247TESOL QUARTERLY Vol 37 No 2 Summer 2003

Mediating Language LearningTeacher Interactions With ESL Studentsin a Content-Based ClassroomPAULINE GIBBONSUniversity of TechnologySydney Australia

This research draws on the constructs of mediation from socioculturaltheory and mode continuum from systemic functional linguistics toinvestigate how teacher-student talk in a content-based (science) class-room contributes to learnersrsquo language development The illustrativetexts show how two teachers through their interactions with studentsmediate between the studentsrsquo current linguistic levels in English andtheir commonsense understandings of science on the one hand andthe educational discourse and specialist understandings of the subjecton the other Through this mediation studentsrsquo contributions to thediscourse are progressively transformed across a mode continuum intothe specialist discourse of the school curriculum The data reveal waysteachers build linguistic bridges to span the two orders of discourse byshowing how the interactions provide sites for L2 learning in terms ofthe development of the new academic register The illustrative textssuggest that in interactions that are effective in terms of L2 develop-ment both teachers and learners are active participants in the co-construction of language and curriculum knowledge The article alsoargues for the value of qualitative interpretive approaches and groundedknowledge in L2 research that is concerned with teacher developmentand educational improvement

For students who are learning ESL in an English-medium schoolEnglish is both a target and a medium of education They are not

only learning English as a subject but are learning through it as well Inthese content-based classrooms the construction of curriculum knowl-edge needs to progress hand-in-hand with the development of EnglishThis article focuses on how this process occurred for 8- and 9-year-oldstudents in two mainstream science classrooms where for more than90 of the students English was a second (or subsequent) language (iethe teaching and learning of language took place in the context of the

248 TESOL QUARTERLY

regular school curriculum not in a separate ESL classroom) The articleexplores in particular the way teacher-student interactions can enable L2learning in the context of subject teaching and the role of the teacherand learners in this process The texts show language learning as asocially mediated process whereby both teachers and learners are activeparticipants in the co-construction of language and curriculum knowl-edge The article draws on the constructs of mediation and modecontinuum to provide a way of conceptualising how this process is playedout in the classroom

MEDIATION FROM A SOCIOCULTURAL PERSPECTIVE

The construct of mediation is central to the sociocultural theory ofVygotsky (1981 1986 Lantolf 2000) because it provides a means ofstudying social processes involved in situated language learning and use(Appel amp Lantolf 1994 Hall 1995 Kramsch 1993 Moll 1994 Toohey2000 Wells amp Chang-Wells 1992) Re ecting Vygotskyrsquos notion thatlearning originates in the social mediation provided by interactions suchresearch in second language acquisition (SLA) questions the metaphorsof input and output (see eg Swain 2000 van Lier 2000) Socioculturaltheory views language learners not as processors of input or producers ofoutput but as ldquospeakershearers involved in developmental processeswhich are realised in interactionrdquo (Ohta 2000 p 51) Recent research(Donato 1994 2000 Engerstrom amp Middleton 1996 Ohta 1995 19992000 Swain 2000 Swain amp Lapkin 1995 Wells 1999) has shown howlearning and language acquisition are realised through a collaborativeinteractional process in which learners begin to appropriate the lan-guage of the interaction for their own purposes Much of this researchhas been concerned with learning through peer-peer interactions ratherthan through interactions between expert and novice whereas this studyexamines how language learning is mediated by language use in thecollaborative interactions between teacher and students

Mediation is central to the study of collaborative interactions Vygotsky(1981) argues that human activities and mental functioning are medi-ated and facilitated by tools cultural practices and artefacts the mostextensive tool being language Mediation is a familiar concept in manysocial contexts A lawyer for example mediates between a clientrsquosaccount of an event and the language principles and categories of thelegal world of the court (Maley Candlin Crichton amp Koster 1995)Mediation of this kind can broadly be described as occurring insituations characterised by difference dif culty or social distance(Baynham 1993) Such characteristics are inherent in most teacher-student relationships because in the great majority of school classrooms

MEDIATING LANGUAGE LEARNING 249

there is considerable linguistic and conceptual distance between teacherand students especially when they do not share the same languageassumptions and life experiences

Constructs related to mediation that are useful for investigatingcollaboration include zone of proximal development (ZPD) scaffolding jointconstruction and contingency The site where social forms of mediationdevelop is the ZPD (Lantolf 2000) The ZPD (Vygotsky 1978) refers tothe distance or the cognitive gap between what learners can do unaidedand what they can do in collaboration with a more competent other InVygotskian terms learning occurs through this assisted performance andin the context of joint activity Human development including languagedevelopment is thus intrinsically a social process and in the broadestsense educational Newman Grif n and Cole (1989) who call the ZPDthe construction zone de ne it as ldquothe changes that take place in sociallymediated instructionsrdquo (p 2)

The metaphor of scaffolding has been used by mother tongue and L2educators to describe the nature of this assisted performance whichinvolves not simply helping to do but helping to know how to do(Brooks 1992 Cazden 1988 Lee amp Smogarinsky 2000 Maybin Merceramp Stierer 1992 Mercer 1994 Webster Beveridge amp Reid 1996 Wells1999) The term was originally used by Wood Bruner and Ross (1976) intheir examination of parent tutoring in the early years In the classroomit can be de ned as the temporary but essential assistance that helpsapprentice learners into new skills concepts or levels of understanding(Maybin et al 1992 p 186) In relation to teacher-student interactionMaybin et al suggest two criteria for determining whether a particularexample of help can be portrayed as scaffolding There must be evidenceof a learnerrsquos successfully completing the task with the teacherrsquos help andevidence of the learnerrsquos having achieved a greater level of independentcompetence as a result of the experience For unlike the lawyerrsquos aimthe teacherrsquos aim is ultimately to hand over knowledge and control to thestudents In the sample texts discussed in this article this process occursthrough the interactional scaffolding that the teacher provides Thisstudy contributes to the ndings of other studies (eg McCormick ampDonato 2000) that focus on ldquothe interactional mechanisms involved inthe obtaining or providing of assistance during language learning tasksrdquo(Ohta 2000 p 52)

In this article I use the term mediation to characterise the linguistic anddiscourse choices made by teachers in the content-based ESL classroomThe texts illustrate how two teachers through their interactions withstudents mediate between the studentsrsquo current linguistic levels (inEnglish) and their commonsense understandings of science on the onehand and the educational discourse and specialist understandings of thesubject on the other In other words mediation involves communication

250 TESOL QUARTERLY

between two different orders of discourse the current levels of learnersrsquoknowledge and L2 abilities and the broader knowledge and specialistlanguage of the science community into which the students are beingapprenticed

MODE CONTINUUM FROM ASYSTEMIC FUNCTIONAL PERSPECTIVE

Describing mediation across orders of discourse however requires ameans of characterizing the different orders I use the construct of amode continuum to describe the different orders of discourse observed inthe classroom as the learners were assisted in moving from registersexpressing their rsthand experience in oral language to those express-ing academic knowledge in writing A description of such registers drawson systemic functional grammar following the work of Halliday andother systemicists because of the need to study the discourse of teachingand learning from the perspective of linguistic theory that explains howlanguage makes meaning Theories of language and language learningthat ignore its social context of use are of little help in an analysis ofeducational talk From the perspective of mentalist accounts of lan-guage which suggest that the talk children hear around them functionsprimarily as a trigger for language acquisition and a testing ground fortheir developing hypotheses children have already passed many of themajor milestones of language by the time they start school Thus thelanguage of teachers and caregivers may not be seen as having greatsigni cance in childrenrsquos language development

Functional and interactional linguistic theories present the view thatchildren have to learn to use language for a range of purposes and in arange of cultural and situational contexts Thus even though as innatistswould argue learners are ldquoprogrammedrdquo or predisposed to learnlanguage social theories of language and learning would suggest thatwhether they learn it how well they learn to control it and for whatpurposes they will ultimately be able to use it are dependent on the socialand thus interactional contexts in which they nd themselves (See egthe longitudinal studies of Halliday 1993 and Painter 1984 1998 whichtrack how the language of young children develops through the interac-tions and semiotic events in which the children participate) Rather thanviewing language as a nite set of rules that must be acquired systemictheorists view language as a semiotic systemmdasha set of choices from whichspeakers select according to the particular context they are in

This semiotic interpretation of language which views language as a setof resources rather than as a set of rules makes it possible to consider theappropriateness or inappropriateness of language choices in a given

