Aljaafreh & Lantolf 1994 Negative Feedback As Regulation and Second Language Learning in The Zone of Proximal

Negative Feedback as Regulation and Second Language Learning in the Zone of Proximal Development Author(s): Ali Aljaafreh and
James P. Lantolf Source: The Modern Language Journal, Vol. 78, No. 4 (Winter, 1994), pp. 465-483 Published by: Wiley on behalf of the National Federation of Modern Language Teachers Associations Stable URL: http://www.jstor.org/stable/328585 . Accessed: 09/05/2013 09:17
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Wiley and National Federation of Modern Language Teachers Associations are collaborating with JSTOR to digitize, preserve and extend access to The Modern Language Journal.
http://www.jstor.org
This content downloaded from 82.18.130.62 on Thu, 9 May 2013 09:17:57 AM All use subject to JSTOR Terms and Conditions
Negative Feedback as Regulation and Second Language Learning in the Zone of Proximal Development
ALI ALJAAFREH Department ofLinguistics University ofDelaware DE 19716 Newark, JAMES P. LANTOLF and Linguistics ofModern Department Languages Cornell University Ithaca,NY 14853 Email:jpl5@cornell.edu confirmation checks,recastingsand the like,or in the formof provisionof thecorrect explicitly form by teacher, peer, or native interlocutor; thatit maybe accompanied by an explanation, especiallyin the classroom setting;and, thatin the classroom settingmore attentionis apparentlygivento discourse and contenterrorsthan to either lexical, grammatical,or phonological errors (7: p. 141). Despite the informativefindings that have emerged fromthe ethnographictradition,this research has not established a strongempirical link between corrective feedback and interlanguage development. Some authors,such as van Lier (p. 182), argue that correction is "an importantvariable in language learning" and further assume it to be a necessary, thoughnot a condition forlearning a second lansufficient, guage. Others, like Chaudron, are less sanguine, pointingout thatresearch on errorcorrectionboth in and out of the classroomsetting has uncoveredinconsistencies, and ambiguities, a general ineffectiveness in termsof its effects on language learning (7: p. 145). Chaudron then calls for "longitudinalresearch . .. to determine the extent of learning possible from feedback" (7: p. 152). The experimental approach to the studyof error correction,derivingmuch of its impetus from the L2 research informed by Universal Grammar (UG), has squarely confronted the mostfundamentalof the above questions: Does negative feedback lead to L2 learning? Research on this question has been spurred, in part,bythe claim thaterrorcorrectionis apparcondientlyneithera necessarynor a sufficient tion forL1 learningin children (e.g., 2; 44; 45). The controversy that has arisen in the experimental L2 literatureis whetherwe can assume
SINCE THE LATE SIXTIES AND EARLY SEVenties, with the ground breaking publications of Pit Corder, Burt and Kiparsky, George, and Richards (14), one of the central themesof second language research has been the studyof learner errorsas a reflectionof hypothesis teston the of second learners (8; ing part language the attentionof those 15; 20; 23;).1 Eventually, workingon learnererrorshas moved awayfrom the analysisof errorsin theirown rightas indications of hypothesis testingand interlanguage developmentto concern withquestions relating to the potentialeffects of correctiveprocedures on language learning. The fundamentalquestion is: does error correction lead to learning, or are corrective moves by teachers or other caretakersineffective? In addition, some corollaryquestions have also been addressed,including how and when errors should be corrected (7: p. 135). Research aimed at answeringthese questions has been carried out in eitheran ethnographic or an experimentalframework.2 Those working within the ethnographic approach have conducted careful observational studies of corrective behaviors both in the classroom and natural settingsinvolvingteachers, learners,peers, native and non-nativespeakers (7; 8; 17; 22; 24; 36). This research has shown, among other things, that corrective feedback may be message-focused or code-focused; that it may be self- or other-initiated and self- or othercompleted; thatcorrectivefeedback may occur implicitlyin the form of comprehension and The Modern 78,iv (1994) Language Journal,
0026-7902/94/465-483 $1.50/0
?1994The Modern LanguageJournal
466 the same situationto hold foradult L2 learning as seems to be the case forchild L1 acquisition with regard to error correction, or whether adults require correctivefeedback (as van Lier assumes) in order to successfully project a grammarof theirL2. The experimentalresearch on error correction, unlike its ethnographiccounterpart,has generated some positive findingson the interrelationship between negative feedback and learning.A series of recentstudies (5; 6; 21; 23; 32; 34; 35), forexample, has demonstratedthat L2 learners provided with correctivefeedback do indeed outperform control groups given minimal or no negative input. DeKeyser somethatalwhatpessimistically concludes, however, in some results error correction perforthough for some learners,it fails mance improvement to achieve much in the wayof across-the-board impact on learning (p. 504 and p. 510). In his studyof high school L1 speakersof Dutch learning French, DeKeyser (p. 511) uncovered no main effectsfor error correction but did find between such factorsas preinteractioneffects and anxvious experience,aptitude,motivation iety,and feedback. DeKeyser is hardlyalone in recognizing that correctivefeedback, if it is to have any impact on learning, has to in some way be attuned to the individuallearner.Birdsong (p. 150),forexample, in reviewingthe research on error correction in L2 learning,concludes thatnegative evidence might be a question of "individual and/or situationalvariation."Day,et al (p. 143) mayhave somesuggestthatlearnerpersonality and "the do with to amount type of corthing rection supplied" by the teacher (p. 43). Sharwood Smith (30; 31) and Schachter argue that and linguisticdevellearners'internalstrategies an importantrole in deteropment may play of negative feedback mining the effectiveness and recommend that these should be investigated in greater depth than theyhave to the present time. Finally, Spada and Lightbown typesof er(p. 219) wonder "whetherdifferent at ror correction strategiesare more effective timesin learners'development. .. and different teachersgo throughdifferent whether phases in theirerror correctionbehaviors depending on theirperceptions of learners' development." As we will argue on the basis of the evidence to be considered below,while error correction comes down to adjustingfeedbackto ultimately the individual learner, adjustmentscannot be determined a priori; rather,theymust be collaboratively negotiated on-linewiththe learner.
TheModern 78 (1994) Language Journal How the negotiationprocess unfoldsand how it leads, or fails to lead, to L2 developmentforms the principal aim of the present paper. Although the findingsof both ethnographicand experimental research on correctivefeedback we are still a long way have been informative, froma full understandingof how feedback interactswith the L2 learning process. From the experimental studies we have some evidence, that corDeKeyser's doubts notwithstanding, rectivefeedback appears to enhance learning. thatcontrolled We mustbear in mind,however, as it may be, as informative experimentation, reallytells us littleabout how individualsreact to and use, or fail to use, feedback to change their interlanguage. The ethnographic research, on the other hand, has provided fairly about the nature of the feedrich information back processes at the local level, but it has not provided us withmuch evidence of if,and how, these processes resultin learning.What is missing is a way of linkinglearning outcomes with specific feedback procedures. We hope the research to be discussed below will serve as an initial step in thisdirection. As for the corollaryquestions, Chaudron (7: p. 152),forone, pointsto the need to determine "the typesof feedback thatwould best succeed in promotingprogressin the targetlanguage." In this regard, Carroll and Swain (p. 361), report thatlearnerswho received explicitcorrection procedures, defined as any feedback that "overtlystates that a learner's output was not generally part of the language-to-be-learned," performedbetter on their experimental tasks than those learners given implicit feedback, construed as any instance of feedback from whichlearnershave to inferthattheirlinguistic performancewas inaccurate (e. g., confirmation checks, failures to understand, and requests forclarification).Carroll and Swain reason that explicit feedback might have been more beneficialbecause it identifiesthe precise location and natureof erroneous performance, while implicitnegativecorrection,requires the learners to engage in a good deal of mental guesswork.3 As encouraging as the results of such research are, it would be prematureto conclude, thatCarroll and Swain and we are not implying do, thatexplicitfeedbackwill alwaysand everywhere have the upper hand over implicitcorrection. Our claim, to be fleshed out below, is forlinthatboth kindsof feedback are relevant guistic development,but their relevance must be negotiated between the novice learner and
andJames P Lantolf Ali Aljaafreh the expert knower of the language. In other words,in some cases implicitcorrectionis sufficient to promote learning,while in others,it is not, and in such cases, explicit feedback is the onlytypeof correctionthatwill elicit a reactive response fromthe learner.As we will argue, the relevance of the type of feedback offered (as marked by a learner's reactive response to the feedback) is as importantan index of development in a second language as are the actual linguisticformsproduced by the learner.
467
