Learning Strategies in Second Language Acquisition J. Michael O’Malley Georgetown University Anna Uhl Chamot Georgetown University “| CAMBRIDGE UNIVERSITY PRESSPublished by the Press Syndicate of the University of Cambridge The Pitt Building, Trumpington Street, Cambridge CB2 1RP 40 West 20th Street, New York, NY 10011-4211, USA 10 Stamford Road, Oakleigh, Melbourne 3166, Australia © Cambridge University Press 1990 First published 1990 Fourth printing 1995 Printed in the United States of America Library of Congress Cataloging-in-Publication Data O'Malley, J. Michael. Learning strategies in second language acquisition / J. Michael O'Malley, Anna Uhl Chamot. p._ cm. ~ (The Cambridge applied linguistics series) Bibliography: _p. Includes index. ISBN 0-521-35286-X (hardback) — ISBN 0-521-35837-X (paperback) 1. Second language acquisition. 1. Chamot, Anna Uhl. Il. Title. UL, Series. PII8.2.043 1990 418'.007—de20 89-9770 cir British Library Cataloging in Publication Data ‘O'Malley, J. Michael Learning strategies in second language acquisition— ~ (The Cambridge applied linguistics series) 1. Foreign language skills. Acquisition 1, Title Il. Chamot, Anna Uhl 401.9 ISBN 0-521-35837-X paperbackContents Series editors’ preface viii Preface ix 1 Introduction 1 Background 3 Research on learning strategies 3 Theoretical background in second language acquisition 8 Overview of the book 13 2 A cognitive theory of learning 16 Background 16 Language as a cognitive skill 19 Representation in memory 20 Stages of skill acquisition, 25 Complements to the stage-related theory of learning 27 Language comprehension 33 Language production 37 Learning strategies as cognitive skills 42 Definition and classification 43 Strategies as cognitive processes 47 Conclusions 54 3. How cognitive theory applies to second language acquisition 56 Background 57 Relationship of cognitive theory to specific constructs 68 Declarative knowledge 68 Procedural knowledge 73 Stages of skill acquisition 77 Conclusions 83 4° Learning strategies: methods and research 85 A framework for data collection on learning strategies 86 Objective of data collection 86 Language task 88vi Contents Temporal relationship 90 Informant training 91 Elicitation procedures 92 Individual versus group data collection 95 Multiple data collection procedures 95 Issues in the use of self-report data 96 Review of research on applications of learning strategies 98 Definition and classification a7 Descriptions of strategy applications 104 Validation of strategy effectiveness 107 Conclusions 112 Strategies used by second language learners 114 Study 1: learning strategies used by beginning and intermediate ESL students 114 Study 2: learning strategies used by foreign language students 123 Study 3: listening comprehension strategies used by ESL students 128 Study 4: longitudinal study of learning strategies used by foreign language students for different language tasks 133 Summary 143 Metacognitive and cognitive strategies 144 Declarative versus procedural knowledge 145 Stages of skill acquisition 147 Experts versus novices 149 Conclusions 150 Instruction in learning strategies. 151 Issues in instruction 152 Separate versus integrated instruction ake Direct versus embedded instruction 153 Instructional implementation 154 Student characteristics 160 Review of representative studies 165 Instruction in learning strategies for second language acquisition 165 Learning strategy instruction in first language contexts 167 Study 1: learning strategy instruction with students of English as a second language 170 Study 2: learning strategies taught by foreign language instructors 175 Conclusions 184Contents vii 7 Learning strategies: models and materials 187 Instructional models in first language contexts: strategic teaching 187 Instructional models in second language contexts: the Cognitive Academic Language Learning Approach 190 Theoretical framework of CALLA 191 The components of CALLA 193 CALLA lesson plan model 201 Second language learning strategy training materials 204 Learning strategy materials for adult language learners 204 Learning strategy materials for content-based ESL 210 Conclusions 212 8 Summary and conclusions 214 Theoretical developments 214 Research 220 Definitional/classification studies 220 Strategy description 222 Validation studies 224 Glossary 229 References 235 Author index 249 Subject index 2531 Introduction This book is concerned with “learning strategies,” the special thoughts or behaviors that individuals use to help them comprehend, learn, or retain new information. It focuses on the application of learning strategies to second language acquisition by students learning English as a second language as well as by students learning foreign languages. The book addresses the need for an integrated treatment of learning strategies in second language acquisition that is based on theory and research. The theory used here describes how second lan- guages are learned and what role