MEDIATING LANGUAGE LEARNING 251

context of use One of the most fundamental features of language froma systemic perspective is that it varies according to the context ofsituation This context is characterised by three features what is beingtalked or written about ( eld) the relationship between the speakers orwriter and reader (tenor) and whether the language is spoken or written(mode) Language-in-use is determined by these contextual featuresand together these three variables constitute what is referred to as theregister of a text (Halliday amp Hasan 1985) In this study the focus was ondevelopment of the academic register of primary school science and todescribe this development I examined how mediation played a role inlearnersrsquo shifting along a mode continuum

A mode continuum is most simply conceived in relation to modedifferences between spoken and written language which have beendiscussed thoroughly by for example Martin (1984) Halliday (1985)Kress (1982) Derewianka (1990) Hammond (1990) and Eggins (1994)In reality however and if the effect mode has on language choices is tobe fully recognised this distinction between spoken and written lan-guage is more accurately viewed as a scale or continuum Martin (1984)suggests how aural and visual contact affect language and notes that

the more speakers are doing things together and engaging in dialogue themore they can take for granted As language moves away from the events itdescribes and the possibility of feedback is removed more and more of themeanings must be made explicit in the text (p 27)

In other words the language itself must contain more informationbecause it cannot depend on the addresseersquos knowing exactly whatoccurred

The four short texts in Table 1 exemplify a mode continuum Theregister of each text changes because the context in which it wasproduced is different Each text is more explicit than the one thatprecedes it Text 1 was spoken by a student in a small group experiment-ing with a magnet to nd out which objects it attracted It demonstrateshow dependent here-and-now language is on the immediate situationalcontext them and those have exophoric referents Text 2 the samespeaker telling the teacher what she had learned is in the form of arecount The increase in explicitness is the result of a context changeThe teacher had not shared in the experiences so more information isembedded in the text Thus participants are now named pins andmagnet Text 3 from the studentrsquos written report contains a generalisationand some eld-speci c lexis Text 4 by way of comparison is from achildrsquos encyclopedia The language has become denser and the processto which the child was referring in Texts 1 2 and 3 is now collapsed intothe nominalisation magnetic attraction

252 TESOL QUARTERLY

The way in which the language is used in the four texts thus differsconsiderably As they begin to refer to events not shared by listeners orreaders they take less for granted the lexical density increases andbecomes more eld speci c the tenor becomes more impersonal andthe language increasingly takes on the characteristics of written lan-guage These four texts illustrate what Martin (1984) refers to as ldquothegeneral concept of contextual dependencyrdquo (p 27) referring to the waysin which as language moves away from the events it describes more andmore of the meanings must be made explicit in the text if the listener orreader is to recover them

The continuum re ects the process of formal education itself asstudents are required to make shifts within an increasing number of elds and to move from personal everyday ways of making meaningstoward the socially shared and more writtenlike discourses of speci cdisciplines The development of literacy within any subject in the schoolcurriculum involves learning the technical language grammatical pat-terns and generic structures particular to the subject As the continuumsuggests these school-related registers tend to involve more writtenlikediscourse which tends to be less personal more abstract more lexicallydense and more structured than the face-to-face everyday language withwhich students are familiar Although more conversational texts tend tohave high personal involvement low explicitness of meaning andinteractive features these more academic texts require a high explicit-ness of lexical content but allow for little interaction or personalinvolvement (Biber 1986)

Vygotskyrsquos (1986) notion of spontaneous and scienti c concepts offersa related and somewhat similar perspective Spontaneous conceptsemerge from a childrsquos everyday experiences (as in Text 1 in Table 1) areembedded within speci c situational contexts and are therefore not

TABLE 1

Texts and Contexts Illustrating a Mode Continuum of Science Registers

Text Context

1 ldquoLook itrsquos making them move A student talking in a small group as theyThose didnrsquot stickrdquo were experimenting with a magnet

2 ldquoWe found out the pins stuck on the magnetrdquo A student telling the teacher what she hadlearned from the experiment

3 ldquoOur experiment showed that magnets A studentrsquos written report about theattract some metalsrdquo experiment

4 ldquoMagnetic attraction occurs only between An entry in a childrsquos encyclopedia aboutferrous metalsrdquo magnets

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

248 TESOL QUARTERLY

regular school curriculum not in a separate ESL classroom) The articleexplores in particular the way teacher-student interactions can enable L2learning in the context of subject teaching and the role of the teacherand learners in this process The texts show language learning as asocially mediated process whereby both teachers and learners are activeparticipants in the co-construction of language and curriculum knowl-edge The article draws on the constructs of mediation and modecontinuum to provide a way of conceptualising how this process is playedout in the classroom

MEDIATION FROM A SOCIOCULTURAL PERSPECTIVE

The construct of mediation is central to the sociocultural theory ofVygotsky (1981 1986 Lantolf 2000) because it provides a means ofstudying social processes involved in situated language learning and use(Appel amp Lantolf 1994 Hall 1995 Kramsch 1993 Moll 1994 Toohey2000 Wells amp Chang-Wells 1992) Re ecting Vygotskyrsquos notion thatlearning originates in the social mediation provided by interactions suchresearch in second language acquisition (SLA) questions the metaphorsof input and output (see eg Swain 2000 van Lier 2000) Socioculturaltheory views language learners not as processors of input or producers ofoutput but as ldquospeakershearers involved in developmental processeswhich are realised in interactionrdquo (Ohta 2000 p 51) Recent research(Donato 1994 2000 Engerstrom amp Middleton 1996 Ohta 1995 19992000 Swain 2000 Swain amp Lapkin 1995 Wells 1999) has shown howlearning and language acquisition are realised through a collaborativeinteractional process in which learners begin to appropriate the lan-guage of the interaction for their own purposes Much of this researchhas been concerned with learning through peer-peer interactions ratherthan through interactions between expert and novice whereas this studyexamines how language learning is mediated by language use in thecollaborative interactions between teacher and students

Mediation is central to the study of collaborative interactions Vygotsky(1981) argues that human activities and mental functioning are medi-ated and facilitated by tools cultural practices and artefacts the mostextensive tool being language Mediation is a familiar concept in manysocial contexts A lawyer for example mediates between a clientrsquosaccount of an event and the language principles and categories of thelegal world of the court (Maley Candlin Crichton amp Koster 1995)Mediation of this kind can broadly be described as occurring insituations characterised by difference dif culty or social distance(Baynham 1993) Such characteristics are inherent in most teacher-student relationships because in the great majority of school classrooms

MEDIATING LANGUAGE LEARNING 249

there is considerable linguistic and conceptual distance between teacherand students especially when they do not share the same languageassumptions and life experiences

Constructs related to mediation that are useful for investigatingcollaboration include zone of proximal development (ZPD) scaffolding jointconstruction and contingency The site where social forms of mediationdevelop is the ZPD (Lantolf 2000) The ZPD (Vygotsky 1978) refers tothe distance or the cognitive gap between what learners can do unaidedand what they can do in collaboration with a more competent other InVygotskian terms learning occurs through this assisted performance andin the context of joint activity Human development including languagedevelopment is thus intrinsically a social process and in the broadestsense educational Newman Grif n and Cole (1989) who call the ZPDthe construction zone de ne it as ldquothe changes that take place in sociallymediated instructionsrdquo (p 2)

The metaphor of scaffolding has been used by mother tongue and L2educators to describe the nature of this assisted performance whichinvolves not simply helping to do but helping to know how to do(Brooks 1992 Cazden 1988 Lee amp Smogarinsky 2000 Maybin Merceramp Stierer 1992 Mercer 1994 Webster Beveridge amp Reid 1996 Wells1999) The term was originally used by Wood Bruner and Ross (1976) intheir examination of parent tutoring in the early years In the classroomit can be de ned as the temporary but essential assistance that helpsapprentice learners into new skills concepts or levels of understanding(Maybin et al 1992 p 186) In relation to teacher-student interactionMaybin et al suggest two criteria for determining whether a particularexample of help can be portrayed as scaffolding There must be evidenceof a learnerrsquos successfully completing the task with the teacherrsquos help andevidence of the learnerrsquos having achieved a greater level of independentcompetence as a result of the experience For unlike the lawyerrsquos aimthe teacherrsquos aim is ultimately to hand over knowledge and control to thestudents In the sample texts discussed in this article this process occursthrough the interactional scaffolding that the teacher provides Thisstudy contributes to the ndings of other studies (eg McCormick ampDonato 2000) that focus on ldquothe interactional mechanisms involved inthe obtaining or providing of assistance during language learning tasksrdquo(Ohta 2000 p 52)

