments play in the development of an intramental sphere of activity mediated by the L2. Implications are drawn from previous chapters and relevant literature on how instruction may contribute to the enactment of favorable conditions for the internalization of the L2 and its employment as a tool for thought. The crucial role of the LI in the thought processes of L2 learners and in the experience of acquiring literacy in the L2 is examined. Various strategies naturally deployed by students in their efforts to internalize the language, such as silent repetition and eavesdropping, are discussed. Several aspects of teaching deemed useful or essential in the development of L2 inner speech are presented: (a) providing learners with opportunities for social engagement with the L2; (b) fostering internalization and externalization of the L2 through repetition, summarizing, planning, taking notes, paraphrasing, and problem solution; (c) helping students develop a conceptual base with links to the L2; and (d) raising awareness about the role, benefits, and uses of L2 inner speech. This chapter does not provide easy formulas on how to help learners develop inner speech in the L2. Acquiring the capacity to think words through another language is a complex and formidable achievement that cannot be interpreted as the simple outcome of "pedagogical" intervention. Learners' motives, goals, attitudes, and behaviors; social and environmental conditions; teaching methods; and teachers alike, all may facilitate or hamper, to a greater or lesser extent, the attainment of L2 inner speech. THE RELATIONSHIP BETWEEN TEACHING AND L2 INNER SPEECH DEVELOPMENT As a secondary phenomenon, inner speech is the consequence of the internalization of social speech, LI or L2. Inner speech is derived from external speech and cannot happen in the absence of exposure to and participation in the particular discursive practices of a social group. Many individuals-migrants, long-term residents in L2 communities, members of bilingual or multilingual groups-develop inner speech in an L2 naturally (without the intervention of deliberate instruction) as a result of

192

CHAPTER 7

immersion in day-to-day activities in which the L2 is the chief or an alternate means of mediation. For these individuals, the L2 becomes not only a much needed tool for maintaining interpersonal relationships but also for conducting intrapersonal intellectual tasks. Classroom L2 or FL learners, on the other hand, depend to a large extent on teachers and methods as well as on their own volitions for the recreation of external conditions of L2 use and participation and for the eventual development of the L2 as a tool for thought. How can teaching, then, affect the learner's process of internalization of an L2 and have an impact on the formation of a learner's internal plane mediated by the L2? This section deals with such an issue: the role of teaching in the development and use of L2 inner speech. To understand the role of teachers in the development of inner speech among L2 or FL learners within the context of purposive activity, one needs to be reminded of the proposed structure of speech acts in the FL74 as envisioned by A. A. Leontiev and discussed in Chapters 3 and 6. In this model (see p. 61 in this book), the transition from motive to utterance in the planning of speech acts in the FL is not direct. It involves, first, mediation of the native language in the programming of the utterance and, then, translation into the FL at the post-programming stage (A. A. Leontiev, 1981, p. 26). Whereas this model applies to beginning learners, more advanced learners should be able to reach a level of FL development where translation from the native language to the FL is no longer needed, at least as it pertains to some cognitive domains. The teacher's task, Leontiev believes, "is to 'get rid' of the intermediate stage as quickly as possible and to bring the psychological structure of the utterance in the foreign tongue as close as possible to that which operates in the mother tongue" (p. 27). To do this, Leontiev suggests providing learners with the rules governing the transition from speech operations in the LI to those of the FL. These rules are specific to the languages involved, but in general they are of three kinds: (a) simple transference in those features where the two languages are identical, (b) transference with modification and clarification, and (c) formation of new operations where the FL is completely different from the LI (p. 28). In Leontiev's view, mastering these rules of transition from one language to another should lead to the automatization of speech operations in the new language and make translation from LI to FL unnecessary. Leontiev offers no proof of the efficacy of the above procedures, but what is interesting about his suggestion is that eliminating translation in producing speech acts in an FL is presented as a real possibility. In discussing the notion of "thinking in a foreign language," Leontiev further elaborates on what he believes are effective ways of engaging the learner in creative intellectual activity in the FL. The idea (cited as Belyayev's) is to relate the target language to the learner's activities "and construct teaching in such a way that language is not used only for communication purposes .. . but as a medium which will engage the thought, perception, and imagination of the 74 In his book, A. A. Leontiev (1981) always uses the term "foreign language" (FL) to refer to the target language. This term will be preserved in discussing Leontiev's ideas; however, it is clear that the status granted by Leontiev to the FL can be extended to an L2, that is, a language learned in environments where it is the predominant or an alternative language of the community.

A PEDAGOGICAL PERSPECTIVE

193

learner" (p. 65). Some ways to achieve this are by teaching the target language through what is known in Western circles as content-based instruction or teaching language for specific purposes, for example, by having learners solve mathematical problems in the target language or by making the content of the speech activity relevant to the professional concerns of the learners (p. 65). Only by introducing a problematic situation of interest to the learner, will his/her thought become engaged in producing the target language. "To think in a foreign language" is thus possible for Leontiev but only when it is understood as the concretization of the cognitive act and not just as the mere attachment of language labels to ready-made thoughts (p. 109). Thus, the often heard precept that learners should think in the L2 or FL needs to be reconceptualized. First, it must be freed from the simplistic idea that thoughts can be neatly wrapped in words as if they were mere contents to be packaged. Second, it must be understood as something more than just switching or translating from the LI to the L2. Thinking in a language should be more properly conceived as using the L2 creatively and dynamically in the act of reasoning from the very first moment vague, wordless thought is invoked. Encouraging students to "think in the L2" may have no real impact on thought other than asking students to translate into the L2 ideas that have been prepackaged in the LI. Teachers should be aware, therefore, that without internalization of the L2 and the reconstruction of the conceptual foundation to accommodate for subjective and objective meanings in the L2 as well as in the LI, properly "thinking in the L2" will not occur. Such a level of L2 intellectual activity usually comes about with great proficiency in the language and in the presence of certain propitious conditions for the building of new semantic linkages to the L2, such as language immersion, contextualized teaching, learner interest, and so forth. Teachers should also be aware of the fact that, in most cases, barring LI aphasia or LI attrition, the LI continues to operate and exert its influence on the intellectual processes of the L2 user. UNDERSTANDING THE ROLE OF THE LI IN COVERT L2 PROCESSES Using the LI to mediate comprehension and production in the L2 seems to be a natural (that is, not explicitly encouraged by teachers) strategy among beginning L2 or FL learners. Learners at early stages of acquisition resort to the LI to scrutinize, reflect on, manipulate, and gain control over the L2, In the diary study conducted by Guerrero (2004), beginning ESL learners reported the intramental use of the LI in several ways: comparing features of the L2 and the LI, pairing up words (finding equivalent words) in LI and L2, reading or listening to the L2 but thinking in the LI, thinking in the LI and then translating into the L2 before speech production, listening to questions in the L2 and answering them in the LI, and the like. Use of the LI does not seem to be restricted to beginning L2 learners. Intermediate, high, and advanced proficiency level students also reported sometimes experimenting L2 inner speech mixed with the LI (Guerrero, 1994, 1999; see also Chapter 5). Evidence that L2 learners, including the most advanced, resort to the L1 in problem-solving tasks was found by Centeno-Cortes and Jimenez (2004). In their study, advanced L2 learners made more use of L2 private

194

CHAPTER 7

speech than intermediate learners but still resorted to the LI when the task became too difficult. There is also evidence that individuals who have attained high levels of bilingual or multilingual development shift between the LI and the L2(s) in their inner speech depending on the type of intellectual task being conducted and the environmental context in which they find themselves immersed (Cook, 1998; Larsen et al, 2002). It may be a feature of growing bilingualism, and particularly of settings where both the LI and the L2 are means of communication or at least easily accessible, that learners will find themselves resorting to both the LI and the L2 in their inner speech. The common teacher advice that students think in the L2 on the presumption that such activity will enhance the learners' fluency or accuracy in the L2 may therefore be, as Cohen (1998, p. 170) expresses, an unrealistic and perhaps detrimental proposition. For mature students, the LI is an already internalized and very effective cognitive resource. The LI is the most readily available tool for making meaning of reality and life experiences, and this includes figuring out a new linguistic system-the L2. Lantolf (2000b) expresses this idea very clearly: If language is a symbolic tool that mediates our thinking, and we develop control over this ability early on in our lives,... then perhaps it is difficult to move away from this use of our native language even when we have the requisite proficiency to do so. Our language is then strongly implicated in our identity as thinking beings and therefore proficiency alone does not determine use of our native language to mediate ourselves, others, and the interrelationship between the two." (p. 87)

The teachers' view of LI use in the mind as going against the learners' learning of the L2 should be attuned to the realization that, as Cook (1992) said, "the LI is present in the L2 learners' mind, whether the teacher wants it to be there or not. The L2 knowledge that is being created in them is connected in all sorts of ways with their LI knowledge" (p. 585). For L2 teachers, it is important to acknowledge not only the presence of the LI in the cognitive activity of the learner when solving L2 tasks but also the critical role it may serve as a tool for thought. As Anton and DiCamilla (1998) observed and others concurred (Guerrero & Villamil, 2000) in relation to the use of LI in collaborative interactions among L2 peers, it may not be wise pedagogically to ban or discourage the employment of the LI in certain classroom tasks. LI emerged as private speech during collaborative interaction in L2 classrooms (Brooks et al., 1997; Di Camilla & Anton, 2004; Guerrero & Villamil, 2000; Ohta, 2001; Villamil & Guerrero, 1996) and in independent problem solving (Centeno-Cortes & Jimenez, 2004). LI in private speech is used for a variety of self- and task-regulatory purposes: self-encouragement, release of emotions, self-evaluation, searching for words, translating, self-directing questions, and reasoning. A pedagogical recommendation shared by researchers is that the LI should not be banned from intra-or interpersonal use in L2 classrooms because it deprives learners of a critical mediating tool. This recommendation is qualified, however. It is not indiscriminate use of the LI in the classroom nor a return to translation methods that is advocated. On the contrary, the suggestion is to acknowledge the presence and instrumentality of the LI as a cognitive resource but to

A PEDAGOGICAL PERSPECTIVE

195

strive for the learners' progressive weaning from sole dependence on the LI as mediator of verbal thought. THE ROLE OF INNER SPEECH IN BECOMING LITERATE IN THE L2 The introduction of children to literacy seems to correlate with their capacity to use inner speech. Vygotsky (1986) believed that it was the introduction to abstract thinking and scientific concepts in elementary school that consolidated the child's growing differentiation of social speech from private and inner speech. Researchers agree that the emphasis placed by school on silent cognitive tasks mediated by language, such as reading, writing, and mathematical problem solving, helps children develop, strengthen, and become aware of inner speech (Kronk, 1994; Liva et al. 1994; Flavell et al, 1997). Success in school, to a great extent, hinges on the child's capacity to "think" words, that is, on the capacity to detach words from sounds and the objects they represent and assign meaning to abstract words in the mind (Gallimore & Tharp, 1990, p. 193). For Vygotsky (1986), the development of inner speech follows that of oral speech and is, in turn, succeeded by written speech, a form of speech activity that is even more abstract, conscious, and deliberate than inner speech (pp. 182-184). Vygotsky, however, believed that instruction always precedes development and that it is only when this happens that real cognitive growth takes place. Thus, the introduction of reading and writing to a school child enhances the development of inner speech, and hence of higher mental functions, by bringing awareness to speech and requiring from the child higher levels of abstraction and conceptual synthesis than ever before (pp. 182-184). Inner speech has been recognized as essential to the processes of writing and reading. In writing, as Vygotsky (1986) pointed out, the child needs to learn to perform the transition from "maximally compact inner speech to maximally detailed written speech" (p. 182). The covert enunciation that occurs before text is written down constitutes, according to Sokolov(1972), "an important factor in determining the syntactic structure and the entire style of the text" (p. 68). In particular, the internal intonation of inner speech contributes to segregate words into meaningful groups that become the basis for the syntactic arrangement of words in written discourse. Similarly, in reading, it is believed that inner speech functions as the prosodic reconstruction of written text, out of which meaning hypotheses are formed (Beggs & Howarth, 1985; Sokolov, 1972). Campbell (1992), who has studied inner speech among the deaf, suggests that teaching deaf children the sound-letter relationships that occur in alphabetic languages fosters the development of phonological (not acoustic) inner speech. According to Campbell, if one's "goal is to enable the deaf person to enter fully into a literate culture" introducing reading and writing to the child "may enhance the deaf child's skill in using inner speech" (p. 86). If developing inner speech is crucial for children to engage in the types of higher mental functions that are characteristic of school, would it be possible to train children to use inner speech? There has been some research on this. Liva et al. (1994) conducted two studies in which they showed that children who are poor readers and

196

CHAPTER 7

writers can be taught to use inner speech and that their reading and writing skills improve with deliberate deployment of inner speech. Overtly verbalizing strategies to oneself, a form of self-directed speech, was also found beneficial for 2nd to 4th grade children with language deficiencies (Schunk & Rice, 1984). Diaz, Neal, and AmayaWilliams (1990), however, reported little or no beneficial effect of training children to regulate themselves through self-verbalization on their spontaneous use of private speech. The authors suggest that further research should be conducted to identify the types of teaching strategies that might promote children's greater self-regulation through private speech when performing cognitive tasks. What transpires from the research on reading and writing, is that the relationship between inner speech and literacy is a two-way alley: Inner speech contributes to better reading and writing skills and is in turn stimulated by the introduction of reading and writing. In L2 reading and writing, inner speech is just as important as it is in LI reading and writing. Two major differences, however, need to be taken into account. One is that, in most cases, learners who are past childhood age confront the experience of literacy in the L2 having already developed their capacity to think abstractly and control their thought processes through inner speech. They have already awakened to the crucial realization that social speech is for others and inner speech is for oneself. This important feature of L2 learning accounts for learners trying to construct meaning of or derive meaning from verbal material in the L2 through covert, rather than overt, verbalization. L2 learners resort to low volume subvocalization or silent inner speech when reading or writing because they know that loud reading and writing are not only cumbersome and quite ineffective for rapid processing but also socially inappropriate and frowned upon. The other major difference between L2 and LI reading and writing is that, whereas monolinguals have only the L1, L2 learners or users have two semiotic systems to draw upon when using inner speech for reading and writing. While for many years knowledge of the LI was viewed as negatively interfering in the processes of L2 reading and writing, there is now a growing acknowledgment among teachers and researchers that having more than one language at one's disposal is beneficial, rather than detrimental, for L2 production and comprehension. The studies of Woodall (2002) and Huh (2002) on L2 writing as well as that of Upton and Lee-Thompson (2001) on L2 reading-discussed in Chapters 3, 4, and 6-suggest using the LI has an important facilitative effect on L2 understanding and composing. These studies have shown that L2 readers naturally resort to their LI when reading or writing and that, although reliance on the LI diminishes with proficiency, learners at all levels continue to profit from the wider and richer semantic base built along two or more internalized languages. For teachers, thus it is important to recognize that becoming literate in the L2 means developing the ability to handle not one but two languages efficiently in the mind and that for L2 readers and writers the LI is an important cognitive resource that can enhance rather than detract from L2 reading and writing.

A PEDAGOGICAL PERSPECTIVE

197

LEARNERS' STRATEGIES FOR THE DEVELOPMENT OF L2 INNER SPEECH As shown in sociocultural theory studies of private and inner speech as internalization of the L2 (Centeno-Cortes, 2003; Guerrero, 1994,1999,2004; Lantolf, 1997; Lantolf, 2003; Lantolf & Yanez, 2003; Ohta, 2001; Saville-Troike, 1988), learners seem to develop naturally strategies that contribute to the development of inner speech in the L2. Such is the case of inward or sub vocal repetition and imitation, eavesdropping and reflecting on others' L2 speech, spontaneously and deliberately manipulating the language, imagining conversations in the L2, and the like. Learners, for example, engage in varied forms of covert language repetition and imitation. In all cases, there is some form of reproduction-verbatim, selective, or innovative (see Lantolf & Yanez, 2003; Lantolf, 2003 )-of L2 affordances in the environment, including teacher models but certainly not restricted to them. There is empirical support in the literature for repetition as a language learning strategy. Concealed verbalization, or the subvocal reproduction of written texts, for example, has been observed to be crucially necessary for reading and understanding FL texts at the initial stages of FL learning (Sokolov, 1972). Subvocal and explicit unfolded articulation has also been seen as necessary for the internalization of phonological features (McKay, 1992), and silent rehearsal has been linked to long-term retention of new vocabulary and formulaic utterances, learning and consolidation of language forms, phonological memory and accuracy, grammatical fluency and accuracy, comprehension, and metalinguistic knowledge (see, for example, Baddeley et al., 1998; Hu, 2003; Ellis & Sinclair, 1996). Methodologically, however, repetition has been in disrepute since the fall of the audiolingual method. Once the most favored classroom techniques, pattern practice and drilling, both grounded on repetition and imitation, were found to be based on what was thought to be the erroneous theory that language learning is a question of habit formation. Cognitive approaches to language learning, fueled by Chomskyan nativist notions and Krashen's belief in the supremacy of comprehensible input, banished any form of repetition from the language classroom. Communicative approaches to language learning equally neglected repetition or imitation on the basis that what counted was negotiation of meaning. To this day, repetition in the language classroom is almost a forbidden practice. As mentioned above, however, research within sociocultural theory has begun to reconsider the role of repetition, imitation, and rehearsal in language learning.75 Sociocultural theory has emphasized, in particular, the importance of imitation in the transformative and agentive reproduction of social speech and therefore in internalization (Lantolf & Yanez, 2003; Lantolf, 2003; Newman & Holzman, 1993). A view of repetition as vital to language learning is also consonant with the notion of "emergent grammar," which postulates that grammar in the individual emerges as a process of borrowing "observed repetition in discourse"

75 It should be pointed out that research on memory within the information-processing paradigm has been extolling the value of silent repetition and rehearsal since the 70s (see discussion of mental rehearsal in Chapter 1).

198

CHAPTER 7

(Hopper, 1998, p. 156): "In [emergent grammar] . . . it is held that utterances are closely similar to previous utterances, and that anything that is said has been said in something like that form before" (p. 165). Repetition and imitation, from an emergent grammar point of view, are essential activities in the appropriation of L2 social speech and the heteroglossic character of L2 inner speech.76 This apologia of repetition should not be taken as a call for the revival of drills and pattern practice in the language classroom. When used routinely and monotonously, repetition can, as Ohta (2001) says, "erode interest in a task, thereby making a task less meaningful to learners" (p. 256). Rather, this is a bid for the recognition that repetition and imitation are fundamental human processes in the perpetuation, reconstruction, and transformation of social practices, including linguistic practices, and that whether teacher-encouraged or not, learners do in fact resort to covert, if not vocal, repetition and imitation of the L2. Lively and meaningful strategies for systematically incorporating repetition in the L2 classroom for the purpose of internalization will be offered further on. Eavesdropping is another behavior that seems to be naturally deployed by L2 learners (certainly deliberately overhearing other people's conversations is not something teachers encourage students to do!) and that seems to contribute to internalization. Eavesdropping has been noted in private speech studies of children (Saville-Troike, 1988) and has been discussed by Lantolf (2003) in relation to L2 learners. The impact of eavesdropping on the mental rehearsal of the L2 is plainly seen in the following verbal reports from beginning ESL learners (author's data for Guerrero, 2004): I thought all of a sudden of the English greeting: How are you today ? It's because sometimes English teachers get on the elevator with me and they greet each other and [this greeting] stays in my mind.77 Today I was at the bakery with my father and he met an American friend. They were talking and my father's friend said, Call me and I go to your house. I kept thinking about it because it was something I understood.