vidual level; first between people (interpsychological),and then inside the child (intrapsychological)" (39: p. 57). Centralto the evolutionof external,or social, functionsinto internal,or mental,functionsis the processof internalization, or more properly for sociocultural theory,appropriation (26: p. 64). Zinchenko (p. 106) refers to thisprocess as "the bridge between external and internalactivity." Critically, appropriationdoes not simplyreproduce the mental activity of another individual; rather, it "transforms the process itself and and functions"(38: p. 163). changes itsstructure THE ZONE OF PROXIMAL DEVELOPMENT Sociocultural theory,then, insiststhat internal AND MICROGENESIS IN L2 LEARNING and externalfunctionsare related,while at the The data to be considered in this paper are same timerejectinganypresumption thatone is taken froma largerstudyon the interactionof simplya copy of the other (46). The studyof errorcorrectionand learningwhichitselfis the development,then, for sociocultural theory,is firststudywe are aware of that attemptsto inthe studyof how mediationalmeans are approvestigatethe correction/learning interfacelonpriated by the individualas a resultof dialogic interaction withotherindividuals. gitudinally from within a theoretical, rather than from a phenomenological, stance. The the appropriationprocess is not Importantly, framework to which we referis that developed only observable during ontogenesisof children into adults, but also during microgenesis in (37), his colleagues, and folby Lev S. Vygotsky lowersthat has come to be knownas the sociowhich processes undergo change "rightbefore cultural theoryof mind. In particular,we rely one's eyes" in the space of a fewdaysor week,or on Vygotsky's notion of zoneofproximal of seconds (39: develop- even a fewseconds, or fractions ment (ZPD) to analyze the interactionbetween p. 61). Research that overlooks microgenetic error correction and the learning process as it oftenfailsto detectwhatis oftenthe most growth unfolds during the dialogic activity collaband informative data on learningand interesting constructedby learner and tutor. mental activity oratively (41: p. 55). It is in microgenesis A fundamentaltenet of sociocultural theory thatwe will search for evidence on the interacis its thesisthathuman mental activity is essention betweenerrorcorrectionand L2 learning. and The transition from inter- to intramental tiallya mediated process in whichsymbolic, the most whetherin ontogenesis or microsocioculturallyconstructed,artifacts, functioning, pervasiveof whichis language, playan essential genesis,is a dynamicprocess of reconstruction role in the mental lifeof the individual.Hence, and qualitativechange in which the novice and a mutual linguisticactivity, including speaking and writ- the expertcollaborate in constructing frame.This activity ing, is an indispensable component of such frame,or ZPD, repactivity mental operations as voluntary volunresents a crucial move by Vygotsky to link his memory, tary attention,planning, monitoring,the for- theoretical concepts with practical psychologimation of intentions, rational thought, and cal and educational problems. In formulating these processes are inthe concept of the ZPD, Vygotsky was criticalof learning. Furthermore, social in origin and theirdevelopment herently psychologicaland educational practices which in children proceeds from the social, or inter- assess development and guide educational inmental domain, to the individual,or intramen- tervention solelyon the basis of the levelof indital domain, as a consequence of the linvidual, independentfunctioning. Instead,he insisted that two developmental levels of the guistically mediated interaction which arises between childrenand other,oftenmore experiindividual mustbe taken into account: the acenced, membersof theirsocioculturalworld,intual developmental level, "established as a recluding parents,teachers,siblings,older peers, sult of certain already completed developmental etc. According to Vygotsky, the ontogenesis of cycles" (39: p. 85), and the level of potential mentalfunctionsis captured in the genetic law of development,the level at which the individual cultural as follows:"everyfunctionin functionswithassistancefrom,or in collaboradevelopment the child's culturaldevelopmentappears twice: tion with, more experienced members of first,on the social level, and later,on the indisociety.
468 the potential level of According to Vygotsky, developmentvaries independentlyfromactual developmentand is more indicative of mental growththan actual development.Thus, for instance, two individuals who achieve the same score on a given test, language or otherwise, may not both be able to make use of the help offeredby a tutor (e.g., teacher or other student) to generalize theirlearning to novel circumstances (4). In Vygotsky's view,the learner who is able to respond to such help must be considered to be at a more advanced developmental level than the one who fails to do so, because the learner who responds to help can be expected to showa more rapid rate of actual then,the ZPD is "the development.Specifically, distance between the actual developmental level as determined by independent problem solving and the level of potential development as determinedthroughproblem solvingunder adult guidance or in collaboration with more capable peers" (39: p. 86). The actual level of development defines development "retrospectively"(39: p. 87), while the ZPD defines development "prospectively"(ibid).4 The ZPD is the framework, par excellence, which brings all of the pieces of the learning settingtogether-the teacher,the learner,their theirgoals and mosocial and cultural history, tives,as well as the resources available to them, including those that are dialogically condraws atstructed together.Indeed, Vygotsky of the ZPD as a tool forthe tentionto the utility researcher to explore and come to an understandingof the internalcourse of development when he states: "By using this method we can take account of not onlythe cyclesand maturation processes that have already been comin the pleted but also those that are currently state of formation,that are just beginning to mature and develop" (39: p. 87). Mechanisms Help in theZPD. On the ofEffective and the empirical statements basis ofVygotsky's and theoretical work of contemporarysociocultural researchers (9; 42), specific mechanwithin the ZPD isms of effectiveintervention should becan be identified.First,intervention graduated. Help providedbya more experienced member in a joint activityis designed to discover the novice's ZPD in order to offerthe appropriate level of assistance and to encourage the learner to function at his or her The purpose here is to potentiallevel of ability. estimate the minimum level of guidance reperforma quired by the novice to successfully normallystartsat a given task. Help, therefore,
TheModern 78 (1994) Language Journal highly strategic, or implicit, level and progressivelybecomes more specific, more concrete, until the appropriate level is reached as determinedby the novices response patternsto the help.5 Second, help should be contingent, meaning thatit should be offeredonlywhen it is needed, and withdrawn as soon as the novice shows signs of self-control and abilityto function independently.Research by Wertschand his colleagues (41-43), for example, has shown that children often overtlyreject help offered by parents once the children realize that they are capable of carryingout a task alone. Graduation and contingency workin tandem in such a waythatthe expert,togetherwiththe novice, tries to discover the ZPD of the novice in order to determineif help is required and if it is, tojointlyworkout the appropriatelevel at which to provide it. The process is thus one of of the novice's needs and continuous assessment of help to those condiabilitiesand the tailoring tions. This process can be accomplished only throughthe collaborativeinteractionof the expertand thenovice,whichbringsus to the third mechanismof help in the ZPD. Discoveringthe potentialdevelopmentallevel of the novice and providingappropriate help accordinglyis at its core-a dialogicactivitythat unfolds between more capable and less capable individuals.Dialogue is an essential component of Vygotskyan theory(40; 41), and hence of the ZPD. Without impossibleto dialogic negotiation,it is virtually discover the novice's ZPD (42). THE STUDY We now consider some of the findingsof a of negthe effects studydesigned to investigate withinsoativefeedback,or more appropriately on the miciocultural theory, other-regulation, crogenetic developmentof a second language among adults. Since the full studyfromwhich this paper is drawn is quite extensive in the scope of itsfindings(1),we willhave to limitour analysishere to onlysome of the data thatillustrate how the negotiation of corrective feedin the ZPD promotes back, or other-regulation, learning.6 As an initial endeavor to investigate Subjects. second language learning in the ZPD, we decided to relyon writtentextsratherthan oral production, not only because we felt written performance would expedite data collection, but also because we thoughtit would facilitate betweenthe expert (researcher) the interaction and the learners. Consequently,the project in-
P Lantolf andJames AliAljaafteh volved studentsenrolled in an eight-week second level (the mostadvanced level being 6) ESL writing and reading course offered by the of English Language Instituteof the University Delaware. Althougha total of nine studentsparticipated in the full study,only the three who were in the ZPD group are considered here. Since itwas necessaryforthisgroup to take part in one extra tutorialper week with one of the researchers outside of their five weekly classroom meetings,the course instructor asked for volunteers, informing the students that they would receive one free tutoring session per week and would be helpingthe tutor/researcher in a studyof how language teachers can help learners. They were given no additional information relevantto the nature of the research project. Since all of the studentsvolunteered, the teacher randomlyselected threeindividuals for the ZPD group-one Japanese (Y in the protocols), one Spanish (N in the protocols), and one Portuguese (F in the protocols) speaker. Two of the students (Y and F) had been in the US fortwomonthsand one (N) had resided in the country for six months at the time of the study. All were female.7 Procedures. As part of the course syllabus,students were expected to writeone in-class essay per week on a topic of theirchoice fora total of eight compositions. The learners in the ZPD group were informed that theywould receive correctivefeedback during their tutorialswith the researcher and not from their instructor. The tutorialswere conducted in a one-on-one formatin the tutor'soffice.Each session lasted to forty-five minutes.All sessionswere authirty for later analysis. diotaped in theirentirety The students' firstcomposition was not corrected. It was used to develop an initial profile of the learners' grammatical competence and to determine, to some extent at least, likely problem areas. Although learners were given of errors, help during the tutorialson a variety for purposes of analysis,those structuresthat had a high probabilityof recurrence in subsequence essayswere selected. Four grammatical featuresmet our-admittedly less than rigidcriteria:articles,tense marking, use of prepositions,and modal verbs. Since thiswas a writing and reading course, the particular instructor did not believe it necessaryto provide much in the way of explicit formal instructionon the grammatical properties of English. On occahe did answergrammaticalquession, however, tions posed by the studentson a wide arrayof topics.8