In this article I use the term mediation to characterise the linguistic anddiscourse choices made by teachers in the content-based ESL classroomThe texts illustrate how two teachers through their interactions withstudents mediate between the studentsrsquo current linguistic levels (inEnglish) and their commonsense understandings of science on the onehand and the educational discourse and specialist understandings of thesubject on the other In other words mediation involves communication

250 TESOL QUARTERLY

between two different orders of discourse the current levels of learnersrsquoknowledge and L2 abilities and the broader knowledge and specialistlanguage of the science community into which the students are beingapprenticed

MODE CONTINUUM FROM ASYSTEMIC FUNCTIONAL PERSPECTIVE

Describing mediation across orders of discourse however requires ameans of characterizing the different orders I use the construct of amode continuum to describe the different orders of discourse observed inthe classroom as the learners were assisted in moving from registersexpressing their rsthand experience in oral language to those express-ing academic knowledge in writing A description of such registers drawson systemic functional grammar following the work of Halliday andother systemicists because of the need to study the discourse of teachingand learning from the perspective of linguistic theory that explains howlanguage makes meaning Theories of language and language learningthat ignore its social context of use are of little help in an analysis ofeducational talk From the perspective of mentalist accounts of lan-guage which suggest that the talk children hear around them functionsprimarily as a trigger for language acquisition and a testing ground fortheir developing hypotheses children have already passed many of themajor milestones of language by the time they start school Thus thelanguage of teachers and caregivers may not be seen as having greatsigni cance in childrenrsquos language development

Functional and interactional linguistic theories present the view thatchildren have to learn to use language for a range of purposes and in arange of cultural and situational contexts Thus even though as innatistswould argue learners are ldquoprogrammedrdquo or predisposed to learnlanguage social theories of language and learning would suggest thatwhether they learn it how well they learn to control it and for whatpurposes they will ultimately be able to use it are dependent on the socialand thus interactional contexts in which they nd themselves (See egthe longitudinal studies of Halliday 1993 and Painter 1984 1998 whichtrack how the language of young children develops through the interac-tions and semiotic events in which the children participate) Rather thanviewing language as a nite set of rules that must be acquired systemictheorists view language as a semiotic systemmdasha set of choices from whichspeakers select according to the particular context they are in

This semiotic interpretation of language which views language as a setof resources rather than as a set of rules makes it possible to consider theappropriateness or inappropriateness of language choices in a given

MEDIATING LANGUAGE LEARNING 251

context of use One of the most fundamental features of language froma systemic perspective is that it varies according to the context ofsituation This context is characterised by three features what is beingtalked or written about ( eld) the relationship between the speakers orwriter and reader (tenor) and whether the language is spoken or written(mode) Language-in-use is determined by these contextual featuresand together these three variables constitute what is referred to as theregister of a text (Halliday amp Hasan 1985) In this study the focus was ondevelopment of the academic register of primary school science and todescribe this development I examined how mediation played a role inlearnersrsquo shifting along a mode continuum

A mode continuum is most simply conceived in relation to modedifferences between spoken and written language which have beendiscussed thoroughly by for example Martin (1984) Halliday (1985)Kress (1982) Derewianka (1990) Hammond (1990) and Eggins (1994)In reality however and if the effect mode has on language choices is tobe fully recognised this distinction between spoken and written lan-guage is more accurately viewed as a scale or continuum Martin (1984)suggests how aural and visual contact affect language and notes that

the more speakers are doing things together and engaging in dialogue themore they can take for granted As language moves away from the events itdescribes and the possibility of feedback is removed more and more of themeanings must be made explicit in the text (p 27)

In other words the language itself must contain more informationbecause it cannot depend on the addresseersquos knowing exactly whatoccurred

The four short texts in Table 1 exemplify a mode continuum Theregister of each text changes because the context in which it wasproduced is different Each text is more explicit than the one thatprecedes it Text 1 was spoken by a student in a small group experiment-ing with a magnet to nd out which objects it attracted It demonstrateshow dependent here-and-now language is on the immediate situationalcontext them and those have exophoric referents Text 2 the samespeaker telling the teacher what she had learned is in the form of arecount The increase in explicitness is the result of a context changeThe teacher had not shared in the experiences so more information isembedded in the text Thus participants are now named pins andmagnet Text 3 from the studentrsquos written report contains a generalisationand some eld-speci c lexis Text 4 by way of comparison is from achildrsquos encyclopedia The language has become denser and the processto which the child was referring in Texts 1 2 and 3 is now collapsed intothe nominalisation magnetic attraction

252 TESOL QUARTERLY

The way in which the language is used in the four texts thus differsconsiderably As they begin to refer to events not shared by listeners orreaders they take less for granted the lexical density increases andbecomes more eld speci c the tenor becomes more impersonal andthe language increasingly takes on the characteristics of written lan-guage These four texts illustrate what Martin (1984) refers to as ldquothegeneral concept of contextual dependencyrdquo (p 27) referring to the waysin which as language moves away from the events it describes more andmore of the meanings must be made explicit in the text if the listener orreader is to recover them

The continuum re ects the process of formal education itself asstudents are required to make shifts within an increasing number of elds and to move from personal everyday ways of making meaningstoward the socially shared and more writtenlike discourses of speci cdisciplines The development of literacy within any subject in the schoolcurriculum involves learning the technical language grammatical pat-terns and generic structures particular to the subject As the continuumsuggests these school-related registers tend to involve more writtenlikediscourse which tends to be less personal more abstract more lexicallydense and more structured than the face-to-face everyday language withwhich students are familiar Although more conversational texts tend tohave high personal involvement low explicitness of meaning andinteractive features these more academic texts require a high explicit-ness of lexical content but allow for little interaction or personalinvolvement (Biber 1986)

Vygotskyrsquos (1986) notion of spontaneous and scienti c concepts offersa related and somewhat similar perspective Spontaneous conceptsemerge from a childrsquos everyday experiences (as in Text 1 in Table 1) areembedded within speci c situational contexts and are therefore not

TABLE 1

Texts and Contexts Illustrating a Mode Continuum of Science Registers

Text Context

1 ldquoLook itrsquos making them move A student talking in a small group as theyThose didnrsquot stickrdquo were experimenting with a magnet

2 ldquoWe found out the pins stuck on the magnetrdquo A student telling the teacher what she hadlearned from the experiment

3 ldquoOur experiment showed that magnets A studentrsquos written report about theattract some metalsrdquo experiment

4 ldquoMagnetic attraction occurs only between An entry in a childrsquos encyclopedia aboutferrous metalsrdquo magnets

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 249

there is considerable linguistic and conceptual distance between teacherand students especially when they do not share the same languageassumptions and life experiences

Constructs related to mediation that are useful for investigatingcollaboration include zone of proximal development (ZPD) scaffolding jointconstruction and contingency The site where social forms of mediationdevelop is the ZPD (Lantolf 2000) The ZPD (Vygotsky 1978) refers tothe distance or the cognitive gap between what learners can do unaidedand what they can do in collaboration with a more competent other InVygotskian terms learning occurs through this assisted performance andin the context of joint activity Human development including languagedevelopment is thus intrinsically a social process and in the broadestsense educational Newman Grif n and Cole (1989) who call the ZPDthe construction zone de ne it as ldquothe changes that take place in sociallymediated instructionsrdquo (p 2)

The metaphor of scaffolding has been used by mother tongue and L2educators to describe the nature of this assisted performance whichinvolves not simply helping to do but helping to know how to do(Brooks 1992 Cazden 1988 Lee amp Smogarinsky 2000 Maybin Merceramp Stierer 1992 Mercer 1994 Webster Beveridge amp Reid 1996 Wells1999) The term was originally used by Wood Bruner and Ross (1976) intheir examination of parent tutoring in the early years In the classroomit can be de ned as the temporary but essential assistance that helpsapprentice learners into new skills concepts or levels of understanding(Maybin et al 1992 p 186) In relation to teacher-student interactionMaybin et al suggest two criteria for determining whether a particularexample of help can be portrayed as scaffolding There must be evidenceof a learnerrsquos successfully completing the task with the teacherrsquos help andevidence of the learnerrsquos having achieved a greater level of independentcompetence as a result of the experience For unlike the lawyerrsquos aimthe teacherrsquos aim is ultimately to hand over knowledge and control to thestudents In the sample texts discussed in this article this process occursthrough the interactional scaffolding that the teacher provides Thisstudy contributes to the ndings of other studies (eg McCormick ampDonato 2000) that focus on ldquothe interactional mechanisms involved inthe obtaining or providing of assistance during language learning tasksrdquo(Ohta 2000 p 52)