Eavesdropping is a good example of what students privately do, independently from teachers' prescribed practices, with features of the L2 they notice in other people's interactions. Again, this is not a suggestion for teachers to start encouraging students to snoop on other people's conversations but an invitation for teachers to acknowledge the fact that students are much more proactive in their covert efforts to internalize the L2 than might be inferred from their external behavior.

76 Both "appropriation" and "heteroglossia"-the reproduction and reverberation of other people's voices in an individual's mind-are well-developed concepts in Bakhtin (1981) and Wertsch (1998). 77 These student reports are translations of diary entries originally written in Spanish. Text in boldface was written in English in the entries.

A PEDAGOGICAL PERSPECTIVE

199

INSTRUCTIONAL MEDIATION IN THE INTERNALIZATION OF THE L2 AND DEVELOPMENT OF L2 INNER SPEECH Providing Opportunities for Engagement and Participation in L2 External Activities Teachers and teaching environments can foster the internalization of the L2 and contribute to the development of L2 inner speech in many ways. High among the recommended teaching practices is the promotion of language use and interaction in the classroom. Without social exposure to and participation in external activities in the L2, there would be nothing to internalize and no social support for sign-mediated activity at the interpersonal level to be eventually recreated at the intramental level. As Vygotsky (1986) put it, "essentially, the development of inner speech depends on outside factors" (p. 94). It may sound banal to say that inner speech is a derivative of external speech and that social interaction in the language classroom should be promoted, yet it is necessary to say so in response to claims, perhaps inspired in Piagetian beliefs, that developing an effective inner voice in the L2 should be emphasized before requiring students to publicly interact in the L2. Tomlinson (2001), for example, believes that L2 learners should be given time to develop an L2 inner voice before they are required to engage in social interaction. Such a view is predicated on the notion that this is what happens in LI learning: "When we learned our LI . . . we talked to ourselves before we talked to others and even when we talked out loud we were often using a private voice which was self directed" (italics added, Tomlinson, 2001, Inner voice and language learning section, para. 2). It is very difficult to conceive how learners, LI or L2, can talk to themselves if they have not been engaged before in interactive and responsive encounters with other speakers. The reverse of Tomlinson's argument is clearly stated by Galperin (1967) in reference to children but also applicable to mature L2 learners: The formation of speech, fully and accurately embodying an action and severed from its material objects and means, is only possible under other people's control. Their requirements teach the child to speak not in the way that seems correct and intelligible to him, but in order to be understood by others and to communicate the objective content of an action. The child learns to hear himself "from the outside" and to evaluate his speech from the viewpoint of others This form of speech activity can already be regularly and systematically transferred to the intellectual plane, (p. 30)

It is in social interaction with L2 speakers and L2 social artifacts (books, TV, the Internet, etc.), in the tension between external L2 models and learners' L2 verbalizations or understandings that learners begin to construct an intramental plane mediated by the L2. The arduous transformation of external L2 speech activity into internal L2 speech activity takes place in the social context of public and private instantiations of the learners' ZPDs. Without the social element, either physically present in the external environment or mentally invoked on the internal plane, development of L2 inner speech cannot take place. An effective L2 inner voice is thus not a prerequisite for L2 learning but rather an outcome of L2 learning. Acknowledgment of the ontogenetically social precedence of inner speech, however,

200

CHAPTER 7

does not mean that certain internal activities suggested by Tomlinson (2000, 2001), such as repeating and planning, cannot be viewed as useful in the development of L2 inner speech. Provided L2 social speech is not neglected and not viewed as the developmental outcome of L2 inner speech, inward repetition and planning do have a role (to be seen below) in the processes of internalizing and externalizing the L2. Fostering Inter nalization and Externalization of the L2 Repetition It has been noted that silent or subvocal repetition is one of the most frequent strategies L2 learners employ in learning the L2. Its value in helping learners internalize the L2 has already been discussed in this chapter. What remains to be seen is what L2 teachers may do to promote the strategic use of repetition in the L2 classroom. Before considering ways of deliberately stimulating repetition, it is important for teachers to realize that L2 learners do privately repeat what they hear or read; however, the greatest benefit of repetition comes about when learners repeat what they think it is necessary to repeat. In sociocultural theory terms, learners advance in their L2 development when they repeat or imitate what is within their ZPDs. Choral repetition as a routine and mechanical classroom practice is therefore not advisable. Parrot-like drilling robs repetition and imitation of intentionality and thus of a clear goal for learners; and without a real purpose, drills soon become meaningless and boring. Following are two lively and ingenious pedagogical techniques showing how repetition can be turned into a useful resource in the internalization of an L2. Repetition as a tool ofL2 communication and cognition. Roebuck and Wagner (2004) conducted a study to find out whether repetition could be taught as a means of conversational and private mediation in the L2. In the study, a group of fourthsemester college Spanish learners, native speakers of English, were instructed on the use of interactive repetition through a series of classroom activities. One of these was "el mentiroso" (the liar), an exercise in which one learner reads an implausible statement to another learner, for example, "I am the daughter of a famous actor," and the listener responds repeating part of the statement, for instance, "Daughter of a famous actor? I don't believe it!" Although this activity was not too cognitively demanding, it served to introduce learners to the concept of repetition as a conversational strategy. A more challenging form of strategic repetition involved getting students to interview each other and, at the end of the interview, to summarize what their partner had said. The researchers found that as students practiced in class varied forms of interactive repetition they produced briefer segments in their repetition and instances of whispered "private speech" repetition. To the researchers, these signs indicated that internalization of repetition as a cognitive tool was taking place. To further determine that repetition was having an effect on the students' process of internalization, Roebuck and Wagner analyzed a series of dramatizations the students produced during

A PEDAGOGICAL PERSPECTIVE

201

the semester, the first one taking place before instruction on repetition had been given. The number of scripted (planned) and unscripted (spontaneous) repetitions increased significantly after the introduction of repetition, again suggesting that students had appropriated the repetition strategy and were able to use it as a strategic cognitive and conversational tool. Shadowing and Summarizing. Another classroom technique that relies on repetition and has great potential for enhancing inner speech processes such as phonological apprehension, fixation in memory, and semantic condensation is "shadowing" (Murphey, 1995,2000,2001). The basic procedure in shadowing consists of listening to a speaker's discourse and repeating (or shadowing), silently or aloud, what the speaker says. Silent shadowing can be employed, for example, when listening to a teacher's lecture;78 loud shadowing is done in dialogic encounters. There are three main forms of shadowing: complete, selective, and interactive. Complete shadowing occurs when all that a speaker (SP) says is repeated by the shadower (SH): SP: SH: SP: SH:

I'd like to tell you about Boston. Boston, yes You'd like to talk with me about Boston. Boston is in America, in the north east part of America Boston is in America in the north east part of America (Murphey, 2001, p. 136)

According to Murphey (2001), complete shadowing is helpful in training people to shadow and showing speakers how to "chunk" their utterances so that the shadower can keep up with the repetition. This type of shadowing is particularly useful for beginning learners, who may need practice with pronunciation and may still not be able to repeat in more creative or selective ways. Murphey (2001) reports results indicating that silent shadowing had great impact on students' "learning, increasing attention and retention of material in short-term memory" (pp. 132-133). Selective shadowing entails repeating only certain utterances at the discretion of the shadower, usually key words or problematic utterances. Interactive shadowing entails not only selecting segments of a speaker's utterances but also adding comments or questions to the repeated segments. These types of shadowing are more appropriate for intermediate and advanced learners. As can be seen below, selective and interactive shadowing offer the possibility of greater involvement and more meaningful processing than complete shadowing. SP: SH: SP: SH: SP:

came to Boston. Uh, from from England, to Boston. came to Boston? England Ah, Plymouth Right, Plymouth Plantation Yes the Pilgrims came to Ah, Primos Primoth yes uh, Plymouth Ah, you know Plymouth Ah, world

78 Although silent shadowing does not usually involve modification of input on the part of the speaker, Murphey (1995) notes that when he is speaking and sees students moving their lips, as if silently shadowing, he becomes more aware of the need to slow down and chunk his speech.

202

CHAPTER 7

SH:

Ah yes I majored in world history (Murphey,2001,pp. 137-138)

Murphey (2001) believes that teaching shadowing may "be an effective way to generate private speech in a foreign language" (p. 148). By repeating silently or aloud a speaker's utterance, the shadower is pushed to process language internally, focusing on salient features of the speech heard and condensing language into brief meaningful segments. Importantly, shadowing may stimulate inner speech because it obliges the shadower to process what is heard and immediately verbalize it again without much chance for translating on-line. Shadowing may also help in shaping inner speech on the basis of external models adding to it the character of heteroglossia. A variant of shadowing is "summarizing," that is, listening to a series of speaker's statements and summarizing what has been understood (Murphey, 2000). Summarizing requires greater attention to incoming speech, increased focus on key ideas, more complex linkages to conceptual representations, and a higher degree of retention in memory than other forms of shadowing. Repetition, in short, is both a spontaneous behavior naturally deployed by L2 learners and a strategy that can be deliberately introduced by teachers to foster attention to L2 social speech, engagement with L2 models, on-line and delayed processing, and internalization of the L2. Murphey's as well as Roebuck and Wagner's instructional uses of repetition seem worthwhile techniques to apply for the development of L2 inner speech. Taking Notes and Paraphrasing An activity that stimulates inner speech is note-taking. John-Steiner (1985b, p. 365) observed that, when mature, well-educated L2 learners who had already developed effective note-taking skills in their LI tried to take notes on a lecture or speech delivered in the L2, their first strategy was to translate to their LI and write down what they heard in translation. Their next, more efficient strategy, was to attempt writing down what the speaker said in the L2, even if it was not completely understood. Only later were the learners able to listen carefully, identify key words or concepts, and write down condensed notes in the L2 that reflected the main ideas. John-Steiner attributes this superior type of note-taking to the learners having internalized the L2 and to their process of weaving the LI and the L2 together at the conceptual level. Although greater proficiency in the L2 no doubt contributes to better note-taking in the L2, learners at all levels could profit from deliberate instruction on note-taking. Stressing the need to concentrate on focal points and to refrain from writing down word for word what a speakers says may lead L2 learners to make better use of their inner speech when attending to speech in the L2. Another skill that requires effective use of inner speech and that encourages appropriation and transformation of L2 speech at an internal level is paraphrasing. Paraphrasing entails not only understanding the meaning of an utterance-as Vygotsky (1986) would say, comprehending the "subtext" or "thought hidden behind" someone's words (p. 50)-but also reproducing that subtext in words that are different from the

A PEDAGOGICAL PERSPECTIVE

203

speaker's and yet faithful to the intended meaning of the utterance. A pedagogical activity recommended by Rohrer (1986) that relies on stimulating meaning connections in inner speech is "articulating the unsaid." Rohrer suggests getting students to respond to such statements as "It's getting late" with rejoinders such as "Do you want to leave?" In this activity, learners have to activate their inner speech by accessing their semantic stores and internalized verbal scripts to guess the illocutionary intent of the speaker's utterance and respond appropriately. Planning Giving students the opportunity, the time, and the language support necessary for them to try out the L2 mentally or privately before actual speaking or writing is a desirable practice in the L2 classroom. As argued in Chapter 6, inner speech verbalizations are necessary not only as a phase in the externalization of the L2 but also as a way to internalize the language by noticing discrepancies with target language forms, finding means of possibly solving them, testing hypotheses, and developing metalinguistic awareness. Sources of strategic assistance should be facilitated to students as they mentally plan for public speaking or writing, for example, providing students with prompts or checklists to guide their thinking, making it possible for students to jot down ideas and organize their text or speech, permitting the use of dictionaries or other sources of reference, allowing students to consult their teachers or classmates, and-not least-resorting to the LI (or other L2s). The implementation of planning in the L2 classroom finds empirical support in a large number of studies pointing out the benefits of pretask planning in L2 production. Although results are mixed in terms of increased grammatical accuracy when planning is manipulated experimentally, in general, research shows that planning enhances L2 production in measures of fluency, grammatical complexity, and lexical richness (Crookes, 1989; Foster & Skehan, 1996; Menhert, 1998; Ortega, 1999; Wigglesworth, 1997; Yuan & Ellis, 2003). Because factors such as length of planning time, type of task, proficiency level, and learner styles may differentially affect the nature of planning and thereby performance, it is suggested that teacher intervention should be geared to show learners how to make best use of their inner speech while planning for different activities under different planning conditions. Problem-solving Tasks Another way to encourage increased inner use of the L2 is to get students involved in challenging problem-solving tasks in which language support is in the L2. As far as possible, the aim of these activities should be for students to mediate their verbal thought through the L2. A. A. Leontiev (1981) believed that teaching methods should be geared not only to foster the communicative use of the L2 but also to engage the learner's intellectual, perceptive, and creative skills. One study (Centeno-Cortes & Jimenez, 2004) that employed problem-solving in the L2 illustrates the extent to which teachers may expect their students to mediate thinking through L2 inner speech. Centeno-Cortes and Jimenez utilized a series of cognitively challenging questions to

204

CHAPTER 7

determine the nature of the private speech that L2 students (adult, English-speaking) employed to solve problems. Three groups of participants were included: native speakers of Spanish, intermediate learners of Spanish as an L2, and advanced learners of Spanish as an L2. The questions included mathematical, algebraic, and logical problems; kinship questions; and visual-spatial problems. Two of the problems (originally stated in Spanish) the students were offered may serve as illustration: Carlos bought a used car for $600 and sold it to Pedro for $800. He later bought it back for $1000 and resold it for $1200. Did Carlos make any profit and if so how much? Explain. (Centeno-Cort6s & Jimenez, 2004, p. 34) Carlos can never tell a lie. Pedro can never tell the truth. One of them said, "the other one said he is Pedro." Which one said that? Explain, (p. 34)

To answer the questions, which were presented via a web page, the participants were allowed to use paper and pencil and a bilingual dictionary. However, points were deducted from the total score for using the paper and pencil. This was done to encourage the use of private speech for thinking (PVT), which was the main focus of the research project.79 The results showed that more than half of the participants made no or little use of PVT, a finding that suggests that the participants were utilizing silent, inner speech to solve the problems. Furthermore, the researchers found that use of the L2 in private speech increased with proficiency, the more advanced learners being able to conduct most of their reasoning in the L2. These learners, however, tended to shift to their LI when the task became too difficult. Intermediate learners mainly used the L2 (Spanish) in reading the problem, repeating parts of it, formulaic expressions, and comments about the task. Their reasoning was mainly conducted in the LI (English). After reading aloud the problem in Spanish, the learners translated the problem literally into English and then re-phrased the problem in more meaningful English. The pedagogical implications of this study are strong. First is the fact that L2 learners do resort to the LI when solving problems in the L2, a fact that has been reiterated in various studies involving intellectually challenging tasks (Brooks & Donato, 1994; Brooks, et al., 1997; Cohen, 1998, see Section 6.4; Swain & Lapkin, 1998). In these studies, the LI emerged as a mediating tool, whether in private speech or in interactional discourse between learners of the same language background. Expecting students to put aside their most essential tool for thought-the Ll-when faced with cognitive difficulty is, as mentioned before, not only unrealistic but also pedagogically unwise. It is unrealistic and useless because teachers (or anyone else) cannot really control what goes on in the privacy of an individual's mind. In addition, if the true aim of the activity is the solution of a task, it makes no sense to bar the LI from the arsenal of tools at the learner's disposal. Another implication is that 79 Paper and pencil would have been an excellent artifact for the participants to mediate their thinking processes. For the purposes of the study, however, written notes would not have been too useful inasmuch as the researchers were interested in analyzing orally produced private speech. Other studies could compare private speech as revealed in written notes and in oral speech to find out differences between these two different forms of mediation.

A PEDAGOGICAL PERSPECTIVE

205

proficiency does make a difference in the extent to which learners use the L2 as a language for thought in cognitive problem solving. Advanced students can mediate their thinking processes through the L2 more than lower level learners; however, they can still fall back on their LI in case of great cognitive difficulty. Offering students the opportunity to engage in challenging problem-solving tasks is still valuable in terms of L2 inner speech development. Meaningful, contextualized, purposeful activities generate more interest in the learner. Having to wrestle with the L2 in order to solve a problem leads to greater involvement with the L2, even if translation into the LI is involved. In the process of understanding the premises of the problem and finding a solution for it, more meaning connections are built between LI and L2. It has been shown that depth of processing-such as occurs when meanings are activated and associations at a semantic level are made-the effects on retention in longterm memory are greater (Craik & Lockhart, 1972). Furthermore, in providing their answers in the L2, learners also need to construct meaning in the L2, even if that implies accessing first a conceptual semantic base in the LI. It seems then that instructional environments that emphasize meaningful, contextualized, creative, and challenging language teaching, such as teaching language through content, language for specific purposes, or language across the curriculum are potentially valuable for the development of the L2 as a tool for thought. Developing a Conceptual Foundation in the L2 Perhaps the most important single way to impact the learner's development of L2 inner speech would be the reconstruction of the learner's conceptual foundation through the integration80 of L2-related concepts and the building of L2 semantic networks and connections to lexical and grammatical forms in the L2. It has been seen that learners who develop a bilingual or multilingual conceptual base have done so by somehow modifying the pre-existing LI-related conceptual foundation.81 Conceptual change, however, has been found to occur mainly in cases of intense participation and immersion in L2, bilingual, and multilingual communities. The question is therefore whether it might be possible to create or replicate in the classroom the conditions necessary for conceptual change to take place. Three important factors seem to work against the recreation of optimal environments leading to conceptual modification on the basis of an L2: decontextualized language learning, an over-reliance on translation as a teaching method, and lack of or insufficient social interaction with L2 speakers or artifacts. 80 The integration of an L2 within an already developed LI conceptual foundation may be the result of three main processes-coexistence of LI and L2 concepts, transfer of LI concepts to the L2, and/or changes in the LI conceptual store as a result of L2 influence -as postulated by Pavlenko (1999) and discussed in Chapter 6. 81 Conceptual change may result from one or more of the following processes: addition of L2 concepts (internalization), shifting from LI conceptualizations to L2 conceptualizations, forming unitary LI and L2 conceptualizations (convergence), restructuring of LI concepts to accommodate for L2 features, and fading (attrition) of LI concepts and substitution by L2 concepts (Pavlenko, 1999).