469 Priorto each tutorial, thetutor read each essay in orderto detectproblems, butitis important to note that at this point he made no attemptto prepare a specificsetof corrective proceduresto be followed with the student. Corrective procedures in the ZPD mustbe negotiatedbetween the novice and the expert. The idea is to offer just enough assistance to encourage and guide the learner to participatein the activity and to assume increased responsibility for arrivingat the appropriate performance.One simplycannot determinethisbeforehand, without comproin the ZPD misingthepotentialthatjointactivity has to promote learning.The expert, however, must tryto be sensitiveto the learners' actual levelof competence,and in Wertsch's (41: p. 176) "lure" theminto functioning in an terminology, without the task frusappropriate way making trating. Before beginning the collaborative in each case the learner phase of each tutorial, was asked to read her essay,underlinewhatever errorsshe could find,and correctwhatevershe could. During this time, the tutorwas present, but was busyinghimself withother tasks.When the learner indicated that she had completed the reading and error correction process, the tutorjoined her and collaborative correction began. In the followingparagraph we outline a tutorial; generalized schemata of a prototypical as the protocols taken fromthe actual however, tutorialsshow, thingsdid not usually proceed and the interactions were ofquite so smoothly ten quite complex. At the outset,the tutorasks the learnerifshe encounteredanyerrorsduringher privatereading of the essay.If the learneridentifies anymistakesbutfailsto correctthem, or does so erroneously,theyare dealt with as the pair considers each sentence of the essay.Whenevera target erroris discoveredor whenever thelearnerasksa question about some aspect of the composition, thereadingprocesshaltsand thecorrection process begins. The tutordirectsthe learner'sattention to a particularsentence containing an error and asks a general question of the type:"Do you notice any problem, or is there anything fails to wrong in thissentence?" If thisstrategy produce a response, the learner's attentionis then narrowed to the line or phrase in which the error appears, using an utterancesuch as: "Is there anythingwrong in this line or segment?"If thisalso fails to prompta response, a more explicitstrategy is adopted. Thus, help is elaborated until the learner shows increasingly signs of responsiveness toward the error at hand. If the narrowingstrategy also fails,a spe-
470 cific clue is offeredwhich indicates the nature of the error,forexample: "Pay attentionto the tense of theverb." If the subjectis stillunable to recognize the error,the tutoridentifiesit and the learner is asked to correctit. If the learner fails here, the tutormoves to even more direct formsof regulation,such as "Use the past participle of the verb." If this also produces no response, clues about the correct answer are if thisproduces no responsiveacgiven. Finally, tion from the learner, the tutor provides the correct answer.This is accompanied by a brief grammaticalexplanation when the tutorfeelsit to be helpful. Two criteriawere used Criteria. Developmental to determine the microgeneticgrowthof the learners' interlanguage.The first, a more traditional product-oriented was to search criterion, for signsof improvement in the subjects' use of the relevant featuresin subsequent eslinguistic says. Here concern focused on a reduction in frequencyor complete eradication of those errors as well as on the generalizationof learning beyond the specific cases forwhich the learner had received help.9 The second criterionis, we believe, quite distinctfrom anythingreported in the L2 literature to date, and this is a criterion that falls out of learning in the ZPD-does the naturally learner show signs of movementawayfromreand toliance on the tutor,or other-regulation, ? This wardsreliance on the self,or self-regulation of and quality was determined by the frequency help that the learner elicited fromthe tutorin the correctionof the same errorin subsequent episodes in the same tutorialsession and in subsequent tutorials dealing with new composiwe observed five gentions. More specifically, eral levels of transition from intermental to intramental functioningas the learners moved the ZPD toward self-regulationand through These levels control over the targetstructures. loosely parallel the transitional stages uncovered by Wertsch and his colleagues for child development (42, 43).10 The levels are characterized by varyinginstantiationsof three parameters,namely,need for intervention, noticing the error. an error,and correcting Level 1. The learner is not able to notice, or from correct the error,even with intervention At thislevel,the learnerdoes not have the tutor. a sufficientbasis from which to interpretthe tutor'smoves to provide help, and probablyhas no awarenessthatthereis even a problem. The mustassume fullresponsibility tutor, therefore, forcorrectingthe error.Thus, ratherthan pro-
TheModern 78 (1994) Language Journal viding corrective help, the tutor's task is to bring the targetforminto focus and, in so doing, begin the process of co-constructingthe ZPD withthe learner. Level2. The learneris able to notice the error, but cannot correct it, even with intervention. This indicatessome degree of development, but more importantly,even though the learner must rely heavilyon the tutor,in contrast to level 1,an opening is providedforthe tutorand the learner to begin negotiating the feedback process and forthe learnerto begin to progress towardself-regulation. The help required tends to be towardthelower,explicit,end of the regulatoryscale given in Figure I below. Level3. The learner is able to notice and correct an error,but only under other-regulation. The learner understands the tutor's intervention and is able to react to thefeedbackoffered. The levels of help needed to correct the error move towardthe strategic,implicit,end of the regulatoryscale. Level 4. The learner notices and corrects an error with minimal, or no obvious feedback from the tutor and begins to assume full refor error correction. However,desponsibility intramental, velopmenthas not yetbecome fully since the learner often produces the target and maystillneed the tutorto formincorrectly confirm the adequacy of the correction. The learnermayeven rejectfeedbackfromthe tutor when it is unsolicited (e.g., "Let me see if I can do it alone"). Level5. The learnerbecomes more consistent in all conin using the targetstructure correctly texts.In most cases, the individual'suse of the correct targetform is automatized. Whenever noaberrantperformancedoes arise, however, ticing and correctingof errors do not require fromsomeone else. Thus, the indiintervention vidual is fullyself-regulated. The five transitionallevels represent,then, three general stages of development.The first stage, encompassing levels 1 through3, represents other-regulation in which the learner mustrelyin some way on another individualin order to perform.Withouthelp fromsomeone else, the individual is not able to notice or correct his or her errors.The next stage is partial self-regulation,encompassing level 4. At this capable of detectingand stage learnersare fully correctingtheir own mistakeswithoutoutside feedback; their performance, however,is not automatized. The third,and final developmental stage, is that in which the learners' performance, including correctivebehavior, is com-
andJames P Lantolf Ali Aljaafreh and automatized and mispletelyself-generated takes emanate from legitimate slips of the tongue, or the pen, rather than from incomplete learning." Figure I presentsa listingof the levelsof help, or regulation, thatwere identifiedin the analthatoccurred during the ysisof the interactions tutorialsessions. The levels are arranged from what we consider to be the most indirect,or implicit(lower numbers),to the most direct,or explicit (the higher numbers). We point out that the levels were not determinedin advance of the study.Moreover,except for level 0, they were not rigidlyfollowed in everycase. In any particular instance of corrective intervention, the collaborativeworkof both participantsdeterminesthe level of assistance to be invoked,if one or more levelswillbe skipped,whereto stop and allow the learner to assume responsibility, and when to withholdassistance. Thus, microgenetic development,according to our second criterion,is evidenced wheneverthe negotiated feedback moves fromthe bottom to the top of the regulatoryhierarchy.12 DATAANALYSIS Before considering protocols that illustrate how microgenesis arises through intervention in the ZPD, we willlook at some protocolswhich demonstratehow feedback is negotiated in the ZPD in terms of the regulatoryscale given in Figure I. Errors for which implicit strategic feedback proved to be effective are considered to be high in ZPD (e.g., level 1, 2, 3), since the learner is close to independent performance, while those that require explicit feedback are said to be low in the ZPD (levels 10, 11,12), because the learner is further awayfromproducing the correctformwithouthelp. The Collaborative Frame as a Source of Feedback.Levels 1 and 0 on the scale entail the same surfacebehavior; thatis, reading of the essayby the learner. At a more abstractlevel, however, they are quite different-the differencebeing that the reading at level 1 takes place when the tutorand the learner are in a collaborativeposture,while at Level 0 the expectation is thatthe learner is to relyon herself.Level 1, therefore, marks the beginning of the collaborative interaction, while level 0 is outside of the collaborative frame. A shiftin orientation thus brings withit a source of regulation thatwas not available to the learner before. The presence of a dialogic partner helps the learner in a subtle, though significant, way.It representsthe mini-
471 FIGURE I to Explicit Regulatory Scale-Implicit (strategic) 0. Tutorasksthelearnerto read,findtheerrors, and correctthemindependently, prior to the tutorial. 1. Construction of a "collaborative frame" promptedby the presence of the tutoras a potentialdialogicpartner. 2. Prompted or focusedreadingof the sentence thatcontainsthe errorby the learneror the tutor. 3. Tutorindicatesthatsomething maybe wrong in a segment(e.g.,sentence, clause, line)-"Is thereanything wrongin thissentence?" 4. Tutor rejectsunsuccessful at recogattempts nizingtheerror. 5. Tutornarrows down the locationof the error (e.g., tutorrepeats or points to the specific whichcontainsthe error). segment 6. Tutorindicates thenature oftheerror, butdoes notidentify theerror(e.g.,"Thereis something withthetensemarking here"). wrong 7. Tutoridentifies the error ("You can't use an here"). auxiliary 8. Tutorrejects learner's unsuccessful at attempts theerror. correcting 9. Tutorprovides clues to help thelearnerarrive at the correct form(e.g., "It is not reallypast but some thingthatis stillgoing on"). 10. Tutorprovidesthe correct form. 11. Tutorprovides someexplanation foruse ofthe correct form. 12. Tutorprovides of thecorrect examples pattern when otherformsof help fail to produce an action. appropriate responsive mal formof otherderivedhelp available to the learner in the activityof error correction. In point of fact,however,even though Level 0 is be connoncollaborative,it can not legitimately strued as nonsocial activity. It is clearly social, since, for one thing,the learner undertakes to correct her composition at the request of the tutor; thus, this activityis situated within the larger event of error correction,which begins when the student enters the tutor's office.'3 Thus, even thoughthetutorand student maybe of doing different thingsat Level 0, the activity searching for errors is still social, but it need not be collaborative.We examine protocol (A), as an illustrationof how this strategicformof regulation functions. (A) N3* 1. N: "It's a littledifficult forme" 2. T(utor): uhum