In this article I use the term mediation to characterise the linguistic anddiscourse choices made by teachers in the content-based ESL classroomThe texts illustrate how two teachers through their interactions withstudents mediate between the studentsrsquo current linguistic levels (inEnglish) and their commonsense understandings of science on the onehand and the educational discourse and specialist understandings of thesubject on the other In other words mediation involves communication

250 TESOL QUARTERLY

between two different orders of discourse the current levels of learnersrsquoknowledge and L2 abilities and the broader knowledge and specialistlanguage of the science community into which the students are beingapprenticed

MODE CONTINUUM FROM ASYSTEMIC FUNCTIONAL PERSPECTIVE

Describing mediation across orders of discourse however requires ameans of characterizing the different orders I use the construct of amode continuum to describe the different orders of discourse observed inthe classroom as the learners were assisted in moving from registersexpressing their rsthand experience in oral language to those express-ing academic knowledge in writing A description of such registers drawson systemic functional grammar following the work of Halliday andother systemicists because of the need to study the discourse of teachingand learning from the perspective of linguistic theory that explains howlanguage makes meaning Theories of language and language learningthat ignore its social context of use are of little help in an analysis ofeducational talk From the perspective of mentalist accounts of lan-guage which suggest that the talk children hear around them functionsprimarily as a trigger for language acquisition and a testing ground fortheir developing hypotheses children have already passed many of themajor milestones of language by the time they start school Thus thelanguage of teachers and caregivers may not be seen as having greatsigni cance in childrenrsquos language development

Functional and interactional linguistic theories present the view thatchildren have to learn to use language for a range of purposes and in arange of cultural and situational contexts Thus even though as innatistswould argue learners are ldquoprogrammedrdquo or predisposed to learnlanguage social theories of language and learning would suggest thatwhether they learn it how well they learn to control it and for whatpurposes they will ultimately be able to use it are dependent on the socialand thus interactional contexts in which they nd themselves (See egthe longitudinal studies of Halliday 1993 and Painter 1984 1998 whichtrack how the language of young children develops through the interac-tions and semiotic events in which the children participate) Rather thanviewing language as a nite set of rules that must be acquired systemictheorists view language as a semiotic systemmdasha set of choices from whichspeakers select according to the particular context they are in

This semiotic interpretation of language which views language as a setof resources rather than as a set of rules makes it possible to consider theappropriateness or inappropriateness of language choices in a given

MEDIATING LANGUAGE LEARNING 251

context of use One of the most fundamental features of language froma systemic perspective is that it varies according to the context ofsituation This context is characterised by three features what is beingtalked or written about ( eld) the relationship between the speakers orwriter and reader (tenor) and whether the language is spoken or written(mode) Language-in-use is determined by these contextual featuresand together these three variables constitute what is referred to as theregister of a text (Halliday amp Hasan 1985) In this study the focus was ondevelopment of the academic register of primary school science and todescribe this development I examined how mediation played a role inlearnersrsquo shifting along a mode continuum

A mode continuum is most simply conceived in relation to modedifferences between spoken and written language which have beendiscussed thoroughly by for example Martin (1984) Halliday (1985)Kress (1982) Derewianka (1990) Hammond (1990) and Eggins (1994)In reality however and if the effect mode has on language choices is tobe fully recognised this distinction between spoken and written lan-guage is more accurately viewed as a scale or continuum Martin (1984)suggests how aural and visual contact affect language and notes that

the more speakers are doing things together and engaging in dialogue themore they can take for granted As language moves away from the events itdescribes and the possibility of feedback is removed more and more of themeanings must be made explicit in the text (p 27)

In other words the language itself must contain more informationbecause it cannot depend on the addresseersquos knowing exactly whatoccurred

The four short texts in Table 1 exemplify a mode continuum Theregister of each text changes because the context in which it wasproduced is different Each text is more explicit than the one thatprecedes it Text 1 was spoken by a student in a small group experiment-ing with a magnet to nd out which objects it attracted It demonstrateshow dependent here-and-now language is on the immediate situationalcontext them and those have exophoric referents Text 2 the samespeaker telling the teacher what she had learned is in the form of arecount The increase in explicitness is the result of a context changeThe teacher had not shared in the experiences so more information isembedded in the text Thus participants are now named pins andmagnet Text 3 from the studentrsquos written report contains a generalisationand some eld-speci c lexis Text 4 by way of comparison is from achildrsquos encyclopedia The language has become denser and the processto which the child was referring in Texts 1 2 and 3 is now collapsed intothe nominalisation magnetic attraction

252 TESOL QUARTERLY

The way in which the language is used in the four texts thus differsconsiderably As they begin to refer to events not shared by listeners orreaders they take less for granted the lexical density increases andbecomes more eld speci c the tenor becomes more impersonal andthe language increasingly takes on the characteristics of written lan-guage These four texts illustrate what Martin (1984) refers to as ldquothegeneral concept of contextual dependencyrdquo (p 27) referring to the waysin which as language moves away from the events it describes more andmore of the meanings must be made explicit in the text if the listener orreader is to recover them

The continuum re ects the process of formal education itself asstudents are required to make shifts within an increasing number of elds and to move from personal everyday ways of making meaningstoward the socially shared and more writtenlike discourses of speci cdisciplines The development of literacy within any subject in the schoolcurriculum involves learning the technical language grammatical pat-terns and generic structures particular to the subject As the continuumsuggests these school-related registers tend to involve more writtenlikediscourse which tends to be less personal more abstract more lexicallydense and more structured than the face-to-face everyday language withwhich students are familiar Although more conversational texts tend tohave high personal involvement low explicitness of meaning andinteractive features these more academic texts require a high explicit-ness of lexical content but allow for little interaction or personalinvolvement (Biber 1986)

Vygotskyrsquos (1986) notion of spontaneous and scienti c concepts offersa related and somewhat similar perspective Spontaneous conceptsemerge from a childrsquos everyday experiences (as in Text 1 in Table 1) areembedded within speci c situational contexts and are therefore not

TABLE 1

Texts and Contexts Illustrating a Mode Continuum of Science Registers

Text Context

1 ldquoLook itrsquos making them move A student talking in a small group as theyThose didnrsquot stickrdquo were experimenting with a magnet

2 ldquoWe found out the pins stuck on the magnetrdquo A student telling the teacher what she hadlearned from the experiment

3 ldquoOur experiment showed that magnets A studentrsquos written report about theattract some metalsrdquo experiment

4 ldquoMagnetic attraction occurs only between An entry in a childrsquos encyclopedia aboutferrous metalsrdquo magnets

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

250 TESOL QUARTERLY

between two different orders of discourse the current levels of learnersrsquoknowledge and L2 abilities and the broader knowledge and specialistlanguage of the science community into which the students are beingapprenticed

MODE CONTINUUM FROM ASYSTEMIC FUNCTIONAL PERSPECTIVE

Describing mediation across orders of discourse however requires ameans of characterizing the different orders I use the construct of amode continuum to describe the different orders of discourse observed inthe classroom as the learners were assisted in moving from registersexpressing their rsthand experience in oral language to those express-ing academic knowledge in writing A description of such registers drawson systemic functional grammar following the work of Halliday andother systemicists because of the need to study the discourse of teachingand learning from the perspective of linguistic theory that explains howlanguage makes meaning Theories of language and language learningthat ignore its social context of use are of little help in an analysis ofeducational talk From the perspective of mentalist accounts of lan-guage which suggest that the talk children hear around them functionsprimarily as a trigger for language acquisition and a testing ground fortheir developing hypotheses children have already passed many of themajor milestones of language by the time they start school Thus thelanguage of teachers and caregivers may not be seen as having greatsigni cance in childrenrsquos language development

Functional and interactional linguistic theories present the view thatchildren have to learn to use language for a range of purposes and in arange of cultural and situational contexts Thus even though as innatistswould argue learners are ldquoprogrammedrdquo or predisposed to learnlanguage social theories of language and learning would suggest thatwhether they learn it how well they learn to control it and for whatpurposes they will ultimately be able to use it are dependent on the socialand thus interactional contexts in which they nd themselves (See egthe longitudinal studies of Halliday 1993 and Painter 1984 1998 whichtrack how the language of young children develops through the interac-tions and semiotic events in which the children participate) Rather thanviewing language as a nite set of rules that must be acquired systemictheorists view language as a semiotic systemmdasha set of choices from whichspeakers select according to the particular context they are in