206

CHAPTER 7

The teaching of language in context, and not as isolated linguistic components such as vocabulary lists or grammatical rules, is generally agreed to be conducive to internalization of an L2. A. A. Leontiev (1981) recognized the "overwhelming necessity of introducing words in context" (p. 56). One way of contextualizing words, according to Leontiev, is presenting lexical items in meaningful texts; another one is introducing new words as part of functional systems or networks of related meanings. Leontiev suggests that semanticization of words (acquiring the basic meanings of words) should be encouraged by providing the different contexts in which a word might occur. Semanticization of words may also be achieved through visual support, that is, "with the help of a few pictures which will cover all the basic uses of a given word" (p. 57). Visual support, however, may not be enough. With certain concepts it might be desirable to stimulate other sensory associations (olfactory, tactile, etc.) as well as connections to specific events and concrete referents. An approach to contextualized L2 teaching should include not just "providing" L2 learners with all the possible meanings of words but also allowing learners to construct linkages between L2 words and their own experiences. In Vygotsky's (1986) terms, learners should not only learn the objective "meanings" of words (socially agreed relationships between signs and referents, as stated in dictionaries) but also attribute to them their own subjective "senses." Kramsch (2000) explains this by postulating, as Vygotsky did, the non-arbitrary nature of signs. Signs, "are created, used, borrowed, and interpreted by the individual for the purposeful actions in which he/she is engaged" (Kramsch, 2000, pp. 133-134). In Kramsch's view, learners are capable not only of using signs with meanings established by others but also of creating signs with new meanings by relating signs to other signs. In learning an FL, the first thing learners acquire is the relationship that signs hold to their referents (objects, events, or abstractions); they learn, for example, that "rain" stands for "water that falls from the sky." Learners should also come to understand, however, that they can relate signs to other signs, for instance, that "rain" can be linked to "a ruined vacation" or "bountiful crops" (p. 135). In this process, they can imbue with new meaning-invest with their own "senses"-other people's words. A very good example of the subjective connections that students spontaneously make in learning new words was offered by the learner (cited in Chapter 5) who associated the new word "robbers" to an incident in which a CD player had been stolen from a friend's car. In fostering L2 inner speech through contextualized language teaching, it is therefore vital to help learners build rich associations of L2 words, not just to concrete or abstract referents, visual images and sensory experiences but to other words with which they are related in meaningful-systemic, logical, and subjective-ways. Importantly, teachers may help students grasp word relationships that may not be transparent from their single use in different contexts. For instance, it might be useful for students to know the metaphorical origins and dimensions of many of the words they encounter in the L2. Students might be alerted, for example, to the fact that in English it makes sense to speak of "enlightening ideas" or "being able to see the truth" because underlying these utterances there is a pervasive metaphor stating that "knowledge is light."

A PEDAGOGICAL PERSPECTIVE

207

Equally useful might be acquainting students with the interrelationships that may exist between the multiple meanings of polysemous words. One such type of relationship is that between core and. peripheral meanings of a word. Verspoor and Lowie (2003) suggest providing learners with the core senses of words, that is, those essential meanings that underlie all senses of polysemous words, so that learners can then infer the peripheral or figurative senses of those same words. For example, giving students the core meaning of the word nugget, as in "gold nugget," can help them infer the meaning of nugget, as in "chicken nugget" or "a nugget of information." Providing learners with the peripheral senses of the word nugget will not enable learners to see how the two senses are linked to each other nor their relationship to the core sense. As the researchers put it, "a small piece of batter-fried chicken has the same shape and color as a gold nugget, and a small piece of information may be valuable as a gold nugget is valuable, but what does a small piece of batter-fried chicken have to do with a useful piece of information?" (p. 568). Results of a study conducted by Verspoor and Lowie confirmed their hypothesis that giving learners a core sense is more effective than giving them a noncore sense when they try to guess the peripheral meanings of unfamiliar polysemous words. The authors believe that "textbooks for beginners would do well to introduce new polysemous vocabulary items by presenting their core sense first, because these will provide a good basis for guessing a more figurative sense encountered later" (p. 569). The authors also speculate that learners might learn to apply on their own the method of finding out first the core sense of a word, as a strategy for looking up words in a dictionary. To develop a rich conceptual base that will accommodate the L2, it seems imperative not only to teach the language in context and to help students develop multiple senses of words but also to free the L2 from its sole connection to the LI. The exclusive or predominant use of translation in teaching a new language does not foster the formation of direct linkages to conceptual images, sensory representations or episodic memory. In translation, the learning of new items in the L2 is always indirect, that is, through the conceptual sieve of the LI. L2-related concepts are learned as they are interpreted in the LI, thereby leading to what is known as conceptual 'transfer," rather than conceptual "coexistence" or "restructuring" (Pavlenko, 1999). Reliance on translation as a method to learn the meaning of new words will inevitably result in what Ushakova (1994) described as the "plugging" of the new lexicon "into the inner speech mechanisms that have been worked out with respect to the first language" (p. 136). Although learners, especially at the beginning stages, might naturally resort to translation in their minds and find in the LI a powerful tool to make meaning, it should be the task of language instructors to ensure that other than LI connections made via translation are established in the L2 or FL classroom. According to Kroll, Michael, and Sankaranarayanan (1998), it might be possible to overcome dependence on the LI if cues (semantic information) for the L2 are sufficiently distinctive. The researchers hypothesize that the success of immersion in breaking the exclusive link to the LI may be attributable not only to increased frequency of exposure to the L2 but also to the presence of unique cues in the environment that might "reduce LI competition and enable conceptual processing in L2M (p. 391).

208

CHAPTER 7

The third factor preventing the development of a conceptual foundation that might enable L2 inner speech is minimal or absence of interaction with L2 speakers. This situation tends to occur in settings where the new language is learned as a "foreign" language and where learners lack, as Pavlenko (1999) points out, "contextualized interactions with members of the L2" capable of providing "enough context to form an experiential multi-modal representation which goes beyond word definition and forms a concept" (p. 222). On the other hand, Pavlenko goes on, "natural environment learners acquire new concepts contextually, interactively and experientially, and, as a result, incorporate them into their restructured conceptual systems, whereby they can be available for both recognition and recall" (p. 222). Instructional settings based on language immersion programs, content-based or bilingual curricula, or L2 classroom learning supplemented by untutored participatory experiences with an L2 community, appear to be particularly fertile environments for the development of an L2-related conceptual base. Although the ties to the LI may always remain and in fact continue to enrich the learner's cognitive and verbal processes, as discussed in an earlier section above, some new concepts that are solely connected to the L2 may be acquired and old concepts may be reshaped on the basis of newly acquired senses in the L2. Particularly in the case of content-based and language immersion programs, a whole new universe of L2 concepts associated to the language of various academic fields may be internalized. Vygotsky (1986) argued that concepts are not static, that they develop over the lifetime of an individual. It seems quite probable then that, given the right instructional circumstances, the concepts that shape, inform, and constrain an individual's thought might also grow and expand with the internalization of an L2 and its interweaving with the LI. Raising Awareness about Inner Speech The last pedagogical suggestion that will be made here concerns the desirability that L2 teachers bring to awareness the topic of inner speech among their students. Awareness of inner speech is not something that comes about automatically with its development. In learning or using an LI or an L2 people may be aware of the fact that they "use words in the head" or that they "talk to themselves" silently, yet full consciousness of the nature and functions of inner speech is not an inevitable realization. Neither are people very conscious of the possible benefits of engaging deliberately in inner speech or of the damaging consequences of negative self-talk. L2 learners may profit from a heightened awareness of inner speech in several ways. Among other benefits, L2 learners may intentionally deploy their L2 resources through inner speech for higher retention, mental practice, or better production; they may increase their confidence as L2 users; and they may create a favorable self-image as L2 intra-speakers. According to Vygotsky (1986), awareness and control of higher mental functions emerge only after the function has been internalized: The general law of development says that awareness and deliberate control appear only during a very advanced stage in the development of a mental function, after it has been

A PEDAGOGICAL PERSPECTIVE

209

used and practiced unconsciously and spontaneously. In order to subject a function to intellectual and volitional control, we must first possess it. (p. 168)

Flavell, et al. (1997) contend that children acquire awareness of inner speech during the early school years, probably as a result of the emphasis placed by instructional contexts on silent verbalization in the activities of reading, writing, adding, and subtracting. L2 learners who are past childhood and already "possess" inner speech in the LI could be introduced to the notion of L2 inner speech. L2 learners may profit from their awareness of L2 inner speech and their capacity to control it. Although many inner speech processes are out of the reach of consciousness, learners can take an active role in those phases of inner speech that are under their attention and therefore available for control. Vocate (1994b) points out that one of the stages of inner speech in its transition from thought to word is the level when inner speech is sufficiently syntactic and lexical to support conscious self-talk, or communication with oneself. L2 learners can intentionally access their inner speech in the self-talk mode and impose some selfcontrol on the processes of internalizing, externalizing, and practicing the L2. Teacher intervention is crucial in the development of self-awareness of L2 inner speech. It is therefore important to introduce the notion of inner speech not only to learners but also to L2 teachers. Streff (1984) points out several reasons for the "teaching of inner speech" in LI instruction, ideas that are also quite relevant to the L2. Teaching about inner speech, according to Streff, is worthwhile because it (a) develops the notion of human communication as process (Learners come to understand the nature and complexity of speaking, writing, listening, and reading and their relationship to inner speech.) (b) facilitates the development of decentering (ability to take another person's point of view) and displacement (ability to transcend time and space through language) (c) contributes to an awareness of the need for semantic and syntactic elaboration in verbalizing inner speech (d) facilitates the learning of contextuality (importance of making contextual features explicit in social communication) (e) fosters an appreciation for the uniqueness of the human condition In particular, learners could be introduced to the benefits of deliberately engaging in rehearsal-type inner speech. Among the benefits that students can be made aware of are higher retention of new vocabulary and formulaic utterances, learning and consolidation of language forms, phonological memory and accuracy, grammatical fluency and accuracy, comprehension, and metalinguistic knowledge (see, for example, Baddeley et al., 1998; Hu, 2003; Ellis & Sinclair, 1996; see also discussion of rehearsal benefits in Chapters 1, 2, and 3). Learners who have limited exposure to social interaction in the L2 or FL could find in dialogic inner speech an additional source of practice in the language. Learners may also profit from an awareness of their capacity to monitor speech production by engaging in preparatory inner speech. Finally, students may learn about the affective component of inner speech and how they can use it to their advantage by engaging in self-encouraging self-talk. Employing the L2 deliberately through covert rehearsal may conduce to the building of a positive L2

210

CHAPTER 7

identity. Learners may find in their efforts and ability to handle the L2 in the mind a source of pride and high self-esteem, learn to feel comfortable with their "bilingual" status, discover the freedom to express themselves and create meaning through another language, and take stock of their own power to control the L2 through mental manipulation. A specific area where deliberate attention to inner speech can be exercised is in the externalization of inner speech in speaking and writing. Knowledge of the characteristic features of speech-for-oneself: brevity, fragmentariness, predominance of sense over meaning, condensed semantics, omission of the "psychological subject" (or known information), under-referenced pronominalization, shifting focus, and overall crypticality would help L2 learners understand the need for full expansion, preciseness, clarity, even redundancy, in speech-for-others. Knowledge of the types of adjustments that are necessary for inner speech to become comprehensible to others would also be useful to learners. Teacher intervention is therefore crucial in helping learners perform the transition from inner to social speech successfully. It is also important, however, for teachers and students to understand that externalized private speech or private writing is not "bad" speaking or writing but an essential aspect of the normal externalization of speech process. In reference to private writing, which is perhaps the most visible form of self-directed speech, DiCamilla and Lantolf (1995) contend: "The written externalization of students' inner dialogue as they attempt to regulate themselves in the writing task is a critical, if not indispensable, part of their writing process" (p. 364). Whereas sophisticated writers learn to exploit this type of writing in notes, outlines, or rough drafts, unsophisticated writers extend private writing to final drafts in their efforts to regulate themselves. The role of private writing and its related processes of planning and rehearsal should be made clear to students so that they can make full and appropriate use of it. Making L2 students aware of the value of inner speech as a fundamental mechanism of the mind and pointing out its functions and uses in the L2 may not only stir their curiosity into experimenting with it consciously but also add a new understanding of what it is that makes humans higher intellectual beings. As Flavell et al. (1997) express in relation to the development of awareness of inner speech among children, As they become increasingly aware of the existence of inner speech as a cognitive activity and increasingly able to notice its occurrence when they engage in it, they should come to realize that it occurs frequently and can take many forms: rehearsing the past or planning the future, verbal problem solving, daydreaming, and fantasizing, worrying and obsessing, and so forth. And with this realization, they will have learned a lot about what people's inner lives are like. (p. 46)

CONCLUSION In this chapter it has been argued that L2 inner speech cannot be developed independently or prior to engagement in social activities mediated by the L2. Whereas development of L2 inner speech is most likely to take place in settings of intense L2 immersion and participation, propitious conditions may be recreated in classrooms that

A PEDAGOGICAL PERSPECTIVE

211

may lead learners to internalize and put the L2 to use in intramental activity. A. A. Leontiev offers some ideas on how L2 teachers can help learners mediate their thinking processes through the L2 without reducing the activity to mere translation of readymade LI propositions into the L2. Throughout the chapter, teachers are encouraged to acknowledge the LI as a critical thinking tool that learners have already internalized and will resort to at all levels of L2 proficiency, principally for meaning making and as support in moments of cognitive difficulty. The presence of the LI in the learners' minds does not mean, however, that the L2 cannot be developed and coexist with the LI as an additional tool for thought. Effective intramental use of the L2, alone or in conjunction with the LI, is essential for becoming literate in the L2. Whereas learners may naturally resort to strategies, such as repetition, eavesdropping, and other types of rehearsal, that may enhance development of L2 inner speech, teachers can deliberately promote internalization and self-directed use of the L2 in various ways. First, it is important to recognize the ontogenetic precedence of social speech in the evolution of inner speech and the need to provide learners with opportunities for interaction in social activities conducted in the L2. Interactive activities that can promote L2 inner speech are conversational repetition, shadowing, summarizing, taking notes, and paraphrasing. Planning and problem-solving tasks can also foster heightened inward use of the L2. The evolution of L2 inner speech could not take place without somehow altering the learner's conceptual foundation acquired through the LI. The fact, clearly stated by Vygotsky, that concept development is a dynamic, life-long phenomenon allows for the possibility of expanding the individual's conceptual and semantic resources through the internalization of another language. Contextually rich environments that can provide multi-modal experiences in the L2, freeing the L2 from its sole connection to the LI, and fostering the construction of new senses for L2 words appear to be auspicious for the development of a conceptual foundation related to the L2. Lastly, it is believed that bringing the topic of inner speech to awareness may contribute to an increase of the learners' intentional use and control of the L2 and to an enrichment of their linguistically-mediated inner worlds.

CHAPTER 8 SYNTHESIS AND DIRECTIONS FOR FURTHER RESEARCH

This book has taken a look at L2 learning as a process of developing the faculty to think words through the L2. This was the faculty that Vygotsky described in relation to the LI as the outcome of the progressive individualization of social speech and its culmination in inner speech. Historically, inner speech has been studied from an LI perspective. In the 1980s, however, largely as a result of the emergence of a sociocultural approach to L2 language learning, inner speech and its audible counterpart, private speech, began to be taken notice of in the L2 learning field. In this book an attempt has been made to gather the existing research on L2 inner speech and related topics in order to provide a comprehensive view of the origin, development, and nature of inner speech in the L2. In this chapter, a synthesis is presented on the fundamental question: How do learners develop the faculty to think words in a language other than their LI, and what does this faculty consist of? This chapter also lays out some of the avenues of research on L2 inner speech that might be attractive, useful, or necessary to pursue in the future. Traditionally associated with the LI, inner speech has been acknowledged since ancient times as a phenomenon in which thinking and speech converge, enabling humans to entertain verbal thoughts. Broadly defined as "a silent manifestation of speech directed to the self (see Chapter 1) inner speech has been found, from the sociocultural theory perspective underlying this volume, to be essential in the development of higher psychological processes and in the making of a self-conscious, sign-mediated mind. Human beings normally develop LI inner speech, that is, thought mediated by the LI, as a result of the interiorization of LI social speech. In this book it has also been postulated that people who learn an L2 may also develop inner speech in the L2 and that this inner speech originates in L2 social speech. The internalization of social speech-in any language-is a process of privatization of speech functions, and the inner speech that results operates primarily for the benefit of the individual, yet inner speech never loses many of its social features: heteroglossia, dialogicity, culturally and historically situated meanings, and social discursive forms of reasoning. In LI learning, the ontogenetic development of inner speech is characterized by a transitional phase known as private speech, a vocalized form of self-directed speech that has mainly a cognitive/regulatory function. In time, private speech goes

214

CHAPTER 8

underground as inner speech but does not disappear. It may resurface as a means of regaining self-regulation under conditions of cognitive stress. Children who learn two languages simultaneously or an L2 in early childhood develop private speech in both languages. Many L2 learners, however, begin learning the new language when they have already internalized their LI and developed the faculty to think words in their LI. These learners have already grown into the distinction that loud speech is for others and silent speech is for the self. Adult or adolescent L2 learners do display loud private speech, both as a means for internalizing the L2 as well as for regaining cognitive control. Yet, apparently, not all L2 learners make use of L2 private speech for internalization purposes, and whether private speech is a necessary phase in L2 development is still an open question to be elucidated with future research. It seems likely that, for some learners, private speech may be substituted by subvocal or covert verbalizations of the L2. The covert process of L2 internalization and development of L2 inner speech is characterized in its early stages by emphasis on the transformative reproduction of L2 social speech. As learners come into contact with the L2 and interact with L2 speakers, they are inwardly reproducing the L2 through concealed verbalizations, a process that is not just the simple copying of external models but a creative and selective process of appropriating different features of external speech and making them the learners' own. Concealed verbalizations happen not only as a reaction to outside L2 stimuli but also as a form of recall or delayed language processing through which the L2 is subjected to further manipulation and rehearsal. At the early stages of L2 development, learners also begin to use the L2 intramentally in preparation for future or imaginary oral production. Beginning L2 learners, however, seem to make little use of L2 inner speech for complex mental operations. The L2 has not been internalized enough for learners to conduct sustained inner speech activity in the L2 during complex cognitive tasks, and learners may in fact rely on LI inner speech to solve or engage in cognitively challenging tasks in the L2. What seems to predominate among beginning L2 learners is an incipient form of L2 inner speech characterized by echoic and transformative attempts to appropriate the L2 in the surroundings. L2 inner speech at this stage has mainly a rehearsal-for-internalization function. Through the inward reproduction and rehearsal of external L2 speech, a temporary internal plane of psychological activity in the L2 is formed until a more permanent stage of L2 inner speech develops on the basis of internalized conceptual, semantic, lexical, grammatical, and phonological aspects of the L2. Because the internalization of an L2 does not occur uniformly across all aspects of an L2 and in all cognitive or discourse domains, there may be partial use of L2 inner speech in those areas in which competence in the L2 has been developed. However, there are multiple other factors that may affect an individual's intramental use of the L2, among these, contextual factors, such as whether the L2 is the dominant language in the environment, as well as personal factors, such as degree of acculturation to the L2 community or personal preference. Some individuals may resist to acculturate to the L2 and rather than reconfigure their LI culturally shaped conceptual foundation may simply opt for transferring LI meanings to the L2; other individuals may feel