472 3. N: 4. T: 5. N: 6. T: 7. N: 8. T: 9. N: 10. T: 11. N: 12. T: 13. N: 14. T: 15. N: 16. T: 17. N: 18. T: 19. N: 20. T: 21. N: 22. T: 23. N: 24. Is good? It's good To tell you uhum Or tell to you said to you To tell to tell you is fine To tell you where or what ... Or ah okay to tell you "Where or whatwill I do ten years later" Okay Okay? That's good "But I will tryexplain to you" ... Tois righthere? aah, yeah Is right? Uhum, it's right Okay Yeah, "I will tryto .. ." Okay, "to explain ..." To explain "To explain to you something about [I prefer"about" no] ofmy inquietudes, about some inquit ...
TheModern 78 (1994) Language Journal sumed the role of dialogic partner, the learner's orientationtowardthe taskof findingerrorsin her essay changed. This is important, because, although it may seem as if the learner continued to act alone, she did not. We referto the help triggeredby the dialogic presence of another,more expert,individualas thecollaborative The collaborativeframeseems to mark a frame. situationas one in whichcorrectionis to occur, even prior to anyovertmove on the part of the tutor,and thus representsthe miminal level of contingenthelp available to the learner in the ZPD. The learners' utterancesduring the tutorial interactionprovide clear signs of the effectof the collaborativeframeas a source of implicit correctivefeedback. In Protocol (A), forexample, the learner clearlyseems to recognize the potential that the tutor'spresence has for testing and confirmingher hypotheses.Although withthe indirect uses the verb tell she correctly object in line 5, in line 7 she volunteersanother way of using the verb as well as an alternative construction with said. This is significantbecause in earlier tutorialsshe had received feedback regarding her incorrect omission of the preposition to with the verb said and similar verbs before indirectobjects as in "I have said you." She offersthe correct form of the conwhere she struction juxtaposed to the verb tell, in and reforto be testing engaged appears about the use of these mulatingher hypothesis in line verbs and the preposition. Importantly, has what she to she is able (15) generalize learned in the earlier interaction to the verb and She is not fully confident,however, explain. asks for confirmationfrom the tutor (lines 15 thislearner showsan orienand 17). Generally, difthatis markedly tation toward joint activity fromworkingalone, or even withthe tuferent tor physically present but not part of the collaborativedyad. Some learners,in fact,openlycommentedon on to notice errorswhenworking theirinability theirown, as protocol (B) illustrates. (B) Y1 1. T: Yeah, you spent time withus. Okay "and I passed over yearwith 2. Okay,here, is there myfamily." wrong here ? "and" anything 3. passed..,. over... year.., with myfamily." 4. Y: (verysoftly)passed over a ? 5. T: Okay,the article... 6. Y: uha I forgetthis mmm
inquit..." 25. T: Okay,what is inquietudes? 26. N: I thinkthis is Spanish (laughs) * The capital letter is the learner's initial (N=Spanish L1; Y=JapaneseL1; F=Portuguese Li). The number following refers to the tutorial session from which the protocol is taken. The textin quotes indicatesreading of the essay. In protocol (A), the learner, with selfinitiation and self-correction, replaces the preposition of which she originallyused, with the more appropriate "about" in line 23. She also offersa number of clarificationquestions checksin lines 3, 7,13,15,and and confirmation 17. She had two prior opportunities to detect errorsin her essay,once on reading it prior to coming to the tutor'sofficeand once on reading it in the office before the tutorial began. When asked if she had found any errors,she replied that she had not. The question, then, is what triggersher atTo be sure, theyare temptsat self-correction? initiated by the learner, butin the presence of the expert tutor.When the learner read and searched forerrorsin her essayon her own, the tutorwas busyinghimselfwith somethingelse and was, therefore, ostensiblyunavailable as a collaborator.When the two came togetherinto a social configuration in which the tutor as-
P Lantolf andJames AliAljaafreh 7. T: 8. Y: Yeah. "pass over a year." Okay. It's But I don't notice by myself
473 (C) and (D) entail the same errormade by two different learnersregardinguse of the definite article with USA. (C) NI 1. T: okay in this,okay,"AlthoughI was preparing mytravelto 2. USA, withsome time almost always we have some thingto do 3. in the last." Do you ... is there ... do you see any thing 4. wrong here in this line here ? "Although I was preparing 5. myself" 6. N: I don't know! 7. T: Okay, "AlthoughI was preparing my ... travelto USA" okay aah 8. N: long travel 9. T: Okay,you say "preparing my..." instead of travel. . . what's 10. a betterword to use ? 11. N: Trip 12. T: Okay 13. N: Is bettertrip ? 14. T: Okay. Yeah "preparing mytrip," okay.There is also something 15. wrongwith the article here. Do you know articles ? 16. N: Articles,yes 17. T: Yeah so what's ... 18. N: eeh on mytrip to ... 19. T: What is the correct article to use here ? 20. N: Isn't tois ... no ... eeh ... article? 21. T: What is the articlethatwe should ... 22. N: It 23. T: No. Article ... you know the articles like the or a or an 24. N: The trip ... my,is not my? no ... the trip ? 25. T: My... yeah it's okay, you saymytrip 26. N: My trip 27. T: Okay 28. N: To United States 29. T: Yeah USA, what article we need to use withUSA ? 30. N: a, an, the 31. T: the, which one ? 32. N: but the ? 33. T: Okay,do we use the ... ah preparUSA? ing mytrip to . . . the 34. N: aaah ah (utterssomethingin Spanish) ah okay when I use when I use USA use witharticle 35. T: Okay 36. N: The
Firstof all, in protocol (B) the learner is able to find and correctthe erroron the basis of the level 3 help provided by the tutor. In line 8, she openly states that she was unimportantly, able to notice the error by herselfduring the prereading.We believe her use of "but" is especially revealingbecause it may indicate a sense of frustration at not being able to detect the mistakeon her own, even though she had been instructedto do so. She clearlyseems explicitly to recognize the differencebetween the two activities-the individual working alone and the individualworking jointly in the collaborative frame-and furtherrealizes that, at this point at least, she is not able to provide correctivefeedbackforherself. The factthatthe tutor is notjust presentin the room, but is acting as a collaborator in the correctionprocess,compels the learner to orient to the activity differently, thus enabling her,with some help, to detect a featureof the L2 that she had taken in earlier but had not yet fullyappropriated. Thus, in some cases-and as faras we can tell these are not predictable in advance-thesimple act of establishing the collaborative frame is an effective formof other-regulation.14 ZPDsfor Learners and Different Different Different An important Structures. dimensionof the negotiation of feedback and microgenesis in the ZPD for which we have uncovered clear evidence is that different learners often have differentZPDs for the same targetlanguage form and will therefore require differentlevels of help. It is importantto rememberthatall three learners under consideration in this studyhad been placed into the same class on the basis of a placementexam. This is an important point because, as the reader will recall fromour earlier discussion of Vygotsky'sformulation of the ZPD, no matter how sophisticated our assessmentinstruments maybe, we cannot arbitrarily assume thatany two learnerswho attain identical scores on a test are necessarilyat the same stage in their interlanguagegrowth,if all that we assess is theiractual developmentallevel.It is imperativeto assess the learners' potentiallevel of developmentas well. The examples that follow illustratehow the same errormade bydifferent learners,more often than not, representsdifferent problemsfor each learner,and consequentlyrequires different levelsof regulationfromthe tutor. Protocols
474 (D) Fl 1. T: 2. 3. 4. F: 5. T: 6. F: 7. T: "In the same day I mailed them ... to .. ." okay alright.What about also ... is there some thing else still in this sentence ? To the hum ? The Okay, "to the" ... yeah, "to the
78 (1994) TheModern Journal Language notice her omission of the infinitive particle to. In thiscase, the collaborativeframe,whichwas in triggering a responsiveaction from effective the same learnerin protocol (A) forthe prepohere when tofunctions sition to,is not effective as a particle. Therefore, the tutor is actively drawn into the correction process. His initial as he sigattemptat providinghelp is strategic, nals throughhis intonationalone (line 2), that something is amiss. The learner immediately locates the apparent source of trouble in line 3 and at the tutor'srecommendation,she rereads the phrase and incorporates the correction. The tutorthen confirmsthe correction. Finally,we consider protocol (F), in which the tutorattemptsto elicit from learner Y the missingpreposition toin a locative clause. (F) Y3 I will studyin Boston 1. T: Okay. "After for nine months,I'll return 2. mycountry."What do you mean "after"here ? Do you mean to previous afterthis (referring 3. or after ... paragraph) 4. you studynine monthsyou go back ? 5. Y: Yes, afternine monthsI mean 6. T: Uhum 7. Y: Afternine months 8. T: Afternine monthsyou go ... 9. Y: "I'll back mycountry" 10. T: You will back 11. Y: "I will be back mycountry..." I will studyin Boston 12. T: Okay,"After for nine months [ah ... 13. (softly)] nine months,I'll return mycountry."Okay,what 14. is ... do you think... is there anythingmissinghere ? "I'll returnmycountry..." 15. 16. Y: Return to? 17. T: Okay In (F), the sentence containing the error is read a numberof times.On some of these readings, the location of the error is narrowed by focusingon the phrase wherethe prepositionis expected to appear; yet,the learner stillfailsto notice thatsomethingis missing.Onlywhen the tutorexplicitlypoints out the precise location of the problem and asks ifsomethingis missing (lines 14 and 15) is a responsiveaction fromthe learner triggered(line 16). It is evidentthatthe subtle sources of feedback thatwere successful withlearner N in protocol (E) are not helpful for learner Y. She clearly needs more explicit
US."