This semiotic interpretation of language which views language as a setof resources rather than as a set of rules makes it possible to consider theappropriateness or inappropriateness of language choices in a given

MEDIATING LANGUAGE LEARNING 251

context of use One of the most fundamental features of language froma systemic perspective is that it varies according to the context ofsituation This context is characterised by three features what is beingtalked or written about ( eld) the relationship between the speakers orwriter and reader (tenor) and whether the language is spoken or written(mode) Language-in-use is determined by these contextual featuresand together these three variables constitute what is referred to as theregister of a text (Halliday amp Hasan 1985) In this study the focus was ondevelopment of the academic register of primary school science and todescribe this development I examined how mediation played a role inlearnersrsquo shifting along a mode continuum

A mode continuum is most simply conceived in relation to modedifferences between spoken and written language which have beendiscussed thoroughly by for example Martin (1984) Halliday (1985)Kress (1982) Derewianka (1990) Hammond (1990) and Eggins (1994)In reality however and if the effect mode has on language choices is tobe fully recognised this distinction between spoken and written lan-guage is more accurately viewed as a scale or continuum Martin (1984)suggests how aural and visual contact affect language and notes that

the more speakers are doing things together and engaging in dialogue themore they can take for granted As language moves away from the events itdescribes and the possibility of feedback is removed more and more of themeanings must be made explicit in the text (p 27)

In other words the language itself must contain more informationbecause it cannot depend on the addresseersquos knowing exactly whatoccurred

The four short texts in Table 1 exemplify a mode continuum Theregister of each text changes because the context in which it wasproduced is different Each text is more explicit than the one thatprecedes it Text 1 was spoken by a student in a small group experiment-ing with a magnet to nd out which objects it attracted It demonstrateshow dependent here-and-now language is on the immediate situationalcontext them and those have exophoric referents Text 2 the samespeaker telling the teacher what she had learned is in the form of arecount The increase in explicitness is the result of a context changeThe teacher had not shared in the experiences so more information isembedded in the text Thus participants are now named pins andmagnet Text 3 from the studentrsquos written report contains a generalisationand some eld-speci c lexis Text 4 by way of comparison is from achildrsquos encyclopedia The language has become denser and the processto which the child was referring in Texts 1 2 and 3 is now collapsed intothe nominalisation magnetic attraction

252 TESOL QUARTERLY

The way in which the language is used in the four texts thus differsconsiderably As they begin to refer to events not shared by listeners orreaders they take less for granted the lexical density increases andbecomes more eld speci c the tenor becomes more impersonal andthe language increasingly takes on the characteristics of written lan-guage These four texts illustrate what Martin (1984) refers to as ldquothegeneral concept of contextual dependencyrdquo (p 27) referring to the waysin which as language moves away from the events it describes more andmore of the meanings must be made explicit in the text if the listener orreader is to recover them

The continuum re ects the process of formal education itself asstudents are required to make shifts within an increasing number of elds and to move from personal everyday ways of making meaningstoward the socially shared and more writtenlike discourses of speci cdisciplines The development of literacy within any subject in the schoolcurriculum involves learning the technical language grammatical pat-terns and generic structures particular to the subject As the continuumsuggests these school-related registers tend to involve more writtenlikediscourse which tends to be less personal more abstract more lexicallydense and more structured than the face-to-face everyday language withwhich students are familiar Although more conversational texts tend tohave high personal involvement low explicitness of meaning andinteractive features these more academic texts require a high explicit-ness of lexical content but allow for little interaction or personalinvolvement (Biber 1986)

Vygotskyrsquos (1986) notion of spontaneous and scienti c concepts offersa related and somewhat similar perspective Spontaneous conceptsemerge from a childrsquos everyday experiences (as in Text 1 in Table 1) areembedded within speci c situational contexts and are therefore not

TABLE 1

Texts and Contexts Illustrating a Mode Continuum of Science Registers

Text Context

1 ldquoLook itrsquos making them move A student talking in a small group as theyThose didnrsquot stickrdquo were experimenting with a magnet

2 ldquoWe found out the pins stuck on the magnetrdquo A student telling the teacher what she hadlearned from the experiment

3 ldquoOur experiment showed that magnets A studentrsquos written report about theattract some metalsrdquo experiment

4 ldquoMagnetic attraction occurs only between An entry in a childrsquos encyclopedia aboutferrous metalsrdquo magnets

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 251

context of use One of the most fundamental features of language froma systemic perspective is that it varies according to the context ofsituation This context is characterised by three features what is beingtalked or written about ( eld) the relationship between the speakers orwriter and reader (tenor) and whether the language is spoken or written(mode) Language-in-use is determined by these contextual featuresand together these three variables constitute what is referred to as theregister of a text (Halliday amp Hasan 1985) In this study the focus was ondevelopment of the academic register of primary school science and todescribe this development I examined how mediation played a role inlearnersrsquo shifting along a mode continuum

A mode continuum is most simply conceived in relation to modedifferences between spoken and written language which have beendiscussed thoroughly by for example Martin (1984) Halliday (1985)Kress (1982) Derewianka (1990) Hammond (1990) and Eggins (1994)In reality however and if the effect mode has on language choices is tobe fully recognised this distinction between spoken and written lan-guage is more accurately viewed as a scale or continuum Martin (1984)suggests how aural and visual contact affect language and notes that

the more speakers are doing things together and engaging in dialogue themore they can take for granted As language moves away from the events itdescribes and the possibility of feedback is removed more and more of themeanings must be made explicit in the text (p 27)

In other words the language itself must contain more informationbecause it cannot depend on the addresseersquos knowing exactly whatoccurred

The four short texts in Table 1 exemplify a mode continuum Theregister of each text changes because the context in which it wasproduced is different Each text is more explicit than the one thatprecedes it Text 1 was spoken by a student in a small group experiment-ing with a magnet to nd out which objects it attracted It demonstrateshow dependent here-and-now language is on the immediate situationalcontext them and those have exophoric referents Text 2 the samespeaker telling the teacher what she had learned is in the form of arecount The increase in explicitness is the result of a context changeThe teacher had not shared in the experiences so more information isembedded in the text Thus participants are now named pins andmagnet Text 3 from the studentrsquos written report contains a generalisationand some eld-speci c lexis Text 4 by way of comparison is from achildrsquos encyclopedia The language has become denser and the processto which the child was referring in Texts 1 2 and 3 is now collapsed intothe nominalisation magnetic attraction

252 TESOL QUARTERLY

The way in which the language is used in the four texts thus differsconsiderably As they begin to refer to events not shared by listeners orreaders they take less for granted the lexical density increases andbecomes more eld speci c the tenor becomes more impersonal andthe language increasingly takes on the characteristics of written lan-guage These four texts illustrate what Martin (1984) refers to as ldquothegeneral concept of contextual dependencyrdquo (p 27) referring to the waysin which as language moves away from the events it describes more andmore of the meanings must be made explicit in the text if the listener orreader is to recover them

The continuum re ects the process of formal education itself asstudents are required to make shifts within an increasing number of elds and to move from personal everyday ways of making meaningstoward the socially shared and more writtenlike discourses of speci cdisciplines The development of literacy within any subject in the schoolcurriculum involves learning the technical language grammatical pat-terns and generic structures particular to the subject As the continuumsuggests these school-related registers tend to involve more writtenlikediscourse which tends to be less personal more abstract more lexicallydense and more structured than the face-to-face everyday language withwhich students are familiar Although more conversational texts tend tohave high personal involvement low explicitness of meaning andinteractive features these more academic texts require a high explicit-ness of lexical content but allow for little interaction or personalinvolvement (Biber 1986)

Vygotskyrsquos (1986) notion of spontaneous and scienti c concepts offersa related and somewhat similar perspective Spontaneous conceptsemerge from a childrsquos everyday experiences (as in Text 1 in Table 1) areembedded within speci c situational contexts and are therefore not

TABLE 1

Texts and Contexts Illustrating a Mode Continuum of Science Registers

Text Context

1 ldquoLook itrsquos making them move A student talking in a small group as theyThose didnrsquot stickrdquo were experimenting with a magnet

2 ldquoWe found out the pins stuck on the magnetrdquo A student telling the teacher what she hadlearned from the experiment