SYNTHESIS AND DIRECTIONS

215

comfortable using the LI for certain private uses-praying, for example-and the L2 for other purposes-self-talking, for instance. Extensive reconfiguration of a person's conceptual and linguistic knowledge, may, however, take place-granted certain conditions, such as very high levels of L2 attainment and acculturation, intensive participation in the L2-speaking community, personal disposition, and the like-and changes in the nature of the inner speech through which a person introspects and becomes self-conscious may even lead to reconstruction of the self and the formation of a new linguistic identity. Two important features, inherent to the nature of inner speech, have been noted a regards L2 inner speech. One is the notion that L2 inner speech is activity, not an internalized L2 system per se but the very act of deploying this system in the on-going and dynamic business of thinking through words and self-communicating. The second feature of L2 inner speech to note is that it involves process, not just in terms of ontogenetic growth-which has been discussed earlier-but also in terms of microgenetic changes in speech production and perception. It has been postulated that the transition from thought to external speech and vice versa at early stages of L2 development is mediated by the LI, the process being essentially one of translating LI meanings to L2 forms when speaking or writing and L2 forms to LI meanings when reading or listening (this does not mean, of course, that speech production and perception are mirror processes). At very advanced levels of L2 development and in the presence of certain conditions such as contextualized language learning and a high degree of acculturation into the L2 community, enough restructuring of the conceptual and semantic foundations of inner speech must take place to make translation from or into the LI unnecessary. However, it has been strongly pointed out, that L2 learners tend not to lose their LI connections, except perhaps in extreme cases of LI attrition or aphasia, and therefore continue to count with their primary language as an important cognitive resource in verbal thought. In form, L2 inner speech, like LI inner speech, seems to be brief and compact when the cognitive task to be mediated so demands it. In other words, inner speech finds in the L2 the same structural elements that allow it to express a maximum of meaning with a minimum of words, particularly during automatic verbal processing. L2 inner speech may also be semantically, structurally, and even phonologically expanded in certain circumstances, for example, when planning what to say or write. L2 inner speech, like LI inner speech, though ostensibly silent, does have auditory and articulatory features accounting for the subjective experience of "hearing" or "saying" words in the mind. In function, L2 inner speech may serve two major purposes or macro functions: (1) as a tool for thought and self-regulation and (2) as a rehearsal mechanism forming a temporary internal plane of L2 mental activity aiding and contributing to the internalization of the L2. In its cognitive-regulatory function, L2 inner speech has been found to be involved in various mental operations such as remembering events encoded in the L2, problem-solving, and the execution of verbal tasks in the L2. Rehearsal-type L2 inner speech plays multiple roles: mnemonic (to memorize or retrieve L2 words from memory), instructional (to self-teach the language), evaluative (to self- and otherevaluate language), preparatory (to mentally prepare for speech production), dialogic

216

CHAPTER 8

(to imagine dialogues with the self and others), play (to playfully manipulate and experiment with language), and affective (to control or express emotions and address self-needs). In this book, it has been also contended that, though L2 learners may develop L2 inner speech independently of L2 instruction, for instance, through everyday life experiences in communities where the L2 is the chief means of communication, teachers and formal educational environments play a critical role in the development of L2 inner speech. One essential requirement that teaching must meet is the provision of opportunities for engagement and interaction with social L2 speech, without which internalization of the L2 could not take place. Very important are instructional techniques that foster inward appropriation and rehearsal of the L2, such as shadowing and assisted planning for speech production. Teaching can also impact development of L2 inner speech by helping students create links to the L2 at the conceptual and semantic levels. Educational environments that stress contextualized language learning, de-emphasize translation as a teaching method, and provide rich sources of interaction with L2 speakers or artifacts, seem most likely to create optimal conditions for the construction of direct meaning connections to the L2. Making learners aware of the role of inner speech in learning an L2 and the benefits they may derive from deliberately engaging in it is another way instruction can positively affect development of L2 inner speech. Four distinct areas where further research is worth pursuing have been identified: (1) aspects concerning the nature, development, and use of L2 inner speech, (2) effects of pedagogical intervention, (3) application of under-used or novel methodological approaches, and (4) continued theorizing. Within the first area, it seems valuable to continue exploring the private speech of L2 learners. One particularly neglected aspect of private speech (LI or L2) is the role of mutterings in the internalization of social speech. As suggested in Chapter 3, by studying mutterings, important information may be obtained on the microgenetic changes that characterize the transformation that social speech undergoes as it turns into inner speech. Another aspect of private speech that is in need of further study concerns the role of private speech in the internalization of the L2. Although some studies have begun to appear in this area, as discussed in Chapters 4 and 6, further research should aim at clarifying whether L2 private speech as internalization constitutes a developmental "phase" in learning an L2, if it is a widespread occurrence, and what type of learners are most likely to experience private speech as internalization oftheL2. The study of L2 inner speech as a developmental phenomenon would profit from longitudinal case studies tracing stages in the interiorization of social speech and documenting the attainment of the faculty to think words in another language. Although longitudinal studies of bilingual acquisition and L2 learning do exist, none has specifically looked at the development of private and inner speech in two (or more) languages and what types of ultimate accomplishments in terms of inner speech such learners display. Two types of longitudinal case studies would be important to be conducted: the study of simultaneous or early bilingual (or multilingual) acquisition and

SYNTHESIS AND DIRECTIONS

217

the study of L2 learning after inner speech in the LI has been attained. In the last case, several interesting possibilities arise in terms of types of L2 learners that might be targeted: children who start learning the L2 in school and might potentially develop very high levels of L2 proficiency throughout their academic instruction, migrants just arrived at an L2 community where they plan to establish life-long residence, learners who have separate contexts of usage for LI and L2 (e.g. LI for home and family, L2 for social, academic, and business relations), and so on. Such longitudinal studies would yield valuable information on the stages of inner speech development among different types of learners and on possible differential outcomes in terms of ultimate nature and use of inner speech. A question in much need of research is how inner speech changes on the basis of the internalization of one or more L2s. If John-Steiner's (1985b) claim that "the relationship of language to thought is not one of static connections; it changes with the shifting lines of development of the two languages" (p. 357) is correct and one agrees with Pavlenko (1999) that, in the study of bilingual and multilingual linguisticconceptual representations, one needs to account for the immense variety of personal, psycholinguistic, and sociocultural factors that are at play, it is important to open the study of inner speech to all the changes that may take place as individuals acquire new languages. Language development in LI or L2, like psychological development, has no "telos" (end point) (Lantolf, in press). Inner speech and its underlying relationship between speech and thought are also likely to be in a constant state of evolution, not just because the conceptual, semantic, and lexical stores in LI are always changing but because they may be significantly impacted by the acquisition of a new language. It is precisely this type of changes in inner speech, LI or L2, that needs to be further explored. A particularly interesting and under-researched aspect of change is the question of self-reconstruction and reconceptualization that may take place as a consequence of L2 learning. Although Pavlenko and Lantolf (2000) have made inroads into this aspect by focusing on the experiences of well-known, sophisticated L2 writers, the study needs to be extended to other types of, perhaps less publicly visible, L2 learners. The second area of further research that is envisioned concerns the impact of instruction on L2 inner speech. Although pedagogical research in L2 learning abounds, there is little research specifically focused on how teaching affects inner speech development, what formal aspects or functions of inner speech are implicated in certain classroom practices, and what instructional techniques might best promote effective use of inner speech among L2 learners. For example, how to foster transformative and creative covert repetition and imitation, how to guide students through planning and rehearsing for future production, and how to teach students to make adjustments to inner speech before externalization in speaking or writing are forms of teacher intervention that could profit from systematic research. Several routine classroom practices, such as note-taking, paraphrasing, and story-retelling, could be investigated to find out how they tap L2 inner speech processes and how they contribute to L2 internalization. Importantly, pedagogical research should be conducted to test proposals aimed at reducing the learners' reliance on translation from/to the LI as a

218

CHAPTER 8

meaning-making mechanism. For instance, would A. A. Leontiev's (1981) three types of rules guiding the transition from operations in the LI to those of the L2 be effective in "getting rid" of the intermediate LI translation phase in speech production? Lastly, it seems worthwhile to conduct research on the effects of making students aware of their own L2 inner speech processes and of their capacity to consciously deploy covert rehearsal and self-talk in the L2. Thirdly, the study of L2 inner speech would be greatly enriched with the application of specific research methodologies, so far explored for the most part in relation to LI inner speech. For example, the cognitive psychophysiology of L2 inner speech, except for the early seminal studies of Sokolov regarding FL texts, is practically unexplored. There is a vast amount of literature on LI inner speech detailing speech interference and electromyographic techniques that could serve as the basis for L2 research in the same area. Psychophysiological recordings of L2 covert linguistic activity could provide valuable information about L2 inner speech processes during particular L2 verbal tasks and cognitive operations. Within the "genetic" approach to inner speech, a methodology of research that has been under-applied in L2 and that would allow to make inferences about the nature of L2 inner speech is the analysis of self-directed, externalized private jottings in the L2, such as lecture notes, annotations in book margins, "to-do" lists, personal reminders, and the like, among classroom learners and more sophisticated L2 or bilingual users. Although a variety of verbal report methods have been used to obtain data on L2 inner speech, some techniques like Q-methodology and randomly sampling thinking have never been tried for that purpose. Thought-sampling, in particular, could provide useful immediate retrospective accounts of L2 inner speech experiences in natural environments. Other verbal report methodologies that have not been utilized in extensive longitudinal studies, such as learner diaries,82 could be conducted on a case study basis to document stages in long-term development of L2 inner speech. It would be very interesting to elicit such type of longitudinal diary data from learners who are likely to attain high levels of L2 proficiency or who might become the subject of bilingual or multilingual learning experiences. Highly promising and productive appears to be the application of neuroimaging techniques to the study of L2 inner speech. Although the neuropsychological study of L2 processing through PET and MRI is burgeoning, there is still a vast territory to be covered. Several hypothesized brain processes and anatomical areas implicated in L2 inner speech activity could be empirically corroborated through the latest brain scanning and imaging techniques whereas new findings may lead to new hypotheses. Particularly valuable about the application of neuroimaging techniques to the study of inner speech is the possibility of "directly" observing processes that are inaccessible through introspection. Where does L2 inner speech arise? What path does inner speech follow in L2 speech production and perception? Does L2 inner speech share brain regions and activation patterns with LI inner speech? These are some of the 82 In Guerrero's (2004) research, implementation of the diary was for just one semester and yielded only information on early stages of inner speech development.

SYNTHESIS AND DIRECTIONS

219

multiple questions to be explored with the new tools of neurolinguistic research. Finally, theoretical propositions about the nature of L2 inner speech should continue to be made. As further research clarifies obscure or contentious areas, brings additional information, and provides fresh new insights, theories about the role of an L2 in verbal thought and the way an L2 comes to enrich a person's intrapsychological plane should move forward to greater coherence, specificity, and accountability. Particularly necessary are continued efforts in refining inner speech production and perception models, taking into account LI, L2, bilingual, and multilingual learning scenarios. The model offered in Chapter 6 is a first attempt at hypothesizing the role of L2inner speech in speech production that could serve as point of departure for further theorizing. As understanding of the conceptual and semantic foundations of inner speech among bilinguals and multilinguals continues to be expanded and neurolinguistic and psycholinguistic research provides clues as to the distinct brain areas and processes implicated in L2 and LI speech activity, the model and hypotheses presented in this book should be subjected to revision. Ultimately, a comprehensive account of L2 inner speech should be sensitive to the great variety of social, cultural-historical, and individual factors involved in its development. The above does not pretend to be an exhaustive list of recommendations to be observed but a sample of topics and approaches that might be desirable to undertake in future studies of L2 inner speech. The author hopes that inspiration to pursue the investigation of this most fascinating and difficult to attain phenomenon of L2 learning, the capacity to think words in the L2, is to be found throughout the pages of this book.

APPENDIX

INSTRUCTIONS FOR KEEPING THE DIARY (ENGLISH VERSION) In this diary you are going to keep a record of your experience related to your inner speech in English [the learners' L2]. An important aspect of learning another language is how the person internalizes it, that is, how the person processes it internally. As you keep this diary, you will probably become more aware of your own role in learning English and of some of the strategies that can help you learn the language. First, let's define inner speech: Inner speech is any type of language in English that occurs in your mind and that is not spoken. Inner speech may include sounds, words, phrases, sentences, dialogues, and even conversations in English. For this diary, you need a notebook or folder with blank sheets to write on. You may also carry around a little notepad to take notes at any moment. You can later include these notes in your diary entries. You are going to write in the diary at the end of the class while you are still in the classroom and after class as often as you please but at least twice a week. Every now and then, I will ask you to show me your diary. You will also have to hand it in at the end of the course. Please write the date and time of each entry (for example, Friday, August 10, 9 p.m.). Try to set aside a specific moment during the day to write in the diary. Also, try to be as specific as possible in your description of your inner speech. You may include your reflections about your inner speech occurrences. Write in Spanish [the learners' LI]. If you want to write some word or phrase in English, you can do it. I hope you enjoy keeping this diary. I also hope it will be of benefit in the process of learning English for this class. The following are questions that can help you as you write your diary: • What kind of inner speech in English did you experience? • Did you practice English in your mind at some moment? • Did you use English in your mind? When? • When the teacher asked a question in class, did you answer it mentally? • Did you "hear" the sounds of English in your mind? In other words, did you hear something of what you heard in class, on TV, or in some other place? • Did you try to repeat words mentally or in a low voice? • Did you try to translate a word mentally? In other words, did you think of its equivalent in Spanish? • Did you think about the meaning, pronunciation, or spelling of a word or phrase?

222

APPENDIX

• • • • • •

Did you visualize the words in your mind? Did you find yourself imitating the way someone spoke in English? Did you try to remember how something is said in English? Did you review the vocabulary mentally? Did you think how to say something in English? Did you imagine talking to someone in English? When you imagined this conversation, what did you say and what did the other person say?

REFERENCES

Abramson, M., & Goldinger, S. D. (1997). What the reader's eye tells the mind's ear: Silent reading activates inner speech, Perception and Psychophysics, 59, 1059-1068. Aitken, J. E. (1997). A typology of inner-speech patterns. In J. E. Aitken & L. J. Shedletsky (Eds.), Intrapersonal communication processes (pp. 129-137). Annandale, VA: National Communication Association. Aitken, J. E., & Shedletsky, L. J. (Eds.). (1997). Intrapersonal communication processes. Annandale, VA: National Communication Association. Akhutina, T. V. (1978). The role of inner speech in the construction of an utterance. Soviet Psychology, 16(3), 3-30. Allen, T., David, N., & Kung, R. (1997). A socio-cultural approach to information processing: Exploring the combined roles of thought, language, and context. In J. E. Aitken & L. J. Shedletsky (Eds.), Intrapersonal communication processes (pp. 105-116). Annandale, VA: National Communication Association. Allison, D. (1998). Investigating learners' course diaries as explorations of language. Language Teaching Research, 2, 24-47. Allwright, D., & Bailey, K. M. (1991) Focus on the language classroom. New York: Cambridge University Press. Anton, M., & DiCamilla, F. (1998). Socio-cognitive functions of LI collaborative interaction in the L2 classroom. The Canadian Modern Language Review, 54,314342. Appel, G., & Lantolf, J. P. (1994). Speaking as mediation: A study of LI and L2 text recall tasks. The Modern Language Journal, 78, 437-452. Apple, C. G. (1989). Freedom of choice: Intrapersonal communication and emotion. In C. V. Roberts & K. W. Watson (Eds.), Intrapersonal communication processes (pp. 319-336). New Orleans, LA: Spectra. Baars, B. J. (1997). In the theater of consciousness. New York: Oxford University Press. Baciu, M. V., Rubin, C , Decorps, M. A., & Segebarth, C. M. (1999). fMRI assessment of hemispheric language dominance using a simple inner speech paradigm. NMR Biomedicine, 12, 293 -298. Baddeley, A. D. (1986). Working memory. Oxford: Oxford University Press. Baddeley, A., Gathercole, S., & Papagno, C. (1998) The phonological loop as a language learning device. Psychological Review, 105, 158-173. Baddeley, A., & Lewis, V. (1981). Inner active processes in reading: The inner voice, the inner ear, and the inner eye. In A. M. Lesgold & C. A. Perfetti (Eds.),

224

REFERENCES

Interactive processes in reading (pp. 107-129). Hillsdale, NJ: Erlbaum. Baddeley, A., & Logie. R. (1992). Auditory memory and working memory. In D. Reisberg (Ed.), Auditory images (pp. 179-197). Hillsdale, NJ: Lawrence Erlbaum. Bailey, K., & Oschner, R. (1983). A methodological review of the diary studies: Windmill tilting or social science? In K. M. Bailey, M. H. Long, & S. Peck (Eds.), Second language acquisition studies (pp. 188-198). Rowley, MA: Newbury House. Bakhtin, M. M. (1981). The dialogic imagination. Four essays. (M. Holquist, Ed., and C. Emerson and M. Holquist, Trans.). Austin, TX: University of Texas Press. Barber, E. J. W. (1980). Language acquisition and applied linguistics, ADFL Bulletin, 12, 26-32. Bedford, D. A. (1985). Spontaneous playback of the second language: A descriptive study. Foreign Language Annals, 18,219-2%!. Beggs, W. D. A., & Howarth, P. N. (1985). Inner speech as a learned skill. Journal of Experimental Child Psychology, 39, 396-411. Behnke, R. R. (1989). Issues of measurement, instrumentation, and analysis of physiological variables. In C. V. Roberts & K. W. Watson, Intrapersonal communication processes (pp. 203-216). New Orleans: Spectra. Belz, J. A. (2002). Second language play as a representation of the multicompetent self in foreign language study. Journal of Language, Identity, and Education, 1, 13-39. Benjafield, J. (1969). Evidence that "thinking aloud" constitutes an externalization of inner speech. Psychonomic Science, 75,83-84. Berk, L. E. (1992). Children's private speech: An overview of theory and the status of research. In R. M. Diaz & L. E. Berk (Eds.), Private speech: From social interaction to self-regulation (pp. 17-54). Hillsdale, NJ: Erlbaum. Berk, L. E., & Garvin, R. (1984). Development of private speech among low-income Appalachian children. Developmental Psychology, 20, 271-286. Bogen, J. (2002). Epistemological custard pies from functional brain imaging. Philosophy of Science, 69 (3), 59-71. Bransford, J, D. (1979). Human cognition. Belmont, CA: Wadsworth. Broner, M., & Tarone, E. (2001). Is it fun ? Language play in a fifth-grade Spanish immersion classroom. The Modern Language Journal, 85, 363-379. Brooks, F. P., & Donato, R. (1994). Vygotskyan approaches to understanding foreign language learner discourse during communicative tasks. Hispania, 77, 262-214. Brooks, F. B., Donato, R., & McGlone, J. V. (1997). When are they going to say "it" right? Understanding learner talk during pair-work activity. Foreign Language Annals, 30, 524-541. Bugelski, B. R. (1979). Principles of learning and memory. New York: Praeger Publishers. Campbell, R. (1992). Speech in the head? Rhyme skill, reading, and immediate memory in the deaf. In D. Reisberg (Ed.), Auditory images (pp. 73-93). Hillsdale, NJ: Erlbaum. Carruthers. P. (1996). Language, thought, and consciousness. An essay in philosophical psychology. Cambridge, UK: Cambridge University Press.