learning strategies play in the lan- guage acquisition process. The theory is also used to organize the presentation of research results, examine the findings, and integrate the results with other studies. The research and theory described in this book are based on a cognitive information processing view of human thought and action. Two fun- damental principles underlying this theory are: (a) that behavior can best be explained by reference to how individuals perceive and interpret their experiences, and (b) that the way in which individuals think and reason has parallels with the manner in which computers process in- formation (Shuell 1986). In cognitive theory, individuals are said to “process” information, and the thoughts involved in this cognitive ac- tivity are referred to as “mental processes.” Learning strategies are spe- cial ways of processing information that enhance comprehension, learning, or retention of the information. This volume presents the view that language is a complex cognitive skill that can be described within the context of cognitive theory. A theory of second language acquisition, to be successful, must be able to describe how knowledge about language is stored in memory and how the process of second language acquisition ultimately results in automatic language comprehension and production. In addition, to be credible, the theory must explain a wide variety of language con- structs that have been discussed in the first and second language ac- quisition literature. For our purposes, we add the further constraint that the theory be able to describe what learning strategies are, how information about learning strategies is stored in memory, how strat- 12 Learning strategies in second language acquisition egies are learned and may become automatic, and why they influence learning in a positive manner. At the onset of our research on learning strategies in 1981, there was no theory to guide our studies and few empirical investigations into the nature of learning strategies and their influence on second language acquisition. What did exist were a few descriptive studies of strategies used by effective second language learners and, quite signif- icantly, over ten years of extensive research in psychology on the in- fluence of learning strategies on reading comprehension and problem solving. The two bodies of research, one in second language acquisi- tion and the other in cognitive psychology, had proceeded fully inde- pendent of each other with little cross-referencing of concepts and approaches across topic areas. Furthermore, the methodologies in the studies were different, the ones in second language acquisition being descriptive, and the ones in psychology being experimental and ori- ented toward training learners to acquire strategies. What these bod- ies of research had in common was an interest in the mental processes of experts compared with novices, and an undeniable pau- city of theory to describe what strategies were or how they influenced learning. The lack of theory to explain learning strategies was com- pounded in second language acquisition studies by the lack of a com- prehensive theory to explain how individuals learn the structures and functions associated with second language use. Only recently have papers appeared in which learning strategies are integrated within cognitive theory (e.g., Rabinowitz and Chi 1987; Garner 1986; Mayer 1988), but when we began our research there was a vacuum with respect to the integration of strategic processing in theories of second language acquisition. We hope to address and at least par- tially resolve this issue in the later chapters. The notion that special learner techniques or strategies might assist second language acquisition is actually quite new, having emerged in the research literature just over ten years ago. The suggestion that the “good language learner” might be doing something special or differ- ent that we could all learn from was introduced at about the same time in work by Rubin (1975) and by Stern (1975). This notion con- trasts sharply with the idea that some people just have an “ear” for language or that some individuals have an inherent ability for lan- guage learning. This early work anticipated what cognitive psycholo- gists were realizing independently, that competent individuals are effective because of special ways of processing information. There was also the suggestion that these strategies are not the preserve of highly capable individuals, but could be learned by others who had not discovered them on their own.Introduction — 3 Background In this section we introduce some of the early studies on learning strat- egies in second language acquisition and cognitive psychology in order to establish a framework for describing the research presented in later chapters. These studies provided the empirical background for the initial investigations we developed. In later chapters we expand upon this se- lective review to