On the face of it, both N and F produced the same error in their respective compositions: omissionof the articlewithUS. When we examine the interactionthat transpiredin the two a more informative picture protocols,however, emerges. While in (C) a wide range of help is necessarybeforethe learnereven begins to realize what the problem is, in (D) simplyreading the sentence containing the error (line 1) and suggestingthat somethingis wrong (lines 2 to 3) is sufficientfor the learner to notice and correctthe mistake. In termsof the regulatoryscale, for learner N, all levels of feedback are used, but for learner F, only levels 1 and 2 are needed. Thus, ZPDs the same featurerepresentstwo different forthe twolearners.In the case of F,the feature is high in the ZPD and the learner is veryclose to being able to control the featureby herself. In the case of (N), on the otherhand, the same featureis low in the ZPD and prospectsare that she will continue to need fairlyexplicit help. Hence, we cannot assume that the errorrepresents the same problem for each learner, because the learnerseach produce itfroma different location in the ZPD. The protocols given in (E) and (F) further our point with regard to the imporexemplify tance of linkingappropriateformsof corrective feedback to the individual learner's responsive moves. (E) N3 1. N: Okay... 2. T: 3. N: 4. T: 5. N: 6. T: 7. N: "I would like spend in
Okay ? Spend... Read again uhum " I would like to spend" Okay,you're missingtohere "To spend in United States two or threeyears."
Upon reading her composition at the outset of the interaction(line 1), the learner does not
P Lantolf andJames Ali Aljaafreh help to correct the error,which comes in the formof a narrowingdown of the location of the errorand explicitly indicatingthatsomethingis missing. in the ZPD. Now that we have Microgenesis some feel forhow feedback is negotiated in the ZPD, we can consider how developmentarises as a resultof correctivehelp. To do thiswe will presentsome sample protocols which illustrate the impact of feedback on microgenesiswithin a particular tutorial,as well as across tutorials. In each case we will be looking for changes from intermental,or other-regulatedperforor self-regulated mance, to intramental, performance, as well as forthe learners' abilityto generalize what theyappropriate in one linguistic context to other relevant contexts. We begin with protocol (G), which elucidates the microgenetic process for the modal can across episodes in the same tutorial. (G) N3 1. T: "To Germany."Do you see anything also wrong here ? "myfutureis can go to Germany" ... What 2. about the use of the auxiliary 3. verb here ? 4. N: Is... is... 5. T: Is can go ? 6. N: Is can go 7. T: Do you see somethingwrong here ? How to say it ? 8. N: No, I don't know 9. N: Okay,how how to use ... 10. N: Is will go 11. T: "One of mydreams for myfuture is . .." (rising intonation) 12. N: Will go? 13. T: No (lengthened vowel) ... 14. N: No 15. T: Okay,is... what...? 16. N: Is... 17. T: To go 18. N: To go not "can" ? 19. T: Yeah, because you have here, like ... this is an auxiliaryand thisis anotherauxiliaryor modal ... 20. 21. N: Yeah 22. T: So you have them together... 23. N: Yes, because I ... the verb form and two verbs together, yes. 24. T: Yeah, so yeah two verbs together. So... 25. N: I know 26. T: One of my... is to go to Germany 27. N: Oh myGod ! (laughs) 28. T: 29. N: 30. T: 31. N: 32. T: 33. N: 34. T: 35. N: 36. T: 37. N: 38. T: 39. N: 40.
475 Okay, "One of mydreams for my futureis to go ..." To go to Germany To Germanyand ... One we can't do right (laughs) No, that'sfine. You're doing fine Filine ? (laughs) Yeah No, please, thisis verybad (laughs). No very... Yeah, but I mean you have done a lot of... This is ... (laughs) Work,you know.Other thingsare gettingbetter,so ... And a lot of work,no, I don't like this.WhyI writebull shit ? I don't like it (laughs)
In protocol (G), the tutoruses severallevelsof implicitregulation-he reads the relevantsentence, hintsat the nature of the error,specifies its location, identifiesits nature,and rejectsinappropriate attemptsby the learner to correct it-but is still unable to triggeran adequate responsivemove fromthe learner.In line 8 the learner overtlystates that she does not know how to correctthe error,at which point the tutor begins to provide more explicit help and finally,in line 17, casts the verb phrase in its correctform.The learnerthen finally responds to the tutor'smoves and theyboth engage in a dialogue about the construction, which the learnerturnsinto an opportunity to expressher frustration her failure to regarding appropriate the feedback. A short time later in the same tutorial,the same problemwas encountered in the sentence "Anotherdream mine is can go to Japan," but thistime the learner'sresponsiveaction is quite differentand shows a shift toward self-regulation, as seen in protocol (H). (H) N3 1. N: "Anotherdream mine is" ... ah ah amm ... what?I can change now. 2. 3. T: Okay 4. N: Okay. "Anotherdream mine is ... is to go" again 5. T: Okay "is to go..." 6. N: "Is to go 7. T: Okay, "Anotherdream of mine is:" ... instead of can, "to go is to 8. go" 9. N: "is to go to Japan. I thinkJapanis an interesting countryin
476 10. culture,metho... methodologia methodology"
TheModern 78 (1994) Language Journal 14. N: Because is ... ifwe have good ... because is no is no only one. 15. T: Okay 16. N: Is general ... 17. T: Okay 18. N: idea... 19. T: Yeah 20. N: general good comprehension 21. T: Yeah, so it's not also count 22. N: Yeah 23. T: It's mass 24. N: It's mass 25. T: Comprehension.You can't say one comprehension,two comprehensions 26. N: Okay In (I) the learneruses the indefinitearticle a with the mass noun compreheninappropriately sion.She relies on the tutor'sfeedbackto generate the correct form of the construction.The tutordoes this withouthaving to resort to explicit levels of help, indicatingthatthe general location of the error is sufficient(line 6). At thispoint,the learner'sresponsiveaction is triggered and she is able to assume responsibility forcorrectingthe error.It seems clear,however, that withoutthe tutor's implicitlevel of interventionthe learnerwould not have noticed the error.Again, in termsof our criterionfordevelopment,we can say thatthe learneris probably at Level 3, since she does notice and correcther error,but only as a result of the tutor's intervention. Be that as it may,even though N is able to locate and correcther errorwithimplicithelp, the tutorapparentlyfeels the need to move to an explicit level of help beginning in line (8) theremainderof the and extendingthroughout protocol. It could be argued that the tutor engages in such a move in order to determineif the learnerunderstood the generalizationat issue and was not just operating locally. Of course, there are otherwaysof making thisdetermination,such as waiting until the pattern recurs in futureperformanceand observingif rendered or if the learner is able it is correctly to rectify anyerroneous patternsthatmayarise. On the other hand, the tutor'smove could simplyhave been a waste of time,or even worse,it might have resulted in confusion and the jettisoning of a perfectlyappropriate hypothesis on article use. We now consider the protocol given in (J), taken froma later episode during the same tutorial session.
In (H) the learnerdisplays evidence of approof in the the earlierepisode priation help given presentedin (G). We notice a markedreduction in the amount of help needed by the learnerto take over and complete the correction.The tutor,in fact,does not have to provideanyhelp for thelearner, who is able to self-initiate (line 1) and self-correct the error(line 4), simply on thebasis of reading the sentencein the tutor'spresence. The learner's metacomments, in our view,provide additional evidence of movementtoward In line (1), forinstance,she asks self-regulation. herselfa question about what she has written ("What?"), which she then proceeds to immediatelyanswer ("I can change now"), indicating form. thatshe knowsthecorrect She also appears to indicate,by her utterance"again" at the end of line (4), recognitionthather errorhere is the same as her earlier mistake discussed in (G). These utterances,in fact, appear to be private whose functionis not to commuspeech--speech nicate withsomeone else but to assistthe selfin situations(37; 40; 41). In terms problem-solving of our criterion, learnerN showssignsof microfromLevel 1 to Level 3, or genetic development perhaps even Level 4. To observe the effects of help on microwe will consider protogenesis across tutorials, cols (I) through(K) as the same learnerstruggles witharticlesand mass nouns. (I) N5 1. T: Okay, "when we read and thinkwe can writeday afterday [okay, 2. right] day afterday better.[Okay] is we have a good 3. comprehension." Do you see anythinghere ? "If we have ... a good comprehension" . . . 4. 5. N: A good comp ... rehension 6. T: uhum, "if we have .. ." Do you see withthe article ? anything 7. N: "If ifwe have ... good comprehension" 8. T: Okay, "good comprehension." You knowwhy? Whywe are not using a? 9. 10. N: When we read... when we... if we have... 11. T: "If we have good comprehension" 12. N: aah because ... wh whydon't we ... 13. T: yeah, whydon't we use a ? Whyit's wrong to use a?
P Lantolf andJames Ali Aljaafreh
477 6. T: 7. N: 8. T: 9. N: 10. T: 11. N: 12. T: 13. N: 14. T: 15. N: 16. T: 17. N: 18. T: 19. N: 20. T: 21. N: 22. T: 23. N: 24. T: 25. N: 26. T: 27. N: But what do you see wrong in these two sentences ... Ah just a moment. "We can ... see we can ., we can ... see" Uhum It... grey Okay Big Okay,greybig Layers Layers Layers in the sky Uhum Because is no one only,is all the Layers,it is not singular.Right, that's good Greybig layers... yes (laughs) In the sky With ... dense Okay (Laughs) Dense, that'sgood Dense smoke With dense smog "Produced by carbon monoxide of the the vehicle."
(J) N5
1. T:
aah and then you can saywhat ? ... "the most import... is ... 2. N: (softly)thing 3. T: is to thinkin the foreignlanguage," right? "But but [what?] be4. make compositions is difficult cause [let me see] . . . 5. because you need to have a good
6. N: (N crosses out the article a) 7. T: Okay,no a 8. N: (laughs) 9. T: "because you need to have good grammar, punctuation and... ? 10. perfectly.. ." perfectly In the above episode, the tutorreads the portion of textcontaining a targeterror and then pauses just before externalizingthe noun. The corrective reading and the pause are sufficient to elicit the help appropriate responsiveaction fromthe learner,who immediately proceeds to the tucross out the indefinitearticle.Actually, tor's pause and N's responsiveaction in line 6 occur almost simultaneously. Given that the tutor's regulation is situated at the upper end of the RegulatoryScale, it appears thatthe learner is close to being able to provide correctivefeedback forherselfand is, thus,developing toward The greater independence, or self-regulation. learner's laughter in line 8 suggests that she is conscious of the error, and that she is consciouslyprovidingthe correctanswer.This level of self-awareness in the process of feedback and error correction is characteristicof the transition fromother-to self-regulated performance (16; 33). it seems thatN is able to generFurthermore, alize withregard to the nonuse of articleswith mass nouns, on the basis of the feedback given at the outset of the tutorial,protocol (I), to a later point in the same tutorial,protocol (J). extends the help given in Hence, she correctly the case of comprehension to good Further grammar. extension is observed in the protocol given in (K) taken fromthe final tutorialfor N, which was conducted one week later. (K) N6 1. T: "We can see a greybig layersin the skywitha dense smog" What is ... do you 2. see anything wrong here ? 3. N: Dense smog withah heavy or... 4. T: That's fine,yeah this is good 5. N: This is good?