REFERENCES

225

Carruthers, P. (1998a). Conscious thinking: Language or elimination. Mind & Language, 13, 457-476. Carruthers, P. (1998b). Thinking in language? Evolution and a modularist possibility. In P. Carruthers & J. Boucher (Eds.), Language and thought. Interdisciplinary themes (94-119). Cambridge, UK: Cambridge University Press. Carruthers, P., & Boucher, J. (Eds.). (1998). Language and thought. Interdisciplinary themes. Cambridge, UK: Cambridge University Press. Centeno-Cortes, B. (2003). Private speech in the second language classroom: Its role in internalization and its link to social production. Unpublished doctoral dissertation, Pennsylvania State University. Centeno-Cortes, B., & Jimenez, A. (2004). Problem-solving tasks in a foreign language: The importance of the LI in private verbal thinking. International Journal of Applied Linguistics, 14, 7-35. Chamot, A. U. (1987). The learning strategies of ESL students. In A.Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 71-83). Englewood Cliffs, NJ: Prentice Hall. Chee, M. W. L., Caplan, D., Soon, C. S., Sriram, N., Tan, E. W. L., Thiel, T., & Weekes, B. (1999). Processing of visually presented sentences in Mandarin and English studied with fMRI. Neuron, 23, 127-137. Chee, M. W. L., Soon, C. S., Lee, H. L. (2003). Common and segregated neuronal networks for different languages revealed using functional magnetic resonance adaptation. Journal of Cognitive Neuroscience, 15, 85-97. Chee, M. W. L., Tan, E. W. L., & Thiel, T. (1999). Mandarin and English single word processing studied with fMRI. Journal of Neuroscience, 19, 3050 -3056. Chen, R. (1988). The private speech of a Chinese-English bilingual child: A naturalistic longitudinal study (Doctoral dissertation, University of Illinois at Urbana-Champaign, 1997). Dissertation Abstracts International-A, 49 (01), 60. Chomsky, N. (1975). Reflections on language. New York: Pantheon Books. Clark, A. (1998). Magic words: How language augments human computation. In P. Carruthers & J. Boucher (Eds.), Language and thought. Interdisciplinary themes (pp. 162-183). Cambridge, UK: Cambridge University Press. Cohen, A. D. (1984). Studying second-language learning strategies. How do we get the information. Applied Linguistics, 5, 107-112. Cohen, A. D. (1987). Using verbal reports in research on language learning. In C. Faerch & G. Kasper (Eds.), Introspection in second language research (pp. 82-95). Cleveland, UK: Multilingual Matters. Cohen, A. D. (1994). The language used to perform cognitive operations during fullimmersion math tasks. Language Testing, 11(2), 1-23. Cohen, A. D. (1998). Strategies in learning and using a second language. New York: Longman. Cohen, A. D., & Olshtain, E. (1993). The production of speech acts by EFL learners. TESOL Quarterly, 27, 33-56. Cole, M. (1990). Cognitive development and formal schooling. The evidence from cross-cultural research. In L. C. Moll (Ed.) Vygotsky and education (pp. 89-110).

226

REFERENCES

New York: Cambridge University Press. Cook, G. (2000). Language play: Language learning. Oxford: Oxford University Press. Cook. V. J. (1992). Evidence formulticompetence. Language Learning, 42,557-591. Cook, V. J. (1998). Internal and external uses of a second language, Essex Research Reports in Linguistics, 100-110. Retrieved May 26,2003, from http://privatewww. essex.ac.uk/~vcook/OBS3.htm Coughlan, P., & Duff, P. A. (1994). Same task, different activities: Analysis of a SLA task from an activity theory perspective. In J. P. Lantolf & G. Appel (Eds.), Vygotskian approaches to second language research (pp. 173-193). Norwood, NJ: Ablex. Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing. A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11,671-684. Crookes, G. (1989). Planning and interlanguage variation. Studies in Second Language Acquisition, 77,367-383. Cunningham, S. B. (1989). Defining intrapersonal communication. In C. V. Roberts & K. W. Watson (Eds.), Intrapersonal communication processes (pp. 82-94). New Orleans, LA: Spectra. Dance, F. E. X. (1994). Hearing voices. In D. R. Vocate (Ed.), Intrapersonal communication: Different voices, different minds (pp. 195-210). Hillsdale, NJ: Lawrence Erlbaum. Dance, F. E. X., & Larson, C. E. (1976). The functions of human communication. A theoretical approach. New York: Holt, Rinehart, & Winston, de Courcy, M. (1995, March). Language learning experiences ofAustralian French immersion students. Paper presented at the Annual Meeting of the American Association for Applied Linguistics, Long Beach, CA. (ERIC Document Reproduction Service No. ED 388 096) Dennett, D. C. (1991). Consciousness explained. New York: Little Brown & Co. Dennett, D. C. (1998). Reflections on language and mind. In P. Carruthers & J. Boucher (Eds.), Language and thought. Interdisciplinary themes (pp. 284-294). Cambridge, UK: Cambridge University Press. De Nil, L. F., Kroll, R. M., & Kapur. S. (2000). A positron emission tomography study of silent and oral single word reading in stuttering and nonstuttering adults. Journal of Speech, Language, and Hearing Research, 43, 1038-1053. Diaz, R. M. (1992). Methodological concerns in the study of private speech. In R. M. Diaz & L. E. Berk (Eds.), Private speech: From social interaction to selfregulation (pp. 55-81). Hillsdale, NJ: Erlbaum. Diaz, R. M., & Berk, L. E. (1992). Introduction. In R. M. Diaz & L. E. Berk (Eds.), Private speech: From social interaction to self regulation (pp. 1-13). Hillsdale, NJ: Erlbaum. Diaz, R. M., Neal, C. J., & Amaya-Williams, M. 1990. The social origins of selfregulation. In L.C.Moll (Ed.) Vygotsky and education (pp. 127-154). New York: Cambridge University Press. DiCamilla, F., & Anton, M. (2004). Private speech: A study of language for thought

REFERENCES

227

in the collaborative interaction of language learners. International Journal of Applied Linguistics, 14,36-69. DiCamilla, F., & Lantolf, J. P. (1995). The linguistic analysis of private writing. Language Sciences, 16, 347-369. Ellis, N. C , & Beaton, A. (1993). Factors affecting the learning of foreign language vocabulary: Imagery keyword mediators and phonological short-term memory. The Quarterly Journal of Experimental Psychology, 46A, 533-558. Ellis, N. C , & Sinclair, S. G. (1996). Working memory in the acquisition of vocabulary and syntax. The Quarterly Journal of Experimental Psychology, 49A, 234-250. Emerson, C. (1996). The outer word and inner speech. Bakhtin, Vygotsky, and the internalization of language. In H. Daniels (Ed.), An introduction to Vygotsky (pp. 123-142). New York: Routledge. Ericsson, K. A., & Simon, H. A. (1980). Verbal reports as data. Psychological Review, 57,215-251. Ericsson, K. A., & Simon, H. A. (1984). Protocol analysis. Verbal reports as data. Cambridge, MA: MIT Press. Ericsson, K. A., & Simon, H. A. (1987). Verbal reports on thinking. InC. Faerch& G. Kasper (Eds.), Introspection in second language research (pp. 24-53). Cleveland, U.K.: Multilingual Matters. Ericsson, K. A., & Simon, H. A. (1998). How to study thinking in everyday life: Contrasting think-aloud protocols with descriptions and explanations of thinking. Mind, Culture, and Activity, 5, 178-186. Feldmann, U., & Stemmer, B. (1987). Thin aloud a retrospective da in C-te taking: diffe languages - diff learners - sa approaches? In C. Faerch & G. Kasper (Eds.), Introspection in second language research (pp. 251-267). Cleveland, UK: Multilingual Matters. Fiez, J. A, Raichle, M. E., Balota, D. A., Tallal, P., & Petersen, S. E. (1996). PET activation of posterior temporal regions during auditory word presentation and verb generation. Cerebral Cortex, 6, 1-10. Flavell, J. H. (1977). Cognitive development. Englewood Cliffs, NJ: Prentice Hall. Flavell, J. H., Green, F. L., Flavell, E. R, & Grossman, J. B. (1997). The development of children's knowledge about inner speech. Child Development, 68, 39-47. Fodor, J. (1975). The language of thought. New York: Thomas Y. Crowell. Fodor, J. (1983). The modularity of mind. Cambridge, MA: MIT Press. Foster, P., & Skehan, P. (1996). The influence of planning and task type on second language performance. Studies in Second Language Acquisition, 75,299-323. Frahm, J., Fransson, P., & Kruger, G. (1999). Magnetic resonance imaging of human brain function. Retrieved March 11, 2003, from http://www.mpibpc.gwdg.de/ abteilungen/NMR/reviews/fmri/neurosci.html Frawley, W. (1989). [Review of the book Thought and Language]. Studies in Second L anguage A cquis it ion, 11, 331-332. Frawley, W. (1997). Vygotsky and cognitive science. Language and the unification ofthe social and computational mind. Cambridge, MA: Harvard University Press.

228

REFERENCES

Frawley, W., & Lantolf, J. P. (1985). Second language discourse: A Vygotskyan perspective. Applied Linguistics, 6, 19-44. Frawley, W., & Lantolf, J. P. (1986). Private speech and self-regulation: A commentary on Frauenglass and Diaz. Developmental Psychology, 22, 706-708. Friedman, L., Kenny, J. T., Wise, A. L., Wu, D., Stuve, T. A., Miller, D. A., Jesberger, J. A., & Lewin, J. S. (1998). Brain activation during silent word generation evaluated with functional MRI. Brain and Language, 64, 231-256. Fry, P. S. (1992). Assessment of private and inner speech of older adults in relation to depression. In R. M. Diaz & L. E. Berk (Eds.), Private speech: From social interaction to self-regulation (pp. 267-284). Hillsdale, NJ: Erlbaum. Gallimore, R., & Tharp, R. (1990). Teaching mind in society: Teaching, schooling, and literate discourse. In L. C. Moll (Ed.), Vygotsky and education. Instructional implications and applications of sociohistoricalpsychology (pp. 175-205). New York: Cambridge University Press. Galperin, P. Ya. (1967). On the notion of internalization. Soviet Psychology, 5(3), 28-33. Gass, S. M., & Mackey, A. (2000). Stimulated recall methodology in second language research. Mahwah, NJ: Erlbaum. Gathercole, S. E., & Baddeley, A. D. (1993). Working memory and language. Hillsdale, NJ: Erlbaum. Gathercole, S. E., & Thorn, A. S. C. (1998). Phonological short-term memory and foreign language learning. In A. F. Healy & L. E. Bourne Jr. (Eds.), Foreign language learning. Psycholinguistic studies on training and retention (pp. 141158). Mahwah, NJ: Erlbaum. Gillet, P, Espagniet, L, Billard, C. (1997). The role of the "phonological loop" in verbal cognition [Original Title: Le Role de la "boucle phonologique" dans la cognition verbale], A.N.A.E., 9, J(43), 107-111. Gillette, B. (1987). Two successful language learners: An introspective approach. In C. Faerch & Kasper (Eds.), Introspection in second language research (pp. 268279). Philadelphia: Multilingual Matters. Gillette, B. (1994). The role of learner goals in L2 success. In J. P. Lantolf & G. Appel (Eds.), Vygotskian approaches to second language research (pp. 195-213). Norwood, NJ: Ablex. Girbau, D. (1996). Private and social speech in communication: Terminology and distinctive traits. Journal of Psycholinguistic Research, 25, 507-513. Glaessner, B. E. (1995). Bilingual children in school: The negotiation of language use in a first-grade classroom. Dissertation Abstracts International, 55(12-B), 5582. Goudena, P. P. (1992). The problem of abbreviation and internalization of private speech. In R. M. Diaz & L. E. Berk (Eds.), Private speech: From social interaction to self-regulation (pp. 215-224). Hillsdale, NJ: Erlbaum. Grabois, H. 1996. Word association methodology in a cross-linguistic study of lexicon. Cornell Working Papers in Linguistics, 14,74-100. Guerrero, M. C. M. de. (1987). The din phenomenon: Mental rehearsal in the second language. Foreign Language Annals, 20, 537-548.

REFERENCES

229

Guerrero, M. C. M. de. (1991). Nature of inner speech in mental rehearsal of the second language (Doctoral Dissertation, Inter American University of Puerto Rico, 1990). Dissertation Abstracts International, 51, 4044A. Guerrero, M. C. M. de. (1994). Form and functions of inner speech in adult second language learning. In J. P. Lantolf & G. Appel (Eds.), Vygotskian approaches to second language research (pp. 83-115). Norwood, NJ: Ablex. Guerrero, M. C. M. de. (1999). Inner speech as mental rehearsal: The case of advanced L2 learners. Issues in Applied Linguistics, 10, 27-55. Guerrero, M. C. M. de. (2000). Activating the ZPD: Mutual scaffolding in L2 peer revision. The Modern Language Journal, 84, 51-68. Guerrero, M. C. M. de. (2004). Early stages of L2 inner speech development: What verbal reports suggest. International Journal of Applied Linguistics, 14, 90-112. Guerrero, M. C. M. de, & Villamil, O. (2000). Activating the ZPD: Mutual scaffolding in L2 peer revision. The Modern Language Journal, 84, 51-68. Gutierrez, X. (2000). El Llenguatge Interior dels Aprenents Adults d'Angles com a Llengua Estrangera: Caracteristiques Lingiiistiques i Funcions. Unpublished master's thesis. Bellaterra: Universitat Autonoma de Barcelona. Hakuta, K. (1986). Mirror of language. The debate on bilingualism. New York: Basic Books. Hamilton, M. A. (1997). Inner voices as alter-ego mental constructs: Auditory verbal hallucinations. In J. E. Aitken & L. J. Shedletsky (Eds.), Intrapersonal Communication Processes (pp. 151-158). Annandale, VA: National Communication Association. Hardyck, L. V., & Petrinovich, L. R. (1970). Sub-vocal speech and comprehension level as a function of difficulty level of reading material. Journal of Verbal Learning and Verbal Behavior, 9, 647-652. Harre, R. (1987). The social construction of selves. In K. Yardley & T. Honess (Eds.), Self and identity: Psychosocial perspectives (pp. 41-52). New York: John Wiley & Sons. Harre, R., & Gillett, G. (1994). The discursive mind. London: Sage. Hellmich, H., & Esser, U. (1975). Innere Sprache, inneres Sprechen und ihre wirksamkeit im Fremdsprachenunterricht [Inner speech, inner language, and their effect on foreign language teaching]. Deutsch als Fremdsprache, 12, 128-142. Hikins, J. W. (1989). Intrapersonal discourse and its relationship to human communication: Rhetorical dimensions of self-talk. In C. V. Roberts & K. W. Watson (Eds.), Intrapersonal communication processes (pp. 28-62). New Orleans, LA: Spectra. Hoffman, E. (1989). Lost in translation. A life in a new language. New York: Penguin Books. Holquist, M. (Ed.). (1981). The dialogic imagination. Four essays by M. M. Bakhtin. Austin, TX: University of Texas Press. Honeycutt, J. M., Zagacki, K. S., & Edwards, R. (1989). Intrapersonal communication, social cognition, and imagined interactions. In C. V. Roberts & K. W. Watson (Eds.), Intrapersonal communication processes (pp. 166-184). New

230

REFERENCES

Orleans, LA: Spectra. Hopper, P. J. (1998). Emergent grammar. In M. Tomasello (Ed.), The new psychology of language. Cognitive and functional approaches to language structure (pp. 155-175). Mahwah, NJ: Erlbaum. Houston, J. P. (1986). Fundamentals ojlearning and memory (3rd ed.). New York: Harcourt, Brace, Jovanovich. Hu, C. (2003). Phonological memory, phonological awareness, and foreign language word learning. Language Learning, 53, 429-462. Huh, M. (2002, April). A perspective on LI inner speech in L2 writing ideology. Paper presented at the meeting of the American Association for Applied Linguistics, Salt Lake City, UT. Hurlburt, R. T. (1990). Sampling normal and schizophrenic inner experience. New York: Plenum Press. Hurlburt, R. T. (1997). Randomly sampling thinking in the natural environment. Journal of Consulting & Clinical Psychology, 65, 941-949. Hurlburt, R. T., Koch, M., & Heavey, C. L. (2002). Descriptive experience sampling demonstrates the connection of thinking to externally observable behavior. Cognitive Therapy & Research, 26, 117-134. Jackendoff, R. (1987). Consciousness and the computational mind. Cambridge, MA: MIT Press. Jackendoff, R. (1996). How language helps us think. Pragmatics & Cognition, 4, 134. James, W.( 1950). The principles of psychology by William James. Vol.1. New York: Dover Publications. Jensen, M. D, (1989). Introspective writings as reflections of intrapersonal communication. In C. V. Roberts &K. W. Watson, Intrapersonal communication processes (pp. 111-134). New Orleans: Spectra. John-Steiner, V. (1985a). Notebooks of the mind. Explorations of thinking. Albuquerque, NM: University of New Mexico Press. John-Steiner, V. (1985b). The road to competence in an alien land. In J. V. Wertsch (Ed.), Culture, communication, and cognition: Vygotskian perspectives (pp. 348371). New York: Cambridge University Press. John-Steiner, V. (1992). Private speech among adults. In R. M. Diaz & L. E. Berk (Eds.), Private speech: From social interaction to self-regulation (pp. 285-296). Hillsdale, NJ: Erlbaum. John-Steiner, V., Panofsky, C. P., & Smith, L. W. (1994). Introduction. In V. JohnSteiner, C. P. Panofsky, & L. W. Smith (Eds.), Sociocultural approaches to language and literacy. An interactionist perspective (pp. 1-33). New York: Cambridge University Press. Johnson, J. R. (1984). The role of inner speech in human communication. Communication Education, 33, 211-222. Johnson, M. (2004). A philosophy of second language acquisition. New Haven: Yale University Press. Joravsky, D. (1989). Russian psychology: A critical history. Oxford, UK: Blackwell.