3 ldquoOur experiment showed that magnets A studentrsquos written report about theattract some metalsrdquo experiment

4 ldquoMagnetic attraction occurs only between An entry in a childrsquos encyclopedia aboutferrous metalsrdquo magnets

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

252 TESOL QUARTERLY

The way in which the language is used in the four texts thus differsconsiderably As they begin to refer to events not shared by listeners orreaders they take less for granted the lexical density increases andbecomes more eld speci c the tenor becomes more impersonal andthe language increasingly takes on the characteristics of written lan-guage These four texts illustrate what Martin (1984) refers to as ldquothegeneral concept of contextual dependencyrdquo (p 27) referring to the waysin which as language moves away from the events it describes more andmore of the meanings must be made explicit in the text if the listener orreader is to recover them

The continuum re ects the process of formal education itself asstudents are required to make shifts within an increasing number of elds and to move from personal everyday ways of making meaningstoward the socially shared and more writtenlike discourses of speci cdisciplines The development of literacy within any subject in the schoolcurriculum involves learning the technical language grammatical pat-terns and generic structures particular to the subject As the continuumsuggests these school-related registers tend to involve more writtenlikediscourse which tends to be less personal more abstract more lexicallydense and more structured than the face-to-face everyday language withwhich students are familiar Although more conversational texts tend tohave high personal involvement low explicitness of meaning andinteractive features these more academic texts require a high explicit-ness of lexical content but allow for little interaction or personalinvolvement (Biber 1986)

Vygotskyrsquos (1986) notion of spontaneous and scienti c concepts offersa related and somewhat similar perspective Spontaneous conceptsemerge from a childrsquos everyday experiences (as in Text 1 in Table 1) areembedded within speci c situational contexts and are therefore not

TABLE 1

Texts and Contexts Illustrating a Mode Continuum of Science Registers

Text Context

1 ldquoLook itrsquos making them move A student talking in a small group as theyThose didnrsquot stickrdquo were experimenting with a magnet

2 ldquoWe found out the pins stuck on the magnetrdquo A student telling the teacher what she hadlearned from the experiment

3 ldquoOur experiment showed that magnets A studentrsquos written report about theattract some metalsrdquo experiment

4 ldquoMagnetic attraction occurs only between An entry in a childrsquos encyclopedia aboutferrous metalsrdquo magnets

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 253

systematic Scienti c concepts on the other hand are located within thestructured and specialised discourse of the subject (as in Text 4) aremore xed and are systematically related and logically organised (Kozulin1998) Thus as Cummins (2000) points out in his discussion of languagepro ciency in academic contexts ldquothe academic tasks [ESL students] arerequired to complete and the linguistic contexts in which they mustfunction become more complex with respect to the registers employedin these contextsrdquo (p 67)

Because fewer linguistic resources are required a young L2 learner islikely to have fewer dif culties producing a text that is embedded in theimmediate situational context than producing more context-reducedtexts which place a greater demand on the learnerrsquos lexicogrammaticalresources Yet in the classroom an oral reporting stage (like Text 2 inTable 1) is often not given much attention and although schoolclassrooms are usually rich in the provision of experiential learningactivities learners are frequently expected to write simply on the basis ofthese personal experiences which represents a very large linguistic step(as can be seen by comparing Texts 1 and 3 in Table 1) that is beyond thelinguistic resources of many young L2 learners In the current study amajor focus is how teachers can support students in developing spokenbut less context dependent language as a way into gaining control of themore formal and often written registers of the curriculum

METHOD

This article is drawn from a larger study with a primarily interpretiveapproach By interpretive I refer to qualitative studies that take asemiotic approach that is one that focuses on the co-construction ofmeaning within a particular social setting (Davis 1995 Hammersley1994) The major concern of that study was to identify factors inclassroom discourse that enabled (or constrained) language develop-ment and to theorise this through instances of language teaching in situ(Gibbons in press)

Context

The data in the larger study were taken from two classes of 9- and 10-year-olds in their fth year of schooling in the same Australian schoolOne of the class teachers had previously worked as a consultant in ESLeducation and had then returned to the classroom and the other hadtaken part in a number of professional development activities in theschool that focused on teaching ESL students Both teachers included

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

254 TESOL QUARTERLY

the teaching of language across the curriculum in their programplanning and were familiar with the mode continuum as a means ofconceptualizing language development

The school was an urban school in a poor socioeconomic area whereat the time of the study 92 of the children were from languagebackgrounds other than English this percentage was approximately thesame for all classes Of the total of 60 children in the two classes manyhad been born in Australia but had entered the rst year of formaleducation with little English other students were rst-generation mi-grants including ve children who had arrived in Australia within theprevious year Usually children with little English very quickly becomeadept at using it in here-and-now contexts where interactions occur face-to-face and often relate directly to what is occurring in the immediatesituation such as playing games in the playground However as Cummins(1984 1988 1996 2000) and others (Collier 1989 McKay et al 1997)have shown children who appear uent in English in such contexts maystill have dif culty understanding and using the registers associated withacademic learning in school described above The learning of theseregisters by ESL students was the particular focus of the study

Data Collection and Analysis

In both classrooms data were collected during one complete unit ofwork or topic consisting of 7 and 11 lessons of approximately 45ndash50minutes Data sources included audio recordings and transcriptions of14 hours of discourse environmental print around the classroom suchas posters charts and childrenrsquos work eld notes and interviews withteachers and students These varied sources facilitated data triangula-tion although in this article the focus is on the transcribed discourse

Handling the extensive amount of data that result from such anapproach is often a major problem in research of this kind In this studythe transcribed data were analyzed at two levels The rst documentedevery teaching and learning activity in the two classrooms and provideda holistic perspective on the total data set indicating the kind of activitythe interaction pattern where the activity fell on the mode continuumand what children were learning about science metalanguage and theiridentity as students This broad analysis indicated how the overall unit ofwork was organised and how it developed and de ned major patterns ofdiscourse and learning From this broad analysis there emerged anumber of themes which were taken up in the second level of analysisThis more detailed analysis drew on the construct of mediation fromsociocultural approaches to learning and mode continuum from systemicfunctional linguistics The texts discussed in this article are taken from

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 255

the second level of analysis and have been selected as representative ofthe regular learning activities and typical discourse patterns identi ed bythe initial more holistic analysis

Because an overall aim of the study was to explore how the register ofscience was constructed in the two classrooms the analysis consideredthe meanings that extended beyond a single lesson and examined howthese were built up over time In any classroom teacher and studentshear and produce language against an extensive background of accumu-lated meanings which researchers are in danger of ignoring if data arecollected on single visits The theoretical need to observe ongoingsequences of lessons has been demonstrated by a number of researchers(Brilliant-Mills 1993 Christie 1995 Floriani 1993 Heras 1994 Lin1993 Mercer 1995) In fact a sociocultural perspective itself demands aldquoholistic qualitative methodologyrdquo (Ohta 2000 p 53) that can explainlearning processes as they occur in interactive settings Thus whole unitsrather than single lessons were the macro units of analysis in the broaderstudy

Focusing on a sequence of lessons is also necessary to avoid inaccurateobservations Some of the science lessons observed consisted entirely ofstudents carrying out experiments in small groups whereas in otherlessons the teacher took a major role in initiating talk the initiation-response-feedback (IRF) pattern was very much in evidence Observingone or another of these lessons might lead to a conclusion that theclassroom was teacher fronted and teacher directed or conversely that itwas totally student centred Neither lesson alone would have provided asense of what the teachers were doing or how knowledge and languagewere being progressively built up

By observing the classroom over a sustained period of time I couldthus observe how a teacher handled all stages of learning for examplehow a topic was introduced if and how studentsrsquo prior learning was builton how new learning and language were developed through thediscourse and what evidence there was that the students took up thisnew learning and language

The Teaching Program

Based on the science topic of magnetism the teachers with someinput from me had planned teaching and learning activities that would(in terms of the language that students would be likely to use) re ectpoints along the mode continuum the assumption being that theactivities would offer a developmental sequence of language learningalong that continuum Broadly three points can be identi ed as stages inthe program doing an experiment in small groups using concrete

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

256 TESOL QUARTERLY

materials reconstructing these personal experiences through reportingto the class (who had carried out similar but not identical experiments)and nally completing an informal writing task in science journals