The learneris immediately able to correcther misuse of the indefinitearticle with the mass in line 1, thusprovidingevidence of noun smog generalization of feedback across tutorials.Of is whatwe observe in lines 6 even more interest and 7, where the learner overtly the interrupts tutor's utteranceand subsequentlyinhibitshis attemptto offerassistance. In so doing, she assumes fullerresponsibility for findingand corin the "a error recting grey big layers.'"15 She does this by externalizingher own corrective feedback process beginning in line 7 and ending in line 9 with the correct form,"grey."Finally,in line 17 she externalizesher understanding that the article cannot be used, in thiscase because the noun withwhichit co-occurs, layers, is singular.Here then we see evidence of psychological, as well as linguistic, development and N shows that she has formeda generalization withregard to nonuse of indefinitearticles with mass nouns and with plural count nouns. Moreover,she curbs the offerof help fromthe tutorand resolves the problem with only minimal evaluativefeedback fromthe tutor. We will consider three final protocols, in whichwe observe interactionbetween the feedback provided by the tutor and the learner's attemptto generalize across episodes as well as
478 tutorials.While the learner,in essence, has difficultiesfullyappropriatingthe help, and thus ultimatelyfails to generalize across contexts, the interaction between the tutor and the learner stimulatedby the failure is quite informative with regard to how generalization occurs. As Vygotsky(39) points out, we often learn more about how a cognitivesystemoperates whenwe observe itunder conditionsoffailure and breakdown than when we observe the In the protocols, systemfunctioningsmoothly. taken fromdifferent episodes of the same tutorial and fromdifferent tutorials,the tutorand learner attemptto work out the correct tense formodal + main verb constructions. markings (L) Fl 1. T: 2. 3. 4. 5. 6. 7. F: 8. T: 9. 10. 11. F: 12. T: 13. F: 14. T: 15. 16. F: 17. T: 18. 19. 20. 21. 22. F: 23. T: 24. F: 25. T: Okay, "to the ... [yeah] to the US. [Okay] In thatmomentI can't ... lived in the house because I didn't have any furniture." Is that ... whatwhat is wrongwith that sentence,too ? What is wrongwith the sentence we just read ? ... "In that momentI can't lived in the house because I didn't have any furniture"... Do you see ? No Okay... ah there is something wrongwith the verb with the verb tense in this thissentence and the modal ... Do you know modals ? Ah yes,I know Okay,so what's what's wrongwhat's wrong here ? The tense of thislive Okay,what about'the the ... is it just in thisor in this,the whole thing? The whole this Okay,how do you correctit? ... Okay, "In thatmoment" . . . What ? ... What is the past tense of can ? whatwas happening.., .what... the past, right? whatwas happening ... what . .. the eventhappened in the past right? so what is the past tense of thisverb can ? ... Do you know ? No Okay,ah could Ah yes Okay, "I could not..." 26. F: 27. T: 28. F: 29. T: 30. F:
TheModern 78 (1994) Journal Language Live Ah exactly, okay.So when you use this in the past then the second verb is the simple ... yes Form,okay ... aah "in thatmoment I could not .. ." Live in the house
In (L), the learnerincorrectly marksthe main verb, ratherthan the modal, for tense. Notice, marks tense in the however,that she correctly case of auxiliarydo + main verb in line 2. The tutorfirst asksifthe learnercan locate her error on simply the basis of his reading the sentence (lines 3 through6), whereupon the learner responds in line 7 that she cannot. In lines 8 to the modal referring through10,by explicitly about modals and askingifshe knowssomething (to which he receives a positive response), the tutorsubsequently narrowsF's search space. He with then again asks whatis wrong [presumably the tense of the modal]. The learner,however, responds by correctingthe tense of the main verb. The tutoraccepts the response, but then refocuses the learner's attention on the full modal + verb constructionand asks a series of questions designed to elicit the correct past tense form of the modal. Finally,the learner replies thatshe does not knowthe past tense of can,and so as a resultthe tutormustprovide it for her. She does appear, at least, to recognize the correctform,in line 24. The tutorpresents her with an explicit rule for tense markingin modal clauses, whichF seems to understand,as in line 28. In this indicated by her interruption of of lack because a interaction, responsiveness on the part of the learner,the tutoroffers very explicit levels of correctivefeedback. The correct tense markingis jointly constructed,but here fallson the tutor. mostof the responsibility The learner,with implicithelp fromthe tutor, recognizes thatthe main verb is not markedfor tense (lines 13 and 30), but she is unable to do about generatingthe correctformof anything the modal, even with explicit help. Thus, the tutor is forced to provide the correct form. Later in the same tutorial,the same problem resurfaces,as seen in protocol (M). (M) Fl 1. T: Okay, "I called other friendswho can't went do the party."Okay, what is wrong here ? 2. 3. F: To 4. T: "Who can't wentdo the partybecause that nighttheyworked at
andJames Ali Aljaafreh P. Lantolf 5. 6. 7. F: 8. T: 9. 10. F: 11. T: 12. F: 13. T: 14. F: 15. T: 16. F: 17. T: the hospital." Okay,fromhere "I called other friendswho can't went do the party."What's wrong in this? To ? Okay,what else ? . . . what about the verb and the tense ? The verb and the tense ... Could Okay,here Past tense All right,okay,"who [alright] could not." Alright? And ? ... To Here [points to the verb phrase], what's the rightform? I... go Go. Okay, "could not go to [that's right] to the party..." 5. T: 6. F: 7. T: 8. F: 9. T: 10. F: 11. T: 12. F:
479 What the verb verb ... something wrongwith the verb ... Ah, yes... That you used. Okay,where ? Do you see it ? (Points to the verb) Took ? Okay Take Alright,take (Laughs)
In thisepisode the intervention of the tutoris considerablyreduced from the episode examined in protocol (L). The level of explicitness required to elicit an appropriate responsiveaction fromthe learneris also reduced. Nevertheless, the learner still has some problems focusing on the error indexed by the tutor. She initiallyassumes the problem to be related to her apparent slip of the pen, and she indicates that do should be rewritten as to (line 3). However,once the tutormanages to get her to focus on tense marking, in lines 8 and 9, she is immediatelyable to correct the modal, in line 10. In addition, she is also able to provide the correct formof the main verb,in line 16,althoughwith some regulationfromthe tutor. Above all, however, the learner accepts more of the responforcorrectingher errorin protocol (M) sibility than is the case in (L). A significant transition appears to have occurred between the firstand second episodes. The learner has appropriated the feedbackprovidedin (L) and showssignsof microgenetic development, given that in (M) she elicits feedback thatis less explicit and less extensive. We can now compare the same learner's performanceon the same structure one week later. (N) F2 1. T: Is there anything wrong here in this sentence ? "I took only Ani 2. because I couldn't took both" . . . Do you see anything 3. here "bewrong ? ... Particularly cause I couldn't took both" 4. F: Or Maki?
The learner continues to experience difficultywith the modal + verb construction.This, in itself,is not too surprising,since when the structurefirstappeared, it was low in F's ZPD, as indicatedbythe need forthe tutorto provide feedbackfromthe explicitlevelsof the scale. As we pointed out earlier,control over featuresof the second language does not move from a stage whereexplicitfeedback is required to full appropriation; ratherlearning evolves through stages of decreasing reliance on the other person towardincreasingreliance on the self. In protocol (N) we see evidence of microgenesis both in production of the modal + verb constructionand the extentof responsibility assumed by the learner forits production. In the earlier tutorials, the learnerwas unable to mark the modal fortense correctly withoutthe tutor's intervention. In (N), however, we observe that the correctformwas produced in the composiseveraldaysafterthe firsttutorial. tion, written F stilldoes not have full mastery over the strucmarksthe main verb ture,since she incorrectly for tense. The tutor focuses her attention on this problem (line 5) through a question, at which point she interrupts him (line 6) and indicates her recognitionof the error and points to its location (line 8). She then proceeds to provide the correction in line 10. Her laughter in line 12, as in our earlier discussion,is a clue that the formmay not yetbe fully automatized and thatthe errorwas not the resultof inadvertent performance.All of this suggests that the learner has appropriated the feedback offered in thefirst tutorialand thatshe now has greater control over the construction. Tracing the same modal + main verb construction in subsequent compositions for the same learnerrevealsevidence of fullappropriation. The learner used the constructionindefor example, in compendentlyand correctly, position three: "The wolfwanted to eat the goat but he couldnotaskhis friend..,. the wolfdrank a lot but he could not to eat the goat." Later in get the same composition,we observe: "I saw peo-
480 livewith a fewmoney." ple who people who could Finallyin composition seven we note: "Another seeis thereare lot of interest to take thingI could in moneyfrompoor people." Most importantly, the final composition,we see that she not only extended the pattern to other verbs but to anothermodal, as well: "If I have money, I would try hard to help poor people." The problem was a complex one in which the learner appears to have focused firston the modal (protocol L) and then on the main verb (protocols M and N), even though the tutor offered corrective feedback on both parts of the complex verb phrase at the same time. CONCLUSIONS Effective error correction and language learning depend crucially on mediation provided by other individuals,who in consortwith the learner dialogicallyco-constructa zone of proximal development in which feedback as regulation becomes relevantand can therefore be appropriated by learnersto modifytheirinFrom thisstance, learning terlanguagesystems. is not somethingan individualdoes alone, but is a collaborative endeavor necessarilyinvolving Here we encounterwhat,on otherindividuals.16 the face of it at least, looks like a potential problem. Van Lier (p. 211) worries--correctly we believe-that too much guidance, or in his words, "other-repair," mightinhibit,or at least whichhe retard,the developmentof self-repair, The views as an "importantlearning activity." in formalthe as interaction of ZPD, verygoal law of culturaldevelopment, ized in Vygotsky's is for novices to appropriate the responsibility for their own linguistic performance. This is what it means to move throughthe regulatory hierarchy.As implicit forms of feedback beand explicitformsbecome come more relevant, less relevantin regulating the novice's correctivebehavior,byimplication,novicesassume inin creased control over theirlinguisticactivity the L2. This is whywe argue that feedback as in the ZPD is not only graduother-regulation ated but is also contingent. At first,responforthe novice's linguisticperformanceis sibility distributedbetween the novice and the expert, with the expert having more control over this performance than the novice. Under the expert's guidance, control is graduallyapproprithe novice moves ated bythe novice. Eventually, awayfromreliance on the expert (other-repair, in van Lier's terms) towardreliance on the self.