REFERENCES

231

Joseph, J., Noble, K., & Eden, G. (2001). The neurobiological basis of reading. Journal of Learning Disabilities, 34, 566-579. Kaplan, A. (1993). French lessons. Chicago: The University of Chicago Press. Kecskes, I., & Papp, T. (2000). Foreign language and mother tongue. London: Erlbaum. Keller, C. M., & Keller, J. D. (1996). Imaging in iron, or thought is not inner speech. In J. J. Gumperz & S. C. Levinson (Eds.), Rethinking linguistic relativity (pp. 115129). New York: Cambridge University Press. Kim, K. H. S., Relkin, N. R., Lee, K., & Hirsch, J. (1997). Distinct cortical areas associated with native and second languages, Nature, 388, 171-174. Klein, D., Milner, B., Zatorre, R. J., Meyer, E., Evans, A. C. (1995). The neural substrates underlying word generation: A bilingual functional-imaging study. Proceedings of the National Academy of Sciences of the United States of America Washington, 92, 2899-2903 Kohlberg, L., Yaeger, J. & Hjertholm, E. (1968). Private speech: Four studies and a review of theories. Child Development, 39, 691-736. Korba, R. J. (1989). The cognitive psychophysiology of inner speech. In C. V. Roberts & K. W. Watson, Intrapersonal communication processes (pp. 217-242). New Orleans: Spectra. Korba, R. J. (1990). The rate of inner speech. Perceptual and Motor Skills, 71,10431052. Kozulin,A. (Ed.). (1986). Thought and language. Cambridge, MA: The MIT Press. Kozulin, A. (1990a). Lapsicologia de Vygotski. Biografia de unas ideas [Vygotksy's psychology. A biography of ideas]. Madrid: Alianza. Kozulin, A. (1990b). [Review of the book Thinking and Speech]. Applied Psycholinguistics, 11, 123-127. Kozulin, A., Gindis, B., Ageyev, V. S., & Miller, S. M. (2003). Sociocultural theory and education: Students, teachers, and knowledge. In A. Kozulin, V. S. Ageyev, S. Miller, & B. Gindis (Eds.), Vygotsky's theory of education in cultural context (pp. 1-11). Cambridge, UK: Cambridge University Press. Kramsch, C. (2000). Social discursive constructions of self in L2 learning. In J. P. Lantolf (Ed..), Sociocultural theory and second language learning (pp. 133-153). Oxford: Oxford University Press. Krashen, S. D. (1983). The din in the head, input, and the language acquisition device. Foreign Language Annals, 16,41-44. Kroll, J., Michael, E., & Sankaranarayanan, A. (1998). A model of bilingual representation and its implications for second language acquisition. In A. F. Healy & L. E. Jr. Bourne (Eds.), Foreign language learning. Psycholinguistic studies on training and retention. Mahwah, NJ: Erlbaum. Kronk, C. M. (1994). Private speech in adolescents. Adolescence, 29, 781-804 Kuczaj, S.A. (1983). Crib speech and language play. New York: Springer-Verlag. Labov, W. (1984). Sociolinguisticpatterns. Philadelphia: University of Philadelphia Press. Lantolf, J. P. (1997). The function of language play in the acquisition of L2 Spanish.

232

REFERENCES

In A. T. Perez-Leroux & W. R. Glass (Eds.), Contemporary perspectives on the acquisition of Spanish. Volume 2: Production, processing, and comprehension (pp. 3-24). Somerville, MA: Cascadilla Press. Lantolf, J. P. (1999). Second culture acquisition. Cognitive considerations. In E. Hinkel (Ed.), Culture in second language teaching and learning (pp. 28-46). Cambridge, UK: Cambridge University Press. Lantolf, J. P. (2000a). Introducing sociocultural theory. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp. 1 -26). New York: Oxford University Press. Lantolf, J. P. (2000b). Second language learning as a mediated process. Language Teaching, 33, 79-96. Lantolf, J. P. (2003). Intrapersonal communication and internalization in the second language classroom. In A. Kozulin, V. S. Ageyev, S. Miller, & B. Gindis (Eds.), Vygotsky 's theory of education in cultural context (pp. 349-370). Cambridge, UK: Cambridge University Press. Lantolf, J. P. (In press). Sociocultural and second language learning research: An exegesis. In E. Hinkel (Ed.), Handbook of Second Language Acquisition. Erlbaum. Lantolf, J. P., & Appel, G. (Eds.) (1994). Vygotskian approaches to second language research. Norwood, NJ: Ablex. Lantolf, J. P., DiCamilla, F. J , & Ahmed, M. (1997). The cognitive function of linguistic performance: Tense/Aspect use by LI and L2 speakers. Language Sciences, 19, 153-165. Lantolf, J. P., & Frawley, W. (1984). Second language performance and Vygotskyan psycholinguistics: Implications for L2 instruction. In A. Manning, P. Martin., & K. McCalla (Eds.), The tenth LACUSForum, 1983 (pp. 425-440). Columbia, SC: Hornbeam Press. Lantolf, J. P., & Frawley, W. (1988). Proficiency. Understanding the construct. Studies in Second Language Acquisition, 10, 181-195. Lantolf, J. P., & Pavlenko, A. (1995). Sociocultural theory and second language acquisition. Annual Review of Applied Linguistics, 15, 108-124. Lantolf, J. P., & Pavlenko, A. (2001). (S)econd (L)anguage (A)ctivity theory: Understanding second language learners as people. In M. P. Breen (Ed.), Learner contributions to language learning. New directions in research (pp. 141-158). Harlow, England: Pearson Education. Lantolf, J. P. & Yanez, M. C. (2003) Talking yourself into Spanish: Private speech and second language learning. Hispania, 86, 97-109. Larsen, S. F., Schrauf, R. W., Fromholt, P., & Rubin, D. C. (2002). Inner speech and bilingual autobiographical memory: A Polish-Danish cross-cultural study. Memory, 10, 45-54. Leontiev, A. A. (1981). Psychology and the language learning process. New York: Pergamon. Leontiev, A. N. (1981). The problem of activity in psychology. In J. V. Wertsch (Ed.), The concept of activity in Soviet psychology (pp. 37-71). Armonk, NY:

REFERENCES

233

Sharpe. Leopold, W. F. (1949). Speech development of a bilingual child. A linguist's record. Vol. 4, Diary from age 2. New York: Northwestern University Press. Lieberman, D. A. (1979). Behaviorism and the mind: A (limited) call for a return to introspection. American Psychologist, 54,319-333. Liva, A., Fijalkow, E., & Fijalkow, J. (1994). European Journal of Psychology of Education, 9, 321-330. Livesay, J., Liebke, A., Samaras, M., & Stanley, A. (1996). Covert speech behavior during a silent language recitation task, Perceptual and Motor Skills, 83, 1355-1362 Luria, A. R. (1973). The working brain. An introduction to neuropsychology (B. High, Trans.). New York: Basic Books. Luria, A. R. (1976). Basic problems of neurolinguistics. The Hague: Mouton. Luria, A. R. (1981). Language and cognition (J. V. Wertsch, Ed.). New York: John Wiley & Sons. Lyons, W. (1986). The disappearance of introspection. Cambridge, MA: MIT Press. MacKay, D. G. 1992. Constraints on theories of inner speech. In D. Reisberg (Ed.), Auditory images (pp. 121-149). Hillsdale, NJ: Erlbaum. MacWhinney, B. (1985). Hungarian language acquisition. In D. Slobin (Ed.), The crosslinguistic study of language acquisition (pp. 1069-1155). Hillsdale, NJ: Erlbaum. Markus, H. & Nurius, P. (1987). Possible selves: The interface between motivation and the self-concept. In K. Yardley & T. Honess (Eds.), Self and identity: Psychosocialperspectives (pp. 157-172). New York: John Wiley & Sons. Mathias, R. (1996). The basics of brain imaging. Nida Notes, 11(5). Retrieved December 12,2002, from http://www.drugabuse.gov/NIDA_Notes/NNVoll 1N5/ Basics.html Mazziotta, J. C. (1994). Mapping human brain activity in vivo. (Neurology: From Basics to Bedside) Western Journal of Medicine, 161, 273-278. McCafferty, S. G. (1992). The use of private speech by adult second language learners: A cross-cultural study. The Modern Language Journal, 76, 179-189. McCafferty, S. G. (1994a). Adult second language learners' use of private speech: A review of studies. The Modern Language Journal, 78, 421-436. McCafferty, S. G. (1994b). The use of private speech by adult ESL learners at different levels of proficiency. In J. P. Lantolf & G. Appel (Eds.), Vygotskian approaches to second language research (pp. 117-134). Norwood, NJ: Ablex. McCafferty, S. G. (1998). Nonverbal expression and L2 private speech. Applied Linguistics, 19, 73-96. McCafferty, S. G., & Ahmed, M. K. (2000). The appropriation of gestures of the abstract by L2 learners. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp. 199-218). New York: Oxford University Press. McCrone, J. (1994, January 29). Inner voice, distant memories. New Scientist, 141, 28-31. McCrone, J. (1995). Vygotsky, inner speech and the bifold approach to the human mind. An easy introduction to Vygotskian psychology. Retrieved December 12,

234

REFERENCES

2002, fromhttp://www.btinternet.com/~neuronat/webtwo_features_innerspeech. htm McCrone, J. (1999). Going inside. A tour round a single moment of consciousness. New York: Fromm International. McDermott, K. B, Ojemann, J. G, Petersen, S. E., Ollinger, J. M., Snyder, A. Z., Akbudak, E., Conturo, T.E., Raichle, M.E. (1999). Direct comparison of episodic encoding and retrieval of words: An event-related fMRI study, Memory, 7, 661678. McDonough, J. & McDonough, S. (1997). Research methods for English language teachers. New York: Arnold. McGuigan, F. J. (1970). Covert oral behavior during silent performance of language tasks. Psychological Bulletin, 74,309-326. McGuigan, F. J. (1978). Cognitive psychophysiology. Principles of covert behavior. Englewood Cliffs, NJ: Prentice-Hall. McGuire, P. K., Silbersweig, D. A., Murray, R. M., David, A. S., Frackowiak, R. S. J., Frith, C. D. (1996). Functional anatomy of inner speech and auditory verbal image. Psychological Medicine. 26,29-40. McNeill, D. (1987). Psycholinguistics. New York: Harper & Row. Menhert, U. (1998). The effects of different lengths of time for planning on second language performance. Studies in Second Language Acquisition, 20, 83-108. Moffet, J. (1985). Liberating inner speech. College Composition and Communication, 36, 304-308. Mori, K. (1997). Polite lies. On being a woman caught between cultures. New York: Henry Holt. Morin, A. (1993). Self-talk and self-awareness: On the nature of the relation. The Journal of Mind and Behavior, 4, 223-234. Morin, A. (1999). On a relation between inner speech and self awareness: Additional evidence from brain studies. Retrieved November 14, 2002, from http://cogprints.ecs.soton.ac.uk/archive/00002557 Morin, A., & Everett, J. (1990). Inner speech as mediator of self-awareness, selfconsciousness, and self-knowledge: An hypothesis. New Ideas in Psychology, 8, 337-356. Murphey, T. (1990). The song stuck in my head phenomenon: A melodic din in the LAD? System, 18, 53-64. Murphey, T. (1995). Conversational shadowing for rapport and interactional language acquisition. In M. Ahmed, T. Fujimura, Y. Kato, & M. Leong (Eds.), Proceedings of the 6th Conference on Second Language Research in Japan (pp. 42-65). Niigata: International University of Japan. Murphey, T. (2000). Shadowing and summarizing [NFLRC video #11]. Honolulu: University of Hawai'i, Second Language Teaching & Curriculum Center. Murphey, T. (2001). Exploring conversational shadowing. Language Teaching Research, 5, 128-155. Newman, D., Griffin, P., Cole, M. (1989). The construction zone: Working for cognitive change in school. New York: Cambridge University Press.

REFERENCES

235

Newman, F., & Holzman, L. (1993). Lev Vygotsky: Revolutionary scientist. New York: Routledge. Nida, E. (1957). Learning a foreign language: A handbook prepared especially for missionaries. New York: Friendship Press, Council of the Churches of Christ. Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231-259. Novak, J. (1994). My typewriter made me do it. In M. Robinson (Ed.), Altogether elsewhere. Writers in exile (pp. 262-266). New York: Harcourt, Brace & Co. Ohta, A. S. 2001. Second language acquisition processes in the classroom. Learning Japanese. Mahwah, NJ: Erlbaum. Omaggio-Hadley, A. C. (2001). Teaching language in context. Proficiency-Oriented instruction (3rd ed.). Boston: Heinle & Heinle. O'Malley, J. M., Chamot, A. U., Stewner-Manzanares, G., Russo, R., & Kiipper, L. (1985). Learning strategy applications with students of English as a second language. TESOL Quarterly, 19, 557-584. Orellana, M. F. (1994). Appropriating the voice of the superheroes: Three preschoolers'bilingual uses in play. Early Childhood Research Quarterly, 9,171193. Ortega, L. (1999). Planning and focus on form in L2 oral performance. Studies in Second Language Acquisition, 21, 109-148. Otte, L. M. (2001). The relationship between children's early literacy skills and awareness of inner speech. Dissertation Abstracts International, 61, 5030-B. Papagno, C , Valentine, T., & Baddeley, A. (1991). Phonological short-term memory and foreign-language vocabulary learning. Journal ofMemory and Language, 30, 331-334. Papagno, C , & Vallar, G. (1995). Verbal short-term memory and vocabulary learning in polyglots. Quarterly Journal of Experimental Psychology, 38A, 98-107. Parr, P. C , & Krashen, S. (1986). Involuntary rehearsal of second languages in beginning and advanced performers. System, 14, 275-278. Patrick, E., & Abravanel, E. (2000). The self-regulatory nature of preschool children's private speech in a naturalistic setting. AppliedPsycholinguistics, 21, 45-61. Paulesu, E., Frith, C. D., & Frackowiak, R. S. J. (1993). The neural correlates of the verbal component of working memory. Nature, 362, 342-345 Pavlenko, A. (1996). Bilingualism and cognition: Concepts in the mental lexicon. Cornell Working Papers in Linguistics, 14, 39-73. Pavlenko, A. (1998). Second language learning by adults: Testimonies of bilingual writers. Issues in Applied Linguistics, 9, 3-19. Pavlenko, A. (1999). New approaches to concepts in bilingual memory. Bilingualism: Language and Cognition, 2, 209-230. Pavlenko, A., & Lantolf, J. P. (2000). Second language learning as participation and the (re)construction of selves. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp. 155-177). New York: Oxford University Press. Pellegrini, A. D., & DeStefano, J. S. (1979). The functions of private speech in

236

REFERENCES

preschool children. International Journal ofPsycholinguistics, 6-3, 21-42 Perani, D., Paulesu, E., Galles, N. S., Dupoux, E., Dehaene, S.,Bettinardi, V., Cappa, S., Fazio, F., & Mehler, J. (1998). The bilingual brain-proficiency and age of acquisition of the second language. Brain, 121, 1841-1852. Plato. (1952). The dialogues of Plato. In R. M. Hutchins (Ed.), Great books of the western world. Vol. 7 (B. Jowett, Trans.) (pp. 1-799). Chicago: Encyclopaedia Britannica. Pomper, M. 1990. The hidden meaning of inner speech. (ERIC Document Reproduction Service No. ED 324 737) Posner, M. I., & Raichle, M. E. (1994). Images of mind. New York: Scientific American Library. Pressley, M., & Afflerbach, P. (1995). Verbal protocols of reading. The nature of constructively responsive reading. Mahwah, NJ: Erlbaum. Price, C. J., Green, D. W., & von Studnitz, R. (1999). Brain, 122, 2221-2235. Puech, C. (2001). Langage interieur et ontologie linguistique a la fin du XIXe siecle [Inner language and linguistic ontology at the end of the nineteenth century], Langue Francaise, 132, 26-47. Raichle, M. E. (1994). Images of the mind: Studies with modern imaging techniques. Annual Review of Psychology, 45, 333-356. Raichle, M. E. (1997). Functional brain imaging and verbal behavior. In J. W. Donahoe, & V. P. Dorsel (Eds.), Neural-network models of cognition: Biobehavioral foundations. Advances in psychology, Vol. 121 (pp. 438-454). Amsterdam: North-Holland/Elsevier Science. Reiss, M. (1985). The "good" language learner: Another look. The Canadian Modern Language Review, 41, 511-523. Richmond, V. P., & McCroskey, J. C. (1989). Willingness to communicate and dysfunctional communicative processes. In C. V. Roberts & K. W. Watson (Eds.), Intrapersonalcommunicationprocesses (pp. 292-318). New Orleans, LA: Spectra. Rivers, W. (1979). Learning a sixth language: An adult learner's daily diary. The Canadian Modern Language Review, 36, 67-82. Roberts, C. V., & Watson, K. W. (Eds.). (1989). Intrapersonal communication processes. New Orleans, LA: Spectra. Rodriguez-Fornells, A., Rotte, M., Heinze, H., Nosselt, T., & Munte, T. F. (2002). Brain potential and functional MRI evidence for how to handle two languages with one brain. Nature, 415, 1026-1029. Roebuck, R. (1998). Reading and recall in LI and L2. A sociocultural approach. Stamford, CO: Ablex. Roebuck, R. (2000). Subjects speak out: How learners position themselves in a psycholinguistic task. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp. 79-95). New York: Oxford University Press. Roebuck, R. F. & Wagner, L. C. (2004). Teaching repetition as a communicative and cognitive tool: Evidence from a Spanish conversation class. Inter nationalJour nal of Applied Linguistics, 14, 70-89. Rohrer, J. (1987). Inner speech. Implications for foreign language learning. Die

REFERENCES

237

Nueren Sprachen, 86, 92-101. Rohrkemper, M. (1986). The functions of inner speech in elementary school students' problem-solving behavior. American Educational Research Journal, 23,303-313. Rubin, J. (1987). Learner strategies: Theoretical assumptions, research history and typology. In A. Wenden & J. Rubin (Eds.) Learner strategies in language learning (pp. 15-30). Englewood Cliffs, NJ: Prentice Hall. Ryding, E., Bradvik,B., & Ingvar, D. H. (1996). Silent speech activates prefrontal cortical regions asymmetrically, as well as speech-related areas in the dominant hemisphere. Brain and Language, 52,435-451. Sachs, J. S. (1967). Recognition memory for syntactic and semantic aspects of connected discourse. Perception andPsychophysics, 2, 437-442. Saville-Troike, M. (1988). Private speech: Evidence for second language learning strategies during the "silent" period. Journal of Child Language, 15, 567-590. Schunk, D. H., & Rice, J. M. (1984). Strategy self-verbalization during remedial listening comprehension instruction. Journal of Experimental Education, 53, 4954. Schwartz, J. (1980). The negotiation for meaning: Repair in conversations between second language learners of English. In D. Larsen-Freeman (Ed.), Discourse analysis: Second language research (pp. 138-153). Rowley, MA: Newbury House. Schweers, W., & Hudders, M. (1993). Towards a reconceptualization of the teaching of English in Puerto Rico. San Juan, Puerto Rico: Biblioteca del Pensamiento Critico. Service, E. (1992). Phonology, working memory, and foreign-language memory. The Quarterly Journal of Experimental Psychology, 45A, 21-47. Service, E., & Kohonen, V. (1995). Is the relation between phonological memory and foreign language learning accounted for by vocabulary acquisition? Applied Psycholinguistics, 16, 155-172. Shergill, S. S., Bullmore, E. T, Brammer, M. J, Williams, S. C. R, Murray, R. M, & McGuire, P. K. (2001). A functional study of auditory verbal imagery. Psychological Medicine, 31, 241 -25 3. Siegrist, M. (1995). Inner speech as a cognitive process mediating self-consciousness and inhibiting self-deception. Psychological Reports, 76,259-265. Slobin, D. (1979). Psycholinguistics (2nd ed.). Glenview, IL: Scott, Foresman & Co. Smagorinsky, P. (1998). Thinking and speech and protocol analysis. Mind, Language, and Activity, 5, 157-177. Smith, F. (1983). Essays into literacy. London: Heinemann Educational Books. Smith, J. D., Reisberg, D., & Wilson, M. (1992). Subvocalization and auditory imagery: Interactions between the inner ear and inner voice. In D. Reisberg (Ed.), Auditory images (pp. 95-119). Hillsdale, NJ: Erlbaum. Sokolov, A. N. (1972). Inner speech and thought. New York: Plenum. Stacks, D. W., & Sellers, D. E. (1994). Research: Expanding our knowledge of intrapersonal communication processes. In D. R. Vocate (Ed.), Intrapersonal communication: Different voices, different minds (pp. 101-119). Hillsdale, NJ: Lawrence Erlbaum.