present a more detailed examination of research and to analyze some of the more recent findings concerning learning strat- egies, leading up to a detailed description of our own work. This section continues with an introduction to relevant theoretical positions on lan- guage competence and second language acquisition that were current when we began our studies. We expand upon this description in far greater detail in subsequent chapters, and indicate implications for in- structional practice. Research on learning strategies The literature on learning strategies in second language acquisition emerged from a concern for identifying the characteristics of effective learners. Research efforts concentrating on the “good language learner” (Naiman et al. 1978; Rubin 1975) had identified strategies reported by students or observed in language learning situations that appear to con- tribute to learning. These efforts demonstrated that students do apply learning strategies while learning a second language and that these strat- egies can be described and classified. For example, Rubin (1981) pro- posed a classification scheme that subsumes learning strategies under two primary groupings and a number of subgroups, as illustrated in Table 1.1. Rubin’s first primary category, consisting of strategies that directly affect learning, includes clarification/verification, monitoring, memori- zation, guessing/inductive reasoning, deductive reasoning, and practice. The second primary category, consisting of strategies that contribute indirectly to learning, includes creating practice opportunities and using production tricks such as communication strategies. Rubin based her strategies on fairly extensive data collection in varied settings, which included about fifty hours of classroom observation, observation of a small group of students working on a strip story, analysis of self-reports from “a few students” instructed to write down what they did to learn a second language, and analysis of daily journal entries of two students who were directed to report on strategies after having been given strategy examples. The classroom observations proved to be the least useful of these methods for identi iSuruvaur Apizep> 01 soztpenaxaqu0D uon>esaIUT Sre|MUIIOY ssp} sareuoo 4o ‘surAuouds ‘suonnsoyumnoz19 sas) 2 AL 01 Sutuaysy] ‘qe] aenBue] ut aun spuads siuopnas MO]P>J YIM UoNesI9AUOD sareTUT soyeads aaneu qm UORends s97v—2I.) aqeyU 01 sa pue AynyoreD suis] Ausea peounouosd jun ssousuos syeadoy spunos mou yim siusuiuiodxg 30194I990-09 JO S9[N JOy $YOO] spiom sdnoas aSenguy] 398302 0) a8enSue] soypo/aapeu soredwioc, “919 4xa1u0o ‘sommaoid “soamaonas ‘spiom ay woxy Surueous sossonsy Aypareadas suroat sailim ‘1uoWoUU e spuy ‘pnoy ino. ssounouoad ‘surayt Mau Jo a}0U oye aydis ‘ewusesd ‘Gunjeds ‘Axejnqeooa ‘uoneounuosd 8,19 10/UMO UT $10139 81991105) Suipueisiapun wayu0s 01 spiom steadar ‘uoissa1dxa 10 puows 8 98n 0} MOY Jo apdwIexo uv 30} sySy Sajdiuxe oanpinasoudoy sypa woRONporg aonsead 104 sontunys0ddo soo sondeIg Suruoseas oanonpag Sutouoxojut sanonput/Butssoney uonezouDy Surronuopy wonesgHaA/uOnIyIE|, satdaqvas Aavpuoras aaywjuasasday Suruzeay or Apoosipur ainguauod ey sossad0Ig Suruzesy soye Apoomp aeyp sodas uonwoypssojs KBarouss Kavuesig (1861) wqny 4oqiny NOLLISINDOV SDVADNVT GNOOMS NI SAIDTLVULS ONINUVT 10 SNOLLVOIISSVID “|| ATEVLsoyessiur yeodon 01 104 Se Os aaosdunt 01 Ske 10) S400] suonsear Oj $400] pur soousiuss sores9U95 sanjnoyyip 405 pasedoad 5] soyeasiu uato 3e ye] 03 9Iqe S| yeads 01 uonrgiyur sau109294Q, sosesyd pur soisaumo> sozu0uoyy syed wad 01 sonagy sioyeads aanvu 7] ptm seIUOD eUOsIOd asopp sayst|quisy aiqissod saasuoya saqvoqunUI0 suonnoopumoz sasp) yeads 01 arensay 10 sa0q, AroBo1e9 9utes Ut S120 01 spsom AreUOHDIP MOU S>IeprY samupiqissod arezoua8 02 sajna s2s-) umouy Apeoaye st eyA SUIS], soreuBoo Sosy) spunos seay 01 auoye speoy sweados pue ysnour s194vads 3e SYOO] DZMOUIDUL 0} SPIOM UMOP SAILTA sade 03 suaasr] sway feuoUIppe speay aSenSur] 198ye) ul yfasioy/unny Sursrourunt 01 rou asinoo v Sayer pue auouruostAus Suruse9] pammonns 40} poou so8pojmouype ruapmMg ypeqpa9y 10 sioyeads aaneu 71 Sunyse pur ssouosojur Sunsoa Aq uraasds 77 sostaag Apueastioc Suruseay uy spuewiap aansaye yim sadoc sSuruvous pesnyjnoopos spul siyeads 7] quim suonens aanvoiunustios $4995 Advinoov s2a0 Aouony saziseydurg, wroashs & s1 aen8ue] aeyp 1983 Jo asn sayeyy saDu9s9fUl ayeur 02 oBen8ury y81e1 sozdyeuy suosuedusos 71/171 SRW swiogord yenprarput sozdqeuy ssonseid ueadoad wooasse|> 1e]nBa1 02 somtAnse Suruzeay aBenBur] parejas sppy swusuuomau Sutures] sitojdxo pur s¥90s 10 Artunaz0ddo Suusea] 01 Apoanisod spuodsay aoururzopiod 77] SuuoMUOpy spuvuiop aansoye Jo 1usue8euEyy uonoezaut pue uoneornunuos jo suvaut e se o8enBue] jo uonezipeoy wiaisks, e se o8endue] jo uonezseay (8261) yproidde yse1 aanoy “Je 19 ueUEN,6 Learning strategies in second language acquisition An alternative classification scheme proposed by Naiman etal. (1978), also shown