In the rst stage students carried out a series of experiments in smallgroups with each group doing a different experiment The second stagethe reconstruction of personal experiences occurred through what Ihave described as teacher-guided reporting (Gibbons 2001) Groups ofstudents with the help of the teacher shared their learning with thewhole class It was at this stage that the teacher and students began to co-construct the more formal register of school science the interactionsthat occurred here are the focus of this article As Driver (1994) pointsout in her critique of inductive methods of teaching science ldquotheoreticalmodels and scienti c conventions will not be lsquodiscoveredrsquo by childrenthrough their practical work guidance is need to help childrenassimilate their practical experiences into what is possibly a new way ofthinking about themrdquo (p 49) Martin (1990) likewise states that

common-sense knowledge can be a very useful starting point for learningscience because it organises the world in ways that can be clearly related toscienti c understandings at the same time it is clear that commonsenseunderstandings differ from scienti c ones and that schools have a crucialresponsibility to induct students into the alternative scienti c world views(p 84)

(In relation to issues of technicality and mode in science discourseeducation see also Halliday amp Martin 1993 Martin amp Veel 1998)

During teacher-guided reporting the teachers modeled and focusedon key lexis (eg attract and repel ) or on signi cant grammaticalstructures either through a brief explanation or in the course of jointlyconstructed interactions Toward the end of these sessions the teachershelped students build up generalisations by directing their attention tothe commonalities in the groupsrsquo ndings Re ecting the principlessuggested by Driver (1994) teacher-guided reporting thus allowed aspace for the teacher to assist learners to construct principled under-standings about the activities in which they had taken part As expectedalthough students had little dif culty in talking about what they weredoing in the face-to-face setting of the experiment it was considerablymore of a challenge for them to reconstruct through language what hadoccurred for the bene t of others As already pointed out they werebeing required at this stage to shift along the mode continuum towardmore writtenlike language The more explicit use of spoken languagerequired in these reporting sessions provided a linguistic bridge into the nal stage of the teaching sequence when students wrote in theirjournals Manipulating the contextual variables through these threestages resulted in a systematic increase in the lexicogrammatical demands

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 257

on the students that was intended to help them move along the modecontinuum This three-stage cycle was repeated several times during theunit of work

FINDINGS

All the texts discussed below come from the teacher-guided reportingsessions when as described above teachers and students togetherreconstructed what had occurred in the face-to-face context of theexperiments and began to recontextualise it in the discourse of scienceThrough the teacherrsquos mediation studentsrsquo contributions to the dis-course of the classroom were progressively transformed into the special-ist discourse of the school curriculum The texts illustrate how theteachersrsquo mediating role played out and indicate how the teachers builtlinguistic bridges to span the dif culty difference or distance referredto earlier Because in content classrooms the learning of science must gohand-in-hand with the development of an L2 classroom interactions area major site for language development so I focus on some of the waysteacher-student interactions can create discourse sites that enable use ofthe more scienti c register In the examples teachers mediate languagelearning in several ways mode shifting through recasting signaling tolearners how to reformulate indicating the need for reformulation andrecontexualising personal knowledge

Mode Shifting and Recasting

The teachers explicitly engaged in mode shifting at a macro levelacross different teaching activities but mode shifts also occurred withinthe discourse by which these activities were realised at the microlevel ofongoing and moment-by-moment interactions between teacher andstudents In Text 1 (Figure 1) for example mode shifting takes placewhen the teacher invites Charbel1 to talk about what his group hadlearned about the behaviour of two bar magnets in relation to theposition of the poles This text is typical of many exhibiting an ongoingprocess of recapping by the teacher who re-represents or recontextualiseslearnersrsquo experiences and the events they are talking about in a way that ts the broader pedagogic objectives of the curriculum There remainshowever a close similarity between what students say and how theirversion of events is recontextualised In such cases the mode shiftingoccurs when the teacher recasts a studentrsquos contribution

1 Studentsrsquo names in the transcripts are pseudonyms

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

258 TESOL QUARTERLY

I refer to this type of mode shift as recasting noting however that thisis a somewhat extended use of the term as used by most SLA researchers2

(see eg Lyster 1998 Oliver 1995) The teacherrsquos recast version in thiscase is thematically related to the studentrsquos version even though differ-ent thematic items are used stickattract not pushingrepelling The samesemantic relations are constructed and the same thematic pattern isrepeated it sticks togetherthey attracted each other you can feel theyrsquore notpushingyou felt it repelling Equivalent words occupy the same or corre-sponding slots in a similar grammatical construction Lemke (1990)refers to this as local equivalence the marking of two expressions asequivalent within the thematic pattern being built up further markedhere by the use of emphasis Here the teacherrsquos response closely followsthe studentrsquos grammatical construction appropriating the studentrsquosmeanings while recoding the everyday wordings and recasting them asattract and repel This recasting and extension of student-initiated mean-

FIGURE 1

Mode Shifting in Text 1

Teacher

SituationallyStudent embedded Everyday Formal

it sticks togetherlike that(demonstrating)

they attracted to eachother

they stuck to eachother

you can feel thattheyrsquore not pushing if we use the other sidewe canrsquot feel pushing

when they were facing oneway you felt the magnetsattract

and stick togetherwhen you turn one of themagnets around you felt itrepelling

or pushing away

2 Recasting has generally been used to refer to reformulations of child or L2 speakerrsquosutterances at the level of morphology or syntax Here I am using it to refer to any piece ofconnected discourse where a teacher rewords student meaning in more registrally appropriateways

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 259

ing depends on the adultrsquos contribution being closely related to andthus following the studentrsquos contribution In other words the teacherrsquosmodeling occurs after and on the basis of what the student hascontributed and thus is semantically contingent upon it Semanticallycontingent speech ne-tuned to an individual learner has been identi- ed by researchers as an enabling factor in both mother tongue and L2development (Snow 1986 van Lier 1996 Webster et al 1996 Wells1985)

Further examination of the recast reveals that it consists of more thana single shift between two points on the mode continuum The teacherrsquoscontributions themselves include instances of three distinct points alongthe mode continuum For the purposes of examining the teacher talk Icharacterise these points as situationally embedded (representing thoseparts of the discourse that contain exophoric reference and are boundup with and rely on the immediate visual context for their interpreta-tion) formal (representing the standard lexis of school science) andeveryday (representing the informal spoken language familiar to thechildren) These mode shifts result in considerable message redundancyan important aspect of discourse in facilitating comprehension for L2learners (see eg Wong-Fillmore 1985) Figure 2 based on Text 1illustrates the mode shifting through which this redundancy is achieved

The mode shifts evident in the teacherrsquos discourse offer a microper-spective on the way the discourse operates as a linguistic bridge betweenstudentsrsquo current language abilities and the demands of the schoolcurriculum The two middle columns of Figure 2 span these two ordersof discourse The bridging might be interpreted as the ZPD in action anoperationalisation of the notion in terms of L2 teaching and learningExploring the mode shifts within the discourse also offers a linguisticperspective on the construct of comprehensible input and indicates onediscourse strategy by which it is achieved in classroom practice Here thestudents can access three sources of meaning the demonstration by theteacher the description of the process using familiar language (stick topushing away) and the technical terms (attract repel) In this sense it is amultimodal text which provides considerable message redundancy forL2 learners3 Clearly however comprehensibility here is not synonymouswith simpli cation rather students are given access to key technicalterms in a context where meanings are made transparent

However although such exchanges seem likely to increase the capac-ity of the discourse to facilitate language learning researchers havesuggested that implicit teacher recasts (ie recasts that simply reformu-late all or part of the studentrsquos utterance with no additional meaning and

3 Although not the usual term perhaps the notion of message abundancy better captures whatis happening here

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

260 TESOL QUARTERLY

without drawing the studentrsquos attention to the reformulation) minimisethe value of studentsrsquo utterances (Lyster 1998 p 55) for such results donot require students to adjust what they have said (Pica 1988 1994)Students appear to need time and opportunity for self-repair (van Lier1988 1996) Thus when listeners signal a need for clari cation this maybene t L2 learning in two ways by providing clues to assist learners inmodifying and actively confronting communication dif culties (referredto here as signaling how to reformulate) and by inviting student- ratherthan teacher-generated repair (referred to here as indicating a need for

FIGURE 2

Text 1

Turn Student Teacher

1 what were your results2 when we put it on one pole em

faces the other one it doesnrsquot stickbut when we turned the other onearound it sticks together