238

REFERENCES

Stern, H. H. (1980). What can we learn from the good language learner? In K. Croft (Ed.), Readings on English as a second language (pp. 54-71). Boston: Little, Brown and Co. Streff, C. R. (1984). The concept of inner speech and its implications for an integrated language arts curriculum. Communication Education, 33, 223-230. Swain, M. (1995). Three functions of output in second language learning. In G. Cook & B. Seidlhofer (Eds.), Principle andpractice in applied linguistics (pp. 125-144). Oxford: Oxford University Press. Swain, M. (2000). The output hypothesis and beyond: Mediating acquisition through collaborative dialogue. In J. P. Lantolf (Ed.), Sociocultural theory and second language learning (pp. 97-114). New York: Oxford University Press. Swain, M., & Lapkin, S. (1998). Interaction and second language learning: Two adolescent French immersion students working together. The Modern Language Journal 82, 320-337. Tarone, E. (1983). Some thoughts on the notion of "communication strategy." In C. Faerch & G. Kasper (Eds.), Strategies in interlanguage communication (pp. 4-14). New York: Longman. Tomlinson, B. (2000). Talking to yourself. The role of the inner voice in language learning. Applied Language Learning, 11, 123-154. Tomlinson, B. (2001). The inner voice: A critical factor in L2 learning. Journal of the Imagination in Language Learning and Teaching, 6. Retrieved September 9, 2002, from http://www.njcu.edu/cill/journal-index.html Trimbur, J. (1987). Beyond cognition: The voices of inner speech. Rhetoric Review, 5,211-221. Upton, T. A., & Lee-Thompson, L. (2001). The role of the first language in second language reading. Studies in Second Language Acquisition, 23, 469-495. Ushakova, T. (1994). Inner speech and second language acquisition: An experimental-theoretical approach. In J. P. Lantolf & G. Appel (Eds.), Vygotskian approaches to second language research (pp. 135-156). Norwood, NJ: Ablex. van der Veer, R., & Valsiner, J. 1991. Understanding Vygotsky: A quest for synthesis. Cambridge, MA: Blackwell van Lier, L. 1996. Interaction in the language curriculum. Awareness, autonomy & authenticity. New York: Longman. Verspoor, M., & Lowie, W. (2003). Making sense of polysemous words. Language Learning, 53, 547-596. Villamil, O. S., & Guerrero, M. C. M. de. (1996). Peer revision in the L2 classroom: Social-cognitive activities, mediating strategies, and aspects of social behavior. Journal of Second Language Writing, 5, 51-75. Vocate, D. R. (1987). The theory of A. R Luria. Functions ofspoken language in the development of higher mental processes. Hillsdale: NJ: Lawrence Erlbaum. Vocate, D. R. (Ed.). (1994a). Intrapersonal communication: Different voices, different minds. Hillsdale, NJ: Lawrence Erlbaum. Vocate, D. R. (1994b). Self-talk and inner speech: Understanding the uniquely human aspects of intrapersonal communication. In D. R. Vocate (Ed.), Intrapersonal

REFERENCES

239

communication: Different voices, different minds (pp. 3-31). Hillsdale, NJ: Lawrence Erlbaum. Volosinov, V. N. (1973). Marxism and the philosophy of language (L. Matejka and I. R. Titunik, Trans.). New York: Seminar Press. (Original work published 1930) Vygotsky, L. S. (1978). Mind in society. The development of higher psychological processes. Cambridge, MA: Harvard University Press. Vygotsky, L. S. (1986). Thought and language. Cambridge, MA: The MIT Press. Vygotsky, L. S. (1997). Educational psychology (K. Silverman, Trans.) Boca Raton, FL: St. Lucie Press. (Original work published 1926) Vygotsky, L. S. (2000). Consciousness as a problem in the psychology of behavior (N. Schmolze, Trans.). (Original work published 1925). Retrieved October 1,2002, from http://www.marxists.org/archive/vygotsky/works/1925/consciousness.htm Weir, R. H. (1962). Language in the crib. The Hague: Mouton & Co. Wells, G. (1998). Using LI to master L2: A response to Anton and DiCamilla's "Socio-cognitive functions of LI collaborative interaction in the L2 classroom." The Canadian Modern Language Review, 54, 343-353. Wenden, A. (1986). What do learners know about their language learning? A second look at retrospective accounts. Applied Linguistics, 7, 186-201. Wenden, A. (1987). How to be a successful language learner: Insights and prescriptions from L2 learners. In A. Wenden & J. Rubin (Eds.), Learner strategies in language learning (pp. 103-117). Englewood Cliffs, NJ: PrenticeHall. Wertsch, J. V. (1979a). From social interaction to higher psychological processes. A clarification and application of Vygotsky's theory. Human Development, 22,1-22. Wertsch, J. V. (1979b). The regulation of human action and the given-new organization of private speech. In G. Zivin (Ed.), The development of selfregulation through private speech (pp. 79-98). New York: John Wiley & Sons. Wertsch, J. V. (1980). The significance of dialogue in Vygotsky's account of social, egocentric, and inner speech. Contemporary Educational Psychology, 5,150-162. Wertsch, J. V. (Ed.). (1981). Language and cognition. New York: John Wiley & Sons. Wertsch, J. V. (1985a). Introduction. In J. V. Wertsch (Ed.), Culture, communication, and cognition. Vygotskian perspectives (pp. 1-18). New York: Cambridge University Press. Wertsch, J. (1985b). Vygotsky and the social formation of mind. Cambridge, MA: Harvard University Press. Wertsch, J. V. (1991). Voices of the mind. A sociocultural approach to mediated action. Cambridge, MA: Harvard University Press. Wertsch, J. V. (1995). Commentary. Human Development, 38, 127-130. Wertsch, J. V. (1998). Mind as Action. New York: Oxford University Press. Wertsch, J. V. (2000). Vygotsky's two minds on the nature of meaning. In C. D. Lee & P. Smagorinsky (Eds.), Vygotskian perspectives on literacy research: Constructing meaning through collaborative inquiry (pp. 19-30). New York: Cambridge University Press.

240

REFERENCES

Wertsch, J. V., & Stone, A. (1985). The concept of internalization in Vygotsky's account of the genesis of higher mental functions. In J. V. Wertsch (Ed.), Culture, communication, and cognition. Vygotskianperspectives (pp. 162-179). New York: Cambridge University Press. Wigglesworth, G. (1997). An investigation of planning time and proficiency level on oral test discourse. Language Testing, 14, 85-106. Williams, J. D. (1987). Covert linguistic behavior during writing tasks. Written Communication, 4, 310-328. Winsler A., Carlton, M. P., Barry, M. J. (2000). Age-related changes in preschool children's systematic use of private speech in a natural setting. Journal of Child Language,27, 665-687. Woodall, B. R. (2002). Language switching: Using the first language while writing in a second language. Journal of Second Language Writing, 11, 7-28. Yaden, D. B., Jr. (1984). Inner speech, oral language, and reading: Huey and Vygotsky revisited. Reading Psychology, 5, 155-166. Yuan, F., & Ellis, R. (2003). The effects of pre-task planning and on-line planning on fluency, complexity and accuracy in L2 monologic oral production. Applied Linguistics, 24, 1-27. Zivin, G. (1976). Removing common confusions about egocentric speech, private speech, and self-regulation. In G. Zivin (Ed.), The development of self-regulation through private speech (pp. 13-49). New York: John Wiley & Sons.

AUTHOR INDEX

Abramson, M., 54 Abravanel, E., 154 Afflerbach, P., 107 Ageyev, V. S., 9 Ahmed, M., 9, 18, 71-72, 86, 92, 169 Aitken, J. E., 8, 49, 98, 133 Akhutina, T. V., 36, 42-43 Allen, T., 98 Allison, D., 100 Allwright, D., 100, 146 Amaya-Williams, M , 196 Anton, M., 169, 176, 194 Appel, G., 10-11,92, 169, Apple, C. G., 133 Baars, B. J., 5 Baciu,M.V., 114 Baddeley, A. D., 23, 53-54, 56, 77, 87, 120, 125-126, 163, 179, 197,209 Bailey, K. M., 100, 146 Bakhtin, M. M., 11, 31, 37-38, 50, 6970, 125, 185, 198 Balota, D. A., 113 Barber, E. J. W., 78-79 Barry, M. J., 92 Beaton, A., 77 Bedford, D. A., 79-81, 98, 120 Beggs, W. D. A., 8, 15, 195 Behnke, R. R., 111 Belz,J.A., 132, 181, 184 Benjafield, J., 101 Berk, L. E., 92, 94, 154 Billard, C , 54 Bogen, J., 116 Boucher, J., xiii, 5-6 Bradvik, B., 57 Bransford, J. D., 23, 77

Broner, M., 81, 120, 132, 155-156, 158, 184 Brooks, F. P., 92, 169, 194,204 Bugelski, B. R., 23 Campbell, R., 23, 54-55, 159, 195 Caplan, D., 88, 114, 171 Carlton, M. P., 92 Carruthers, P., xiii, 5-6, 19, 27, 51-53, 56 Centeno-Cortes, B., 72,86,92,156-157, 159, 176-177, 193-194, 197, 203204 Chamot, A. U., 77, 120 Chee, M. W. L., 83-84, 88, 114, 171172 Chen,R., 155 Chomsky, 5, 48 Clark, A., xiii, 5,23,27, 51-52, 55, 120, 124 Cohen, A., 8, 20, 62-64, 85-86, 96, 9899,100-101,164-165,170,177-178, 181, 188, 194,204 Cole,M., 10, 11 Cook, G., 81 Cook, V. J., xii, 64-65, 85,98,133,177179, 181-182, 194 Coughlan, P., 14 Craik, F. I. M., 22-23, 77, 205 Crookes, G., 203 Cunningham, 21, 131 Dance,F.E.X.,7,21,49, 117 David, N., 98 de Courcy, M , 155-156, 163 Decorps, M. A., 114 Dennett, D. C , 5-6, 23 DeNil,L.F., 113

242

AUTHOR INDEX

DeStefano, J. S., 154 Diaz, R. M., 93-94, 196 DiCamilla, F., 92-93,169,176,194,210 Donato, R., 92, 169,204 Duff, P. A., 14 Eden,G., 112 Edwards, R., 23 Ellis, N. C , 77, 126, 179, 197, 209 Ellis, R., 203 Emerson, C , 38 Ericsson, K. A., 95-96, 101-104 Espagniet, L, 54 Esser, U., 8 Evans, A. C , 84 Everett, J., 47 Feldmann, U., 96 Fiez,J.A., 113 Fijalkow, E., 8 Fijalkow, J., 8 Flavell, E. R., 99 Flavell, J. H., 24, 99,119,138,159,195, 209-210 Fodor, J., 5, 20, 26, 52-53 Foster, P., 203 Frackowiak, R. S. J., 56 Frahm,J., 112 Fransson, P., 112 Frawley, W., xiii, 5-6, 8, 11, 14, 20, 24, 26-27, 33, 51-52, 55-56, 72-73, 92, 104, 107, 147, 153, 159, 169, 176, 183-184 Friedman, L., 114 Frith, C. D., 56 Fromholt, P., 9 Fry, P. S., 92 Gallimore, R., 195 Galperin, P. Ya., 4,11,14,160,169,199 Garvin, R., 92 Gass, S., 99-100, 139 Gathercole, S., 23, 54, 77, 119, 125-126 Gillet, P., 54 Gillett,G., 185, 187 Gillette, B., 14, 128 Gindis, B., 9

Girbau, D., 24 Glaessner, B. E., 169 Goldinger, S. D., 54 Goudena, P. P., 24, 92, 138, 154 Grabois, H., 66 Green, D. W., 82 Green, F. L., 99 Griffin, P., 11 Grossman, J. B., 99 Guerrero, M. C. M. de, xv, 8,24,73,77, 79-81, 86-87, 98-100,108, 120-123, 126-127, 129, 131-132, 134, 137139, 149, 150-151, 161-167, 170, 175, 178-179, 183-184, 193-194, 197-198,218 Gutierrez, X., 81, 98-99, 138, 149, 183 Hakuta,K., 154 Hamilton, M. A., 99, 133 Hardyck, L. V , 8, 124 Harre,R., 184-185, 187 Heavey, 104 Heinze, H., 83 Hellmich, H., 8 Hikins, J. W., 8 Hirsch, J, 83 Hjertholm, E., 24 Hoffman, E., 69, 186 Holquist, M., 37 Holzman, L., 197 Honeycutt, J. M., 23, 49, 98, 130-131, 133,150 Hopper, P. J., 198 Houston, J. P., 23, 77, 120 Howarth, P. N., 8, 15, 195 Hu, C , 197,209 Hudders, M., 187 Huh, M., 9, 74, 76, 87, 100, 136, 182, 196 Hurlburt, R. T., 53, 97, 103-104, 106 Ingvar, D. H., 57 Jackendoff, R., 5, 23 James, W., 93 Jensen, M. D., 93

AUTHOR INDEX

Jimenez, A., 92,176-177,193-194,203204 John-Steiner, V., 8-9, 13, 19-20, 36, 6768, 70, 72, 85-86, 92-93, 99, 136, 169, 173, 179, 185,202,217 Johnson, J. R., 21,49 Johnson, M., 10 Joravsky, D., 3 Joseph, J., 112 Kaplan, A., 186-187 Kapur, S., 113 Kecskes, I., 136, 170-171, 173, 181 Keller, C M . , 19 Keller, J. D., 19 Kim, K. H. S., 83, 88, 114, 171 Klein, D., 84, 88, 114, 172 Koch,M., 104 Kohlberg, L., 24, 154 Kohonen, V., 77 Korba, R. J., 7,15,21-22, 107, 110-111, 123 Kozulin, A., 2-4, 9-10, 16-17, 184 Kramsch, C , 206 Krashen, S. D., 24, 63, 78-80, 87, 120, 197 Kroll, J., 207 Kroll, R. M., 113 Kronk, C. M., 92, 195 Kruger,G., 112 Kuczaj, S. A., 81 Kung, R., 98 Kupper, L., 77 Labov, W., 105 Lantolf, J. P., 8-14,24-25, 63,66,68-73, 81-82,86-87,92-93,97-98,101,108, 120, 127, 130, 132, 138, 147, 155159, 165, 169, 173, 179, 183-185, 194, 197-198,210,217 Lapkin, S., 204 Larsen, S. F., 9, 64, 85, 98-99, 106, 173, 178-179, 183, 185, 194 Larson, C. E., 7, 49 Lee, H. L., 83 Lee, K., 83

243

Lee-Thompson, L., 9, 74, 76, 87, 100, 102, 136, 182, 196 Leontiev, A. A., xiii, 4, 14, 16, 20, 2728, 34, 42-43, 49, 60-63, 68, 85, 170-174, 184, 192-193, 203, 206, 211,218 Leontiev, A. N., 4, 10-11, 13-14, 38 Leopold, W. F., 154 Lewis, V., 23, 54 Lieberman, D. A., 95, 96 Liebke, A., 111 Liva, A., 8, 133, 195 Livesay, J., I l l Lockhart, R. S., 22-23, 77, 205 Logie, R., 54 Lowie, W., 207 Luria, A. R., xiii, 4, 6-8, 13-14, 16-17, 25, 27-28, 31, 38-42, 47-49, 116, 167, 173 Lyons, W., 95, 97 MacKay, D. G., 55, 124, 160, 162-164 Mackey, A., 99-100, 139 MacWhinney, B., 81 Markus,H, 185 Mathias, R., 112, 114 Mazziotta, J. C , 56, 112 McCafferty, S. G., 8-9, 18, 71-72, 86, 92, 169 McCrone, J., 5, 38 McCroskey, J. C , 133 McDermott, K. B., 114 McDonough, J., 100, 150 McDonough, S., 100, 150 McGlone, J. V , 92 McGuigan, F. J., 7, 22, 108-110 McGuire, P. K., 113 McNeill, D., 18,71, 104 Menhert, U., 203 Meyer, E., 84 Michael, E., 207 Miller, S. M , 9 Milner, B., 84 Moffet, J., 8 Mori, K, 70