in Table 1.1, contains five broad categories of learning strat- egies and a number of secondary categories. The primary strategies were found to be common to all good language learners interviewed, whereas the secondary strategies were represented only in some of the good learners. The primary classification includes an active task approach, realization of language as a system, realization of language as a means of communication and interaction, management of affective demands, and monitoring of second language performance. Naiman et al. based this classification scheme on interviews with thirty-four good language learners and an initial strategy scheme suggested earlier by Stern (1975). Naiman et al. also identified what they referred to as “techni second language learning, which differed from strategies in their scheme by being focused on specific aspects of language learning. The techniques, with selected examples of each, are as follows: Sound acquisition repeating aloud after a teacher, a native speaker, or a tape; listening carefully; and talking aloud, including role playing. Grammar following rules given in texts; inferring grammar rules from texts; comparing L1 and 1.2; and memorizing structures and using them often. Vocabulary making up charts and memorizing them; learning words in context; learning words that are associateds using new words in phrases; using a dictionary when necessary; and carrying a notebook to note new items. Listening comprehension listening to the radio, records, TV, movies, tapes, etc.; and exposing oneself to different accents and registers. Learning to talk not being afraid to make mistakes; making contact with native speakers; asking for corrections; and memorizing dialogues. Learning to write having pen pals; writing frequently; and frequent reading of what you expect to write. Learning to read reading something every day; reading things that are familiar;Introduction 7 reading texts at the beginner’s level; and looking for meaning from context without consulting a dictionary. Among the various techniques Naiman’s group identified, those asso- ciated with vocabulary learning were used most frequently. This is sig- nificant because it suggests that learners either have difficulty in identifying what techniques they use to learn other tasks or have few strategic processes for doing so. More recently, Oxford (1985) has com- piled an extensive list of strategies identified in these studies and in our early studies that will be described in this book (e.g., O’Malley, Chamot, Stewner-Manzanares, Kipper, and Russo 1985a). However, Oxford’s approach was not available when we began our work. As can be seen from an inspection of the strategies in Table 1.1 and from the Naiman group’s techniques, a number of highly useful delib- erate approaches to learning a second language have been identified. Most of these emerged from interviews or, with Rubin, from interviews and diaries. The Rubin and Naiman et al. classification schemes are substantially different, however, and do not have any grounding in the- ories of second language acquisition or cognition. Consequently, it is difficult to winnow out from the extensive listing of strategies and tech- niques which ones are fundamental for learning, which ones might be most useful to other learners, and which should be combined with others to maximize learning effectiveness. Research on training second language learners to use learning strat- egies has been limited almost exclusively to applications with vocabulary tasks. Dramatic improvements in vocabulary learning tasks presented in one-on-one training have been reported in these studies. The typical approach in this research has been either to encourage students to de- velop their own associations for linking a vocabulary word with its equivalent in the second language (Cohen and Aphek 1980, 1981) or to train students to use specific types of linking associations to cue the target word, such as the keyword method (e.g., Atkinson and Raugh 1975; Levin 1981; Pressley et al. 1980; Pressley et al. 1981). The strategy training in these vocabulary studies was given individually or was pro- vided to groups by an experimenter using special audio equipment for each subject. There were no instances in which training in learning strategies in second language acquisition was performed in a natural classroom instructional setting or by the teacher of the students who served as subjects. In cognitive psychology, studies of learning strategies with first lan- guage learners have concentrated on determining the effects of strategy training on different kinds of tasks and learners. Findings from these studies generally indicated that strategy training is effective in improving the performance of students on a wide range of reading comprehension and problem-solving tasks (e.g., Brown et al. 1983; Chipman, Segal,8 Learning strategies in second language acquisition and Glaser 1985; Dansereau 1985; Segal, Chipman, and