TheModern 78 (1994) Journal Language For this to happen, however,the expert must be willing to relinquish control (itselfdialogicallynegotiated) to the novice at the appropriate time. There can be no real development otherwise. From thisperspective,the typesof errorcorrection (i.e., implicitor explicit) thatpromote learning cannot be determined independently of individuallearnersinteracting withotherindividuals. Moreover,all types of feedback are relevantforlearning,but theirrelepotentially vance depends on where in the learner's ZPD a particular propertyof the L2 is situated. The of regulationis an attemptto capture hierarchy this dynamic character of feedback. Developnot only mentin a second language is therefore reflected in the learner's abilityto generalize what had been appropriated, but is also revealed through the kind of help that is jointly negotiated between experts and novices. Thus, a learner who is able to produce a particular structureas a consequence of more strategic (i.e., implicit)formsof regulation (e.g., the colmore adlaborative frame) is developmentally vanced than one who needs direct and explicit feedback for the same property.This means that linguisticforms alone do not provide us withthe fullpictureof a learner's developmental level. It is essential to know the degree to or mediation, impacts which other-regulation, on the learner's production of the particular forms. We are, of course, not uncritical of the research laid out in thispaper. For one thing,we of this recognize that one of the shortcomings initial attemptto investigatefeedback and L2 learning in the ZPD is that the linguisticfeatures considered are "surfacy" and language specific. Schwartzand her colleagues (28; 29) contend thatcorrectionmayonly stimulatethe development of those properties of a second such as verb morphologyand language system, lexis, which do not arise from abstractprinciples specified by UG. The lattertypeof properties (e.g., syntacticmovement), according to Schwartz (28), will be unaffectedby negative feedback and can only develop if learners receive positive and contextuallyembedded inhowever,sugput.17 Carroll and Swain's study, gests that corrective feedback administered under controlled conditions can impact poson the developmentof more abstractsynitively tactic properties (i.e., dative alternation).It remains to be seen if collaborativeinteractionin the ZPD can also enhance the acquisition of abstractpropertiesof second languages. In this
P Lantolf andJames Ali Aljaafreh regard we intend to extend our work on the ZPD to two specific cases-unaccusative constructions in Spanish and word order in German. Another drawback of our study is that the in audio format, data were collected exclusively imwhich eliminatedfromanalysisa potentially Since portantsource of nonverbalinformation. interaction between individuals also entails a rich gestural component, future research on learning in the ZPD requires the analysis of video recordings to capture the meaning displayed by speakers on theirhands (25). Finally,it is not our intent to sanction the for tutorialas the uniquely endowed framework ZPD. The the tutorial format co-constructing represents only one means for realizing this process. It is necessaryto explore the full array of possibilitiesavailable in the classroomsetting for enhancing learning in the ZPD. Collaborative interaction between learners engaged in and diatasks,use of portfolios, problem-posing logue journals are among the other avenues through which a ZPD can be co-constructed and learning can emerge. We hope that the presentstudywill stimulateadditional research in these domains. ACKNOWLEDGEMENTS We would like to thank MerrillSwain forher insightful suggestionsand usefulnegativefeedback on an earlierversionof thispaper. We are also gratefulfor the helpful commentaryprovided byan anonymousreviewer. Of course, the authors accept full responsibility for any shortcomings in the presentversion of the paper. NOTES 1 Prior to the interestin error analysis,the of those general assumption working from a more behavioristic slant had been that errors resulted from insufficientlearning of target language rules (36: p. 181). 2 The inspiration for the distinctionwe are drawingbetween the ethnographicand experimental approaches to corrective feedback we owe to the commentsof an anonymousreviewer. SAlong similar lines, Sharwood Smith proposes thatcorrectivefeedback,or "negative input enhancement" serves to flag specific er-
481 rors,while elaboration has to do withthe manner in which the error is made salient. That is, some formsof elaboration are explicit and entail metalinguisticexplanations and othersare more indirect, involving only some means of wheretheerroroccurred but givingno marking specific informationon its nature. Sharwood Smith suggests that the various levels of input enhancement should be examined through controlled experimentation (31: p. 177), prelevel of sumablyto determinethe mosteffective input enhancement. charac4 Metaphoricallyspeaking, Vygotsky terizesactual developmentas the "fruits"of the maturationalprocess and the ZPD as representing the "buds" or "flowers"of thatprocess (39: p. 87). the kind of help provided is very 5 Actually, much influenced by sociocultural factors.Research has shown thatmothersfromrural economicallyunderdeveloped settingstend to be more directive in helping their children performtaskswhen compared to middle class, educated urban mothers, who are much more indirect and strategic in regulating the mental of theirchildren (42; 43). activity 6 Althoughthe fullstudyalso entailed a comparison of the performanceof learnerswho received corrective feedback in the ZPD with those who received either explicit or implicit correction,we concern ourselveshere withthe ZPD group only and we will not consider the results from the other two groups. We point out, however,that even though three different treatmentgroups participated in the study,it was not implemented as a controlled experiment. To do so, we believe, would not have allowed us to uncover the processes at work as learning emerged in the ZPD. 7 Even though we refer to learners who received tutorial help as the ZPD group, we not claiming that the only way to create a ZPD is through tutorials.There are types of interaction that can create a ZPD. For instance, research byDonato showsthatlearnersworking in collaboration can jointlyconstructa ZPD without intervention froma tutor. We chose the tutorial procedure, in this particular case, because it seemed to be the mostexpedientwayof the ZPD forpurposes of our study. constructing 8 The learners' use of text-basedproperties, such as cohesive devices, were not considered for our purposes since we were not directly interested in their abilityto write per se; rather our concern was withtheirgrammaticalcompetence as reflectedin the writtenmedium.
482
The Modern Language Journal 78 (1994)
9 The abilityof learnersto generalize on the basis of negativefeedback is important because it providesevidence thatthe learner's linguistic in fact, representationhas changed. Vygotsky, insisted that there can be no masteryin learning if learners cannot extend what they have learned in one context to new contexts. In a studyof L2 Frenchlearners,Carroll,Swain,and Roberge report that while negative feedback helped theirsubjectsreduce theirerrorsin a set of morphologicalendings,the learnerswere apunable to constructthe necessarygenparently eralizationson the basis of the correctivefeedback. In their study of dative alternation in English as a second language, Carroll and Swain (5) show thatlearnerscan indeed generalize fromnegative input. 10The data to be considered in the present instancesof all fivetranpaper do not illustrate sitional levels. For a full empirical account of the levels see Aljaafreh. 11Peter Coughlan (personal communication) proposes that we may not want to dismissperformance errors merelyas slips of the tongue and suggests that there may be some unconscious intentionunderlying such behaviors.He raises the question of whylearnersmake errors even after theyhave learned a particular feature of the L2. This question can also be extended to native speakers as well. 12The hierarchy of regulationgivenin Figure I is not intendedto be exhaustive;rather, it representsthe range of help offeredin the specific tutorialsstudied here. 13 We acknowledgePeter Coughlan forpointing thisout. 14 Of the three ZPD learners,N is the only one to have shown a responsive action at the withregard to the four level of collaborative frame At this point,we have no way targetstructures. of knowingthe potential source of this difference among the learners. 15Wertsch,and Wertschand Hickmann similarlyshow how children,throughtheirspeech, for their own assume increased responsibility activity. problem-solving 16 What this argues for is an "informant"based, rather than a "text" (i.e., data)-based, approach to second language learning (19), and supportsthe contentionthatlanguage learning does not take place inside of someone's head but arises in the interaction that is co-constructedbetween individuals.In the lattercase, only positiveevidence is required, while in the former,both positive evidence and negative feedback are necessary.Space does not permit
us to explore the full implications of the two positionsforsecond language learning. Suffice it to say,that second language researchersare not in agreement as to which approach correctly characterizes nonprimary language learning in adults. 17To provide the details of Schwartz'sinterclaim would take us esting,ifnot controversial, too far afield from our present purpose. We reader to examine her encourage the interested on the topic (28; 29). writings BIBLIOGRAPHY
1. Aljaafreh, Ali. Negative Feedback in SecondLanguage and the ZoneofProximal Diss., Learning Development. 2. Birdsong,David. Metalinguistic and InterPerformance Berlin: Springer-Verlag, linguistic Competence.
Univ. ofDelaware, 1992. Newark,
1989. 3. Burt, MarinaK. & CarolKiparsky. A TheGooficon:

MA: Newbury RepairManualforEnglish. Rowley,
House,1972. AnnL. Brown, 4. Campione, Roberta A. FerJoseph, rera& Nancy R. Bryant. "The ZoneofProximal DifforIndividual Development: Implications San Francisco: Rogoff& JamesV. Wertsch. 1984:77-92. Jossy-Bass, Susanne& Merrill 5. Carroll, Swain."Explicitand Implicit Negative Feedback: An Empirical oftheLearning ofLinguistic GeneralizaStudy
tions." Studies in Second 15 LanguageAcquisition ferences in Learning." Children's in the Learning "Zone of ProximalDevelopment." Ed. Barbara
6. -,
(1993):357-86. Merrill Swain& YvesRoberge."The Role inAdult ofFeedback SecondLanguage Acquisition: Error Correctionand Morphological
13 Generalizations." AppliedPsycholinguistics
Re7. Chaudron, Craig. SecondLanguage Classrooms: search on Teaching and Learning. Cambridge: Cam-
(1992):173-98.
8.
27 (1977): 29-46. guageLearning 9. Children's DevelopLearningin the "ZoneofProximal
1988. Press, bridgeUniv. -. "A Descriptive Model of Discoursein the Treatment ofLearners Errors." LanCorrective
Ed. Barbara & James ment." V.Wertsch, Rogoff San Francisco: 1984. Jossy-Bass, ErofLearners' 10.Corder, S. Pit."The Significance
Review 5 rors." International Linguistics ofApplied
(1967):161-70. AnnE. Chun& N. AnnChenoweth, 11.Day, Richard, StuartLuppescu. "CorrectiveFeedback in Native-Nonnative Discourse." Learning Language 34 (1984):19-45. ofError CorrecM. "The Effect Robert 12.DeKeyser, and OralProtionon L2 Grammar Knowledge
Ali Aljaaftehand JamesP Lantolf ficiency." ModernLanguageJournal77 (1993): 501-14. Ra13. Donato, Richard. Beyond Group: Psycholinguistic in SecondLanguage tionale Activity for Collective Diss., Univ.of Delaware, Newark,1988. Learning. on Second 14. Error Analysis: Perspectives Language Acquisition.Ed. Jack C. Richards, London: Longman, 1974. 15. Fanselow, JohnF "The Treatmentof Errorin Oral Work." Foreign LanguageAnnals10 (1977): 58393. 16. Frawley, William. "The Cross-language Study of Private Speech." Paper, First Conference for Socio-cultural Research. Madrid,June, 1992. 17. Gaskill,William H. "Correction in Native SpeakerNonnative Speaker Conversation." Discourse Ed. Diane in SecondLanguageResearch. Analysis Larsen-Freeman.Rowley, MA: NewburyHouse, 1980: 125-37. 18. George, H. V. Common Errors in LanguageLearning. MA: NewburyHouse, 1972. Rowley, 19. Gold, Edward M. "Language Identificationin the Limit." Information and Control 10 (1976): 467-74. 20. Hendrickson,James M. "Error Correction in Foreign Language Teaching: Recent Theory, Research, and Practice." Modern Language Journal 62 (1978): 387-97. 21. Herron, Carol & Michael Tomasello. "Learning Grammatical Structures in a Foreign Language: Modelling VersusFeedback." TheFrench Review 61 (1988): 910-23. 22. Kasper, Gabriele. "Repair in Foreign Language in Second 7 Teaching." Studies Language Acquisition (1985): 200-15. 23. Lightbown, Patsy M. & Nina Spada. "Focus-onForm and CorrectiveFeedback in Communicative Language Teaching: Effects on Second in Second Language Learning." Studies Language 12 (1990): 429-48. Acquisition 24. Long, Michael H. "Teacher Feedback on Learner Error: Mapping Cognition." On TESOL '77. and LearningEnglish as a Second Teaching Language: Trends in Research and Practice. Ed. H. Douglas Brown, Carlos A. Yorio & Ruth H. Crymes. Washington,DC: TESOL, 1977: 278-93. 25. McNeill, David. Hand and Mind. WhatGestures Reveal about Thought. Chicago: Univ. of Chicago Press, 1992. 26. Newman,Dennis, Peg Griffin & Michael Cole. The Construction in Zone: Working for Cognitive Change School. Cambridge: Cambridge Univ. Press, 1989. 27. Schachter, Jacquelyn. "Corrective Feedback in Historical Perspective."Second Research Language 7 (1991): 89-102. 28. Schwartz, Bonnie. "On Explicit and Negative Data Effecting and Affecting Competence and in Second Linguistic Behavior." Studies Language 15 (1993): 147-63. Acquisition 29. and Magda Gubala-Ryzak."Learnability &8& GrammarReorganizationin L2A: AgainstNega-
483 tive Evidence Causing the Unlearning of Verb Research 8 (1992): 1Movement."Second Language 38. 30. Sharwood-Smith, Michael M. "Input Enhancement in InstructedSLA. Theoretical Bases." Studies in 15 (1993): 165-79. Second Language Acquisition -. 31. "Speaking to Many Minds: On the Relevance of Different Types of Language Information for the L2 Learner." SecondLanguageResearch 7 (1991): 118-32. 32. Spada, Nina & PatsyM. Lightbown. "Instruction and the Development of Questions in L2 Classrooms." Studies in Second 15 LanguageAcquisition (1993): 205-24. 33. Tharp, Richard G. & Robert Gallimore. Rousing and Schooling Minds to Life: Teaching, in Learning Social Context. Cambridge: Cambridge Univ. Press, 1988. 34. Tomasello, Michael & Carol Herron. "Feedback for Language TransferErrors." Studies in Second 11 (1989): 384-95. Language Acquisition 35. & Carol Herron. "Down the Garden Path: Inducing and Correcting Overgeneralization Errorsin the Foreign Language Classroom. Ap9 (1988): 237-46. pliedPsycholinguistics 36. van Lier, Leo. The Classroom and the Language London: Longman, 1988. Learner. 37. Vygotsky, Lev S. Thought and Language. Cambridge: MIT Press, 1986. 38. -. "The Genesis of Higher Mental Functions." The Concept in Soviet ofActivity Psychology. Ed. James V. Wertsch.New York: M.E. Sharpe, 1981: 144-88. 39. -. Mindin Society: The Development of Higher PsyProcesses. chological Cambridge: Harvard Univ. Press, 1978. 40. Wertsch, James V. Voices of theMind. Cambridge: Harvard Univ. Press, 1991. 41. -. and theSocial Formation Vygotsky of Mind. Cambridge: Harvard Univ. Press,1985. . "The Zone of Proximal Development: 42. Some Conceptual Issues." Children's in Learning the Zone of ProximalDevelopment. Ed. Barbara Rogoff & James V. Wertsch. San Francisco: 1984: 7-18. Jossey-Bass, 43. & Maya Hickmann. "Problem Solving in Social Interaction: A Microgenetic Analysis." Social and Functional to Languageand Approaches Ed. Maya Hickmann. Orlando: AcaThought. demic Press, 1987: 251-66. 44. White, Lydia. "AdverbPlacement in Second Lanof Positiveand guage Acquisition: Some Effects Negative Evidence in the Classroom." Second 7 (1991): 133-61. LanguageResearch -. 45. Universal Grammar and Second LanguageAcAmsterdam: quisition. John Benjamins, 1989. 46. Zinchenko,V. P. "Vygotsky's Ideas about Unitsfor the Analysis of Mind." Culture, Communication and Cognition: Ed. JamesV. Vygotskian Perspectives. Wertsch. New York: Cambridge Univ. Press, 1985: 94-17.

Aljaafreh & Lantolf 1994 Negative Feedback As Regulation and Second Language Learning in The Zone of Proximal

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Aljaafreh & Lantolf 1994 Negative Feedback As Regulation and Second Language Learning in The Zone of Proximal

Uploaded by

Copyright:

Available Formats

Negative Feedback as Regulation and Second Language Learning in the Zone of Proximal Development Author(s): Ali Aljaafreh and

?1994The Modern LanguageJournal

476 10. culture,metho... methodologia methodology"

P Lantolf andJames Ali Aljaafreh

The Modern Language Journal 78 (1994)

Univ. ofDelaware, 1992. Newark,

1989. 3. Burt, MarinaK. & CarolKiparsky. A TheGooficon:

27 (1977): 29-46. guageLearning 9. Children's DevelopLearningin the "ZoneofProximal

You might also like