244

AUTHOR INDEX

Morin, A., 7, 15,38,47,56 Munte, T. F., 83 Murphey, T., 78, 162, 201-202 Neal, C. J., 196 Newman, D., 11 Newman, F., 197 Nida,E., 19 Nisbett, R. E., 95-97, 105-106 Noble, K., 112 Nosselt, T., 83 Novak, J., 70 Nurius, P., 185 Ohta, A. S., xii, 8, 72-73, 86, 92, 94, 128-129, 138, 147, 156-161, 184, 194, 197-198 Olshtain, E., 62, 100, 181 Omaggio-Hadley, A. C , 120 O'Malley, J. M., 77, 120 Orellana, M. F., 188 Ortega, L., 203 Oschner, R., 100 Otte,L. M., 8, 138 Panofsky, C. P., 9 Papagno, C , 54, 77 Papp,T., 136, 170-171, 173, 181 Parr, P. C , 79, 120 Patrick, E., 154 Paulesu, E , 56, 113 Pavlenko, A., 9-10,13-14,68-72, 86,97, 101, 108, 172-174, 178-179, 183185, 187,205,207-208,217 Pellegrini, A. D., 154 Perani, D., 84, 88, 114, 172 Petersen, S. E., 113 Petrinovich, L. R., 8, 124 Plato, xiii, 1-2, 14, 21 Pomper, M., 92-93 Posner, M. L, 56, 112-113, 115 Pressley, M., 107 Price, C. J., 82, 88, 113 Puech, C, 3 Raichle, M. E.5 56, 112-113, 115 Reisberg, D., 15 Reiss, M., 128

Relkin, N. R., 83 Rice,J. M., 196 Richmond, V. P., 133 Rivers, W., 100, 170-171 Roberts, C. V., 8, 49 Rodriguez-Fornells, A., 83, 88, 114 Roebuck, R, 93, 169, 176, 200, 202 Rohrer, J., 8, 78, 203 Rohrkemper, M., 8, 99 Rotte, M., 83 Rubin, C , 114 Rubin, D. C , 9 Rubin, J., 77, 120 Russo, R., 77 Ryding, E., 57, 113 Sachs, J. S., 106 Samaras, M., I l l Sankaranarayanan, A., 207 Saville-Troike, M., 72-73, 86, 92, 139, 155, 158-161, 164, 184, 197-198 Schrauf, R. W., 9 Schunk,D.H., 196 Schwartz, J., 128 Schweers, W., 187 Segebarth, C. M., 114 Sellers, D. E., 117 Service, E., 77, 127 Shedletsky, L. J., 8, 49 Shergill, S. S., 57, 114-115, 124 Siegrist, M., 47, 98 Simon, H. A., 95-96, 101-104 Sinclair, S. G., 77, 126, 179, 197, 209 Skehan, P., 203 Slobin, D., 106 Smagorinsky, P., 102-103 Smith, F., 23, 78, 98, 120,130-131, 150 Smith, J. D., 15,23,54 Smith, L. W., 9 Sokolov, A. N., xiii-xiv, 2-8, 13, 15-16, 18-19, 27-28, 31, 33, 43-47, 49, 7476,87,101,109-110,112,116,123125, 130, 140-141, 143-144, 147, 150, 156, 161-162, 174, 176, 182, 195, 197,218

AUTHOR INDEX

Stacks, D. W., 117 Stanley, A., 111 Stemmer, B., 96 Stern, H. H , 20 Stewner-Manzanares, G., 77 Stone, A., 168 Streff, C. R., 209 Swain, M., 160, 165,204 Tallal,P., 113 Tan, E. W. L., 84, 88, 172 Tarone, E., 77, 81, 120, 132, 155-156, 158, 184 Tharp,R., 195 Thorn, A. S. C , 77, 125-126 Tomlinson, B., 199-200 Trimbur, J., 8, 125 Upton, T. A., 9, 74, 76, 87, 100, 102, 136, 182, 196 Ushakova, T., 4, 8-9, 65-67, 72, 86,136, 171,207 Valentine, T., 77 Vallar, G., 77 Valsiner, J., 4, 10,90-91 van der Veer, R., 4, 10,90-91 vanLier, L., 148, 161 Verspoor, M., 207 Villamil, O. S., 194 Vocate, D. R., xiii, 2, 8, 15-16, 18, 21, 27-28,41,47-49,128,131,173,180, 185,209 Volosinov, V. N., 37-38, 185 von Studnitz, R., 82 Vygotsky, L. S, xi-xiv, 2-4, 6-19,24-25, 27-38, 40-43, 46-47, 49, 51-53, 55, 57, 59, 61-62, 66, 68-69, 71-73, 7678, 81, 89-91, 94,102-103,105,117119, 123-124, 130, 139, 143, 153154, 157-159, 162-163, 168, 172174, 176, 184-185, 195, 199, 202, 206,208,211,213 Wagner, L. C , 200, 202 Watson, K. W., 8, 49 Weir,R. H., 81, 154 Wells, G., 94

245

Wenden, A., 97, 107, 140, 150 Wertsch, J. V., 7, 9-11, 13-14, 16, 18, 24, 32-34, 36-38, 70,90-91, 94,165, 168, 176, 187, 198 Wigglesworth, G., 203 Williams, J. D., 110 Wilson, M., 15 Wilson, T. D., 95-97, 105-106 Winsler, A., 92 Woodall, B. R., 102, 105, 182, 196 Yaden, D. B. Jr., 8 Yaeger, J., 24 Yanez, M. C , 8,25, 72-73, 86, 92, 127, 132, 138, 156-158, 197 Yuan, F., 203 Zagacki, K. S., 23 Zatorre, R. J., 84 Zivin, G., 24

SUBJECT INDEX

activity inner speech as, 16-17, 26, 38-39, 41,50,215 theory, 4, 10, 13-14,38,42 agency, learner, 14, 70, 86, 158, 184 appropriation, 11-12, 70 of an L2, 130, 148, 165, 185-187, 198,202,214 of L2 gestures, 71, 86 articulation experimental conditions, 74-75 subvocal, 15, 23, 45-46, 50, 53, 5758, 75-76, 87, 109-110, 147, 160-162, 176, 182, 197 attrition, LI, 58, 69-70, 72, 172, 184, 193,205,215 automatization/automaticity of speech processes, 11, 15-16, 44-45, 50, 75,87, 107, 113, 147, 156, 162, 167, 174, 176, 192,215 awareness metaconscious, 20, 51-52, 58 metalinguistic, 63, 86, 128, 159, 164-166, 180,203 of inner speech, 8, 191, 195, 208-211,216,218 self-, 21, 36-38, 47, 50, 56, 58, 107, 128,131,184-185,213 behaviorism/behaviorists, 2, 7, 19, 22, 95,219 bilingual speakers/bilingualism, 8, 58, 64, 67, 69-72, 82-83, 85-86, 99, 101, 108, 114-116, 121, 136, 154-156,160,169,171-175,177, 181, 184-188, 191, 194, 205, 210,216-218

brain imaging, see neuroimaging brain scanning, see neuroimaging cognitive conception of language, 5-6, 19,52-53 communicative conception of language, 5,52 computed tomography (CT), 112 conceptual change, xiv, 9,64,66,70-72, 86, 172-175, 185-187, 193,205208,211,215,217 consciousness as social construct, 9, 12-13, 25, 3738,51 mediated by inner speech, 5, 12, 23, 28,36-38,44,47,50-51,55,5758,98, 119, 131, 185 metaconsciousness, 37-38, 51-55, 58, 105, 185 context/contextualization, 14, 52, 181, 194,206-209,215-216 decontextualization, 66, 172, 175, 189,205 continuous access, 11, 159, 169 covert linguistic behavior, 1, 7, 9, 22, 26, 108-111 cued recall, 89, 97, 99, 106 deafness, and inner speech, 54-55, 116, 195 dialectical materialism, 5 dialogic nature of inner speech, 4, 21, 26, 31, 37, 47-50, 125, 131-132,213 of L2 inner speech, 131-132 din in the head phenomenon, 24, 59, 63, 78-80, 87, 98, 120 eavesdropping, 155, 191, 197-198,211

248

SUBJECT INDEX

echolalia, 176 echolalic responses, 41, 167 egocentric speech, xi, xiv, 4, 11, 16-17, 24,29-34,36,40-41,49,57-58, 72,90-91, 153 electrocardiography (EKG), 109 electroencephalography (EEG), 43,109, 112,114 electromyography (EMG), 43, 45, 87, 89, 109-112,218 electrooculography (EEO), 109 electrophysiology,xiv,83,108-109, 111 emergent grammar, 197-198 experimental-developmental method, xiv, 90 externalization ofanL2, 191,200,203 of inner speech, 8, 72, 92, 94, 105, 210 of thought, xiii, 18, 33, 35-36, 90, 102-103, 168, 170, 181 of thought through L2 inner speech, xv, 147, 153, 168-175, 189 first language (LI) operational definition, xi role played by, 9, 58, 63-68, 72-73, 76, 85-87, 136-138, 148, 151, 163, 170-175, 181-182, 189, 193-196,204-205,211,215 first-person narratives, see narratives, autobiographical foreign language (FL) operational definition, xi functional magnetic resonance imaging (fMRI), 83-84, 88, 114-115 functions of inner speech, 15, 30-31, 36, 44-46, 49, 58 of L2 inner speech, 80-81, 86-87, 119-120, 125-135, 137, 143, 148-149,151,153,177-181,183, 214-215 galvanic skin response (GSR), 109 genetic approach to mind, 10, 13

genetic method, xiv, 9, 89, 91, 93-94, 117-118,218 gestures, xiv, 9, 12, 18, 25, 71, 86, 95, 169 heteroglossia, 125, 137, 164, 198, 202, 213 higher mental functions, xii, 7, 10-14, 17-18, 31, 49, 153, 158, 167, 177, 195,208,213 historical materialism, 3-4 idealism, 2 identity, xiv-xv, 68-70, 86, 101, 108, 153,180-181,184-189,210,215 images/imagery auditory, 3,6,15,23,50,53-58,124, 147, 160, 176 motor, 3, 15 visual/graphic, 3, 44, 46, 150 imagined interactions, 49, 98, 131, 133 imitation, 73, 81, 86,127-128,132,137, 139-140,157-158,162-163,180, 184, 197-198,200,217 immersion, language programs, 63-64, 100,156,163,177,188,207-208 information-processing approach, xi, xiii, 6,23,27, 51, 53-57, 95,126, 161, 165, 197 inner ear, 15, 23, 53-56, 77, 116, 176 inner eye, 54 inner voice, 15, 23, 53-56, 77, 116, 176 internalization, 10-12 of an LI, xi, 154 of an L2, xi, xiv-xvi, 24, 58-59, 7273,84,86,88,92,127-128,139, 146,148,152-168,178,188-189, 191-193,197-200,202-203,206, 208,214,216-217 of social speech, xii-xiii, 11, 15-17, 29,31,33-34,36,40,47,49,51, 89,94,125,138,213,216 intra-observer paradox, 105 intrapersonal communication, xiii, 1, 78,21,25-26,47-49,58 introspection, 3, 9, 36-38, 50, 53, 58,

SUBJECT INDEX

90,93,95-97,99, 101, 104-106, 112, 115, 118, 151, 178, 183, 185, 189,215,218 language learning strategies, xvi, 20,59, 67,77,87, 120, 179 language of thought, 5, 20-21, 26, 100, 182 preference for, 8, 62-65, 72, 85, 98, 177-178 language of/for thought, 1, 20, 26, 51 language play, xii, xiv, 8, 24, 59, 81-82, 86-87,98, 132, 156, 180-181 language switching, 102, 113, 177, 193 literacy, and inner speech, xvi, 25, 55, 159, 191, 195-196,211 lower mental functions, 12, 17-18 macrofunctions, of L2 inner speech cognitive/regulatory function, xv, 153, 177-179, 183, 189,215 rehearsal function, xv, 153,179-181, 183-184, 189,215-216 magnetic resonance imaging (MRI), xiv, 9,56-57, 112, 114,218 magneto-encephalography (MEG), 114 mastery, of psychological functions, 11 mediation, through signs, xii, 10,12-13, 18,31,49,51, 161,213 Mentalese,5,20-21,26,53,62 microgenesis , 13,33-34, 36, 50,90,94, 117-118,153,158,175,215-216 modularist approach to language, xiii, 56,51-53 multilingual speakers/multilingualism, 62-63, 82, 84, 116, 132, 154, 170-171,173-175,177,181,187, 191, 194,205,216-219 mutterings, 94-95, 216 narratives, autobiographical, xiv, 69,89, 97, 101, 106, 108, 118 nativist approach to language, 5-6, 20, 52-53 neurobiological basis of inner speech, 38,41-42,56-57 neuroimaging, xiv, 9, 15, 22, 56-57, 89,

249

108, 112-118 and L2 inner speech, 59, 82-84, 88, 113-114,171,218-219 nonverbal thought, 6, 18-19, 28-29, 46, 71 note-taking, instructional use of, 191, 202,211,217 ontogenesis, xiv, 13, 16-17, 33, 36, 40, 44, 50, 90, 117-118, 153, 159, 168-169,199,211,213,215 output, mental, 165-166, 180 paraphrasing, instructional use of, 191, 202-203,211,217 phonological loop, 23, 54-56, 77, 113, 116, 125, 147, 162-163 planning, instructional use of, 191, 203, 211,217 playback, xiv, 59, 120, 141-142, 145, 164 positron emission tomography (PET), xiv, 9, 56, 82, 84, 88, 112-114 predicativeness, 31-32, 35, 39, 41, 43, 48,90,153,176,183 private speech, xii, xiv-xv, 1, 6, 9, 11, 16-17, 20-21, 24-26, 33, 51-52, 58, 72, 89, 91-95, 97, 117-118, 139,153-155,158,188,194-196, 198,210,213-214,216 in an L2, 8,12, 72-73, 81-82, 86, 9192,95, 117, 124, 128, 132, 138, 147,153-161,163-164,168-169, 175-177,188-189,193-194,197198,200,202-204,213-214,216 written modality, 93, 210, 218 problem solving and inner speech, 8, 46-47, 99, 111, 177, 195 instructional use of, 191, 203-205, 211 proficiency and brain research, 82-84, 88 and L2 inner speech, xv, 65, 80-81, 85-88, 119-120, 122-132, 135138, 151, 153, 177, 183-184,

250

SUBJECT INDEX

189, 192-193,205 conceptualization, 183 levels in author's studies, 120, 139 psychophysiological research, xiii-xiv, 7-9, 22, 43-47, 89, 108-109, 111-118,218 Q-methodology, 89, 97-98, 106, 218 randomly sampling thinking, see thought-sampling reflexive thinking faculty, 53, 56 reflexology/reflexologists, 3, 7, 22 regional cerebral blood flow (rCBF), 57 rehearsal, elaborative, 22, 147, 162 maintenance, 22 mental, xii, 1, 5, 22-24, 26, 52, 55, 164, 179,214 mental, L2, xiv-xv, 8, 23, 26, 59, 77-82, 84, 87-88, 98-99, 119138, 156, 160, 165, 170, 176, 179-181, 184, 198,211 overt, 139, 155, 159, 162, 164 subvocal, 6, 53-54, 56, 58, 87, 113, 147, 197 unrehearsability, 55, 160 relativity, linguistic, 51, 60 repetition, 73, 77, 86, 120, 128, 132, 139,155,157-158,161,197-198 covert, xv-xvi, 22, 45, 73, 77, 86, 126,140-143,147,161-162,191, 197,211 instructional use of, 200-202, 211, 217 retrospection, xiv, 8, 96-104, 106-109, 139, 149,218 second language (L2) operational definition, xi self, 184-188 (re)construction, 68-71, 101, 184, 186-189,215,217 development of the, 47-48, 58, 131, 133, 178, 180-181, 183 self-talk, xii, 1, 7-8, 15-16, 21, 25-26, 47-50,58,98,131-133,178,180,

184-185,187,189,208-209,218 sense/meaning, 4, 32, 47-48, 69, 153, 173-174,206-207 shadowing, instructional use of, 201202,211 sociocomputational theory, xiii, 27, 51 sociocultural theory perspectives on inner speech, xiii, 27-50 principles, xii, 9-14 sociohistorical approach, xii, 4-5, 10, 28,37-38,69,91, 117 speech interference, xiv, 43, 45, 50, 7475, 87, 89, 108-109, 111-112, 218 stimulated recall technique, xv, 76, 99100, 139, 149-150, 152 structural aspects, of inner speech, 11, 16-18, 21, 31, 33,36,40-41,44-48,50,74,153 of L2 inner speech, S6, 119-120, 123-125,137-138,148-149,151, 153, 175-177, 183, 189,215 subvocal speech, 3, 22, 29, 111 summarizing, instructional use of, 191, 201-202,211 supracommunicative view of language, xiii, 5,27, 51-52 think-aloud technique, xiv, 74, 76, 8789, 101-106 thought-sampling, xiv, 53, 89, 97, 103104, 106,218 translation, as teaching method, 60, 63, 194,205,207,216 ventriloquation, 70 verbal reports, xiv-xv, 73, 89, 95-107, 117-119,139,148,150-152,169, 175,218 verbal thought, xii, 1, 4, 6-8, 15, 17-19, 22, 26, 28-30, 33-34, 36, 39, 46, 49, 91, 99, 104, 115, 136, 159, 177, 181, 184, 189,213 in an L2, xiv, 18, 59-73, 84-86, 88,

195,203,215,219

SUBJECT INDEX

vicarious response, 73, 128, 139, 157, 158 vocabulary acquisition, 9, 54, 56, 58, 65, 77, 87, 125-126, 179, 197, 209 whispering, 40-41, 74, 94, 200 working memory, 23, 52-54, 56, 77, 87, 116, 147 Wurzburg school of psychology, 2-3 zone of proximal development (ZPD), 12,81, 157, 162, 180,199-200

251

EDUCATIONAL LINGUISTICS 1. J. Leather and J. van Dam (eds.): Ecology of Language Acquisition. 2003 ISBN 1-4020-1017-6 2. P. Kalaja and A.M. Ferreira Barcelos (eds.): Beliefs about SLA. New Research Approaches. 2003 ISBN 1-4020-1785-5 3. L. van Lier: The Ecology and Semiotics of Language Learning: A Sociocultural Perspective. 2004 ISBN 1-4020-7904-4 4. N. Bartels (ed.): Applied Linguistics and Language Teacher Education. 2005 ISBN 1-4020-7905-2 5. E. Llurda (ed.): Non-Native Language Teachers: Perceptions, Challenges and Contributions to the Profession. 2005 ISBN 0-387-24566-9 6. M.C.M.d. Guerrero, Inner Speech - L2: Thinking Words in Another Language. 2005 ISBN 0-387-24578-2