Glaser 1985). One of the more important outcomes of these psychological studies was the formulation of learning strategies in an information-processing the- oretical model. This model contains an executive, or metacognitive, function in addition to an operative, or cognitive-processing, function. Metacognitive strategies involve thinking about the learning process, planning for learning, monitoring of comprehension or production while it is taking place, and self-evaluation after the learning activity has been completed. Cognitive strategies are more directly related to individual learning tasks and entail direct manipulation or transformation of the learning materials (Brown and Palincsar 1982). A third type of learning strategy identified in the literature on cognitive psychology concerns the influence of social and affective processes on learning. Examples of so- ciallaffective strategies are cooperative learning, which involves peer interaction to achieve a common goal in learning (e.g., Dansereau et al. 1983; Slavin 1980), and asking questions for clarification. Affective strategies are represented in the exercise of “self-talk,” the redirecting of negative thoughts about one’s capability to perform a task with as- surances that the task performance is within reach. Cooperative strat- egies have been shown to enhance learning on a variety of reading comprehension tasks (Dansereau et al. 1983). Research in metacognitive and cognitive learning strategies suggests that transfer of strategy training to new tasks can be maximized by pairing metacognitive strategies with appropriate cognitive strategies (Brown et al. 1983). Students without metacognitive approaches are essentially learners without direction or opportunity to plan their learn- ing, monitor their progress, or review their accomplishments and future learning directions. As will be seen in later chapters, the issue of transfer is far from being resolved. Theoretical background in second language acquisition There has been no comprehensive analysis describing the influence of cognition in second language acquisition. Nevertheless, at the time of our initial investigations, a number of theorists had articulated positions that included a cognitive component in second language processes. The- oretical efforts that can assist in identifying the role of cognition in second language acquisition had emerged in two general areas: the at- tempt to describe language proficiency or language competence, and the attempt to explain influences on second language acquisition. In each of these areas, cognition had been described and defined but, as will be seen, was not discussed in sufficient detail to delineate the role of cog- nition or strategic processing in second language acquisition. Language proficiency has been described by Cummins (1984) in termsIntroduction 9 of two continua that concern task difficulty and the context in which language occurs. Difficulty may vary from cognitively undemanding tasks, such as learning definitions or reading road signs, to cognitively demanding tasks, such as reading or making an oral presentation on an academic topic. The context for language use may vary from contexts that are embedded, or enriched with linguistic or paralinguistic cues for meaning, to contexts that are reduced, or absent of such additional cues to meaning. Academic tasks, for example, tend to be cognitively de- manding and usually require language in which contextual cues for meaning are reduced. Tasks outside the classroom, on the other hand, are often undemanding cognitively and are characterized by language that either has rich contextual clues or is formulaic and therefore easy to comprehend or produce. The task difficulty dimension, although based on the cognitive demands of the task, has not been used by Cum- mins to describe the potential role of strategic cognitive processes in enhancing learning or task performance. The fundamental concept of language competence expressed by Cum- mins was extended by Tikunoff (1985) in a model intended to elaborate on the description of student functional proficiency in academic settings. To Cummins’s (1984) notion of academic language proficiency, Tikunoff added three intersecting concepts: interactional, academic, and partici- pative competence. For example, successful participation in a classroom setting requires that a student: (1) observe classroom social rules of discourse, (2) function at increasingly complex cognitive levels, and (3) be competent in the procedural rules of the class. As Cummins did with language competence, Tikunoff included a cognitive component but did not elaborate on the significance of strategic behavior for enhancing student comprehension or learning. Other models of language competence also contained cognitive com- ponents but left the role of learning strategies ambiguous. Canale and Swain (1980) proposed a theoretical framework in which communica- tive competence has three major components. The first is grammatical competence, which includes vocabulary and pronunciation as well as grammatical structures and word forms. The second is sociolinguistic competence, which is made up of sociocultural rules for using language appropriately and discourse rules for linking parts of a language text coherently and cohesively. The third component of the Canale and Swain model is strategic competence, which consists of yerbal and non-verbal communication strategies that may be called into ac- tion to compensate for breakdowns in communication due to performance variables or to insufficient competence. (p. 30) In this model, the strategic component refers to communication strat- egies, which can be differentiated from learning strategies by the intent10 Learning strategies in second language acquisition of the strategy use. That is, learning strategies have learning as a goal, and communication strategies are directed toward maintaining com- munication (Tarone 1981). The second area in which studies have assisted in identifying the role of cognition in second language acquisition was the theoretical efforts to identify important influences on second language acquisition. These efforts varied considerably in their attention to cognitive and strategic processing. One of the theorists who included an articulated cognitive component, Bialystok (1978), identified four categories of learning strat- egies in her model of second language learning: inferencing, monitoring, formal practicing, and functional practicing. In this model, learning strat- egies are defined as “optimal means for exploiting available information to improve competence in a second language” (p. 71). The type of strategy used by the learner depends on the type of knowledge required for a given task. Bialystok discussed three types of knowledge: explicit linguistic knowledge, implicit linguistic knowledge, and general knowl- edge of the world. She hypothesized that inferencing may be used with implicit linguistic knowledge and knowledge of the world. Monitoring, formal practicing (such as verbal drills found in a second language class), and functional practicing (such as completing a transaction at a store) contribute both to explicit and implicit linguistic knowledge. That is, strategies introduced explicitly in a formal setting can contribute to implicit linguistic knowledge and therefore to students’ ability to com- prehend and produce spontaneous language. Bialystok’s model can be contrasted to Krashen’s Monitor Model (1982), which does not allow for contributions of explicit linguistic knowledge (learning) to implicit linguistic knowledge (acquisition). The Monitor Model includes two types of language processes: “acquisition” and “learning.” “Acquisition” is described by Krashen as occurring in, spontaneous language contexts, is subconscious, and leads to conver- sational fluency. “Learning,” on the other hand, is equated with con- scious knowledge of the rules of language derived from formal and traditional instruction in grammar. The “monitor” is a conscious process in which the learner applies grammatical rules to language production (either oral or written), which means that the monitor is a highly delib- erate form of processing. In Krashen’s view, “learning” does not lead to “acquisition,” because the sole function of learning is to act as a monitor or editor of the learner’s output. Therefore, the inescapable conclusion of this model is that conscious use of learning strategies will make little contribution to the development of language competence. A comprehensive effort to integrate linguistic with affective and cog- nitive components of learning by Wong Fillmore and Swain reserved an important role for learning strategies in the cognitive component (Wong Fillmore 1985; Wong Fillmore and Swain 1984). Learning strategiesIntroduction il were said to be the principal influence on the rate and level of second language acquisition for children, whereas inherent developmental and experiential factors were considered to be primarily responsible for first language acquisition. The types of strategies described by Wong Fillmore (1985) appear to be more global than those usually described in cognitive psychology, and include knowledge and mental skills as well as strategic processes. Wong Fillmore (1985) suggests that strategies include associative skills, memory, social knowledge, inferential skills... analytical skills... pattern recognition, induction, categorization, generalization, infer- ence, and the like. (p. 37) Wong Fillmore noted that differences in the rate and level of second language learning are due to the involvement of general cognitive pro- cesses, especially those that are important in language learning. In con- trast, the consistency of first language acquisition across individuals is purportedly linked to inherent language acquisition mechanisms. The role that the strategies play with regard to the other model components or to mental processes in second language learning was not identified. Movement toward a more cognitive view of second language acqui- sition was evident in the information processing approach suggested by McLaughlin, Rossman, and McLeod (1983). The learner is viewed as an active organizer of incoming information, with processing limitations and capabilities. While motivation is considered to be an important element in language learning, the learner’s cognitive system is central to processing. The learner is able to store and retrieve information de- pending on the degree to which the information was processed. Evidence for aspects of the information processing model comes from studies of language processing and memory. One implication of information pro- cessing for second language acquisition is that learners actively impose cognitive schemata on incoming data in an effort to organize the infor- mation. McLaughlin et al. (1983) drew on cognitive theory in suggesting that learners may achieve automaticity in second language acquisition by using either a top-down approach (or knowledge-governed system), which makes use of internal schemata, or a bottom-up approach (or an input-governed system), which makes use of external input. In either case, cognition is involved, but the degree of cognitive involvement is set by the interaction between the requirements of the task and the knowledge and mental processes used by the learner. More recently, Spolsky (1985) proposed a model of second language acquisition based on preference rules in which cognitive processes play an important role. In his view, three types of conditions apply to second language learning: necessary conditions, gradient conditions, and typi- cality conditions. A necessary condition is one that is required for learn- ing to occur. Examples of necessary conditions in second language12 Learning strategies in second language acquisition acquisition are target language input, motivation, and practice oppor- tunities. The gradient condition is one in which the more frequently the condition occurs, the more likely learning is to take place. Examples of a gradient condition might be the greater or lesser degree to which a learner actively seeks out interactions with native speakers of the target language, or the greater or lesser degree to which a learner can fine tune a learning strategy to a specific task. The third type of condition is one that typically, but not necessarily, assists learning. An example of a typicality condition might be that of risk taking; thus, outgoing person- alities tend to be good language learners as a rule, although in some cases quiet and reflective persons can be equally or more effective learners (Saville-Troike 1984). Spolsky’s model of second language acquisition contains two clusters of interrelated conditions representing these three types. The first cluster contains social context conditions, such as the learning setting and op- portunities. The second cluster consists of learner factors, such as ca- pability, prior knowledge, and motivation. The learner makes use of these latter conditions to interact with the social context of learning, and this interaction leads to the amount of language learning that takes place. Thus, this model accounts for variability in second language learn- ing outcomes through differing degrees of (or preferences for) application of gradient and typicality conditions. In Spolsky’s model, learning strat- egies, while not specifically identified as such, would be part of the capabilities and prior learning experiences that the learner brings to the task. A precise description of the role of strategic processing in second language learning was missing from these theories of second language proficiency and acquisition. Although some of the theories proposed a cognitive component, and some indicated that cognitive processes influ- ence proficiency or the rate and level of acquisition, the manner in which the influence of cognitive processes is exerted with respect to other mental processes or with respect to language tasks had not been described. Further, although information processing theory had been used to clas- sify strategies into metacognitive and cognitive categories, agreement on the assignment of individual strategies to these two broad groupings had been difficult to achieve (Brown et al. 1983), and neither the theory nor the research had been extended to second language acquisition. There was a need for clarification of the role of learning strategies in second language acquisition from both an empirical and a theoretical standpoint. One step that would help to clarify the definition and as- signment of discrete strategies to a classification scheme would be to describe the correspondence between mental processes that have been identified in cognitive theory and strategic processes described in the learning strategies literature. Another step that would help in under-