3 OK can I just clarify something yoursquove gottwo magnets theyrsquore in line when you put the two together

4 yes Miss5 like that (demonstrating) they attracted to each

other they stuck to each other is thatright

6 (nods) OK can you then tell me what you had to donext

7 when we had em the things the rstone like if you put it up in the airlike that the magnets you canfeel feel the em that theyrsquorenot pushing

8 when you turn the magnet around you feltthat

9 pushing and if we use the other sidewe canrsquot feel pushing

10 OK so when they were facing one way they you felt the magnets attract and sticktogether when you turn one of the magnetsaround you felt it repelling or pushingaway OK thank you well done Charbel

Note The transcripts are set out in columns to enable a clearer focus on the signi cance of eachspeakerrsquos contribution (eg the changes evident in a studentrsquos verbal behaviour or theteacherrsquos scaffolding) Reading across the columns shows the co-construction of the dialogue(eg the coherence and semantic relationship between speakers) Transcription conventionsare as follows

approximate 1-second pause meaning group boundary

underlining marked emphasisitalics procedural language such as turn nominations and evaluative comments

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 261

reformulation) The following two sections illustrate both these ways ofsignaling a need for clari cation

Signaling How to Reformulate

In Text 2 (Figure 3) the teacher signals a need for clari cation andsupplies a recoded version of the studentrsquos meaning only after thelearner has had opportunities for self-correction Julianna is attemptingto describe how when she placed a piece of aluminium foil between amagnet and a nail the magnet still attracted the nail Mediation by theteacher here is focused less on providing new language than onproviding pointers for the student to reformulate her own wordingAgain the co-constructed nature of the text is signi cant The teacherrsquosscaffolding is contingent on the meanings the student is trying toconstruct

FIGURE 3

Text 2

Turn Student (Julianna) Teacher

1 what did you nd out2 if you put a nail onto the piece of

foil and then pick it pick it up the magnet will that if youput a nail under a piece of foil and then pick pick the foil up withthe magnet still still with thenail under it it wonrsquot

3 it what4 it wonrsquot it wonrsquot come out5 what wonrsquot come out6 itrsquoll go up7 wait just a minute can you explain that a

bit more Julianna8 like if you put a nail and then foil

over it and then put the nail ontop of the foil the nailunderneath the foil Miss I canrsquotsay it

9 no yoursquore doing ne I I can see10 Miss forget about the magnet em

the magnet holds it with the foil upthe top and the nailrsquos underneathand the foilrsquos on top and put themagnet in it and you lift it up andthe nail will em hold itstick withthe magnet and the foilrsquos in between

11 oh so even with the foil in between the magnet will still pick up the nail alrightdoes the magnet pick up the foil

12 no

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

262 TESOL QUARTERLY

Juliannarsquos rst explanation (Turn 2) is extended but extremelyhesitant and unclear At Turn 3 the teacher could have closed theexchange by recasting what Julianna is attempting to say (as she doeseventually in Turn 11) Instead she increases the demands made on thestudent as interactant by asking a question in order to determine theprocess to which the student is referring ldquoit whatrdquo (Turn 3 ie it wonrsquotdo what) When the studentrsquos meaning is still not explicit the teacherasks a further question this time to elicit the missing participant ldquowhatwonrsquot come outrdquo (Turn 5 ie what thing wonrsquot come out) When thisquestion does not result in a clearer explanation she asks for furtherclari cation ldquocan you explain that a bit morerdquo (Turn 7) The studentrsquosnext attempt repeats much of the information of Turn 2 but is noticeablyless hesitant However she stops with the very telling words ldquoI canrsquot say itrdquo(Turn 8) The teacher encourages her to continue this time byforegrounding the interpersonal ldquoyoursquore doing nerdquo (Turn 9) Juliannarsquos nal attempt is considerably more complete and is the least hesitantincorporating the key fact that the foil was between the nail and themagnet Compared with the initial attempt (Turn 2) it is a far moreexplicit and comprehensible piece of information

As is evident from the visual layout of the transcript alone the studenttalks far more than the teacher throughout the sequence The teacherhands over to the student the responsibility for clari cation whichresults in increasingly explicit information from the student or whatSwain (1985 1995) has referred to as comprehensible outputmdashcompare forexample Turns 2 4 and 11 Swain also argues for the need for stretchedlanguagemdashlearners must have opportunities to use language that stretchesthem to the outer limits of their capabilities Here Juliannarsquos linguisticresources are clearly being stretched her comment ldquoI canrsquot say itrdquoindicates just how much

In Vygotskian terms in relation to this task Julianna is at the outerlimits of what she can do alone Yet because of the precise andcontingent nature of the teacherrsquos scaffolding the text is characterisedby the studentrsquos rather than the teacherrsquos reformulations The teacher nally does the job of recontextualising the studentrsquos meaning in moreconcise wording but not until the 11th move offering a greatlyincreased opportunity for negotiation of meaning over what would haveoccurred in a three-part exchange And as is well attested such negotia-tion of meaning particularly when achieved through requests forclari cation in which learners must adjust what they have said appears tofacilitate L2 learning (Pica 1988 Pica Holliday Lewis amp Morgenthaler1989) As Hall (1998) points out language learning does not dependonly on studentsrsquo abilities or on their knowledge and skills or motivationfor learning but is ldquotied to the teacherrsquos motivation for and interest inproviding her individual learnersrsquo with lsquoof cial participatory rightsrsquo to

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 263

engage fully in the opportunities for exhibiting and building on theirknowledge and skills in their classroom practicesrdquo (p 308)

In Text 2 the increase in negotiation of meaning is achieved by asmall but highly signi cant adaptation of the usual three-part classroomexchange Here in place of the third feedback move the teacher asks aquestion designed to elicit additional information from the student afeedback move that Wells (1993 1996 1999) refers to as a pivot to thenext exchange He describes how the division of labour typical of the IRFexchange in which the teacher does most of the talking can beredistributed through this move with the student rather than theteacher taking responsibility for making what she says more comprehen-sible to her audience In this way the teacher helps the learner extendher initial response in her subsequent move thereby extending theexchange in Text 2 the teacher does this not once but several times AsCazden (1988) has also observed a relatively minor change in thetraditional IRF sequence can have signi cant effects on the process ofthe discourse as a whole Consider for example what opportunities forlanguage learning would have been lost if the teacher had recast whatJulianna was trying to say at Turn 3 It is particularly signi cant that inthe childrenrsquos journal writing after such talk with the teacher thelexicogrammatical choices of the children re ected these negotiatedand co-constructed texts indicating at least some take-up of the lan-guage and suggesting the importance of the prior rehearsal of morewrittenlike discourse afforded by the teacher-guided reporting episodes(for discussion of evidence of take-up in this context see Gibbons 19982001)

Indicating a Need for Reformulation

In the exchanges shown in Texts 3 and 4 (Figure 4) the teacherindicates the need for reformulation in terms of a more registrallyappropriate response but knowing that the learner can achieve it aloneshe hands the responsibility over to the student The examples illustrateagain how a request for clari cationmdashand perhaps the consequent extratime for formulating a response that this request allows the studentmdashmay in itself result in longer and more complete learner discourse

As in Text 2 the teacherrsquos third move in Text 3 is not an evaluation ora reformulation but a further question that results in the studentrsquoslanguage being stretched As a result of the teacherrsquos contributionBeatrice makes more of her reasoning explicit in the discourse assumingless shared knowledge on the part of her listeners The adjunct still forexample which requires some shared understanding of the context tobe interpreted is recoded more explicitly as a logical conjunction

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

264 TESOL QUARTERLY

indicating a causal relationship even though As pointed out earlier thisability to use explicit discourse is usually required in the academicregisters and written language of school Similarly in Text 4 the promptfrom the teacher which again explicitly draws the studentrsquos attention toregister results in an extended response that is more appropriate for theregister

Recontextualising Personal Knowledge

Text 5 (Figure 5) occurred toward the end of a teacher-guidedreporting episode when the teacher was focusing more explicitly on thelanguage needed to talk about the studentsrsquo ndings in more registrallyappropriate ways speci cally here the wording of a generalisation Atthis point studentsrsquo individual ndings are also being recontextualised interms of the broader principles and framework of science On the boardwas a matrix (Figure 6) and as each student responded the teachermarked the appropriate box with a tick

FIGURE 4

Texts 3 and 4

Turn Student Teacher

Text 31 tell us what happened2 Beatrice em we put three magnets

together it still wouldnrsquot hold thegold nail

3 can you explain that again4 Beatrice we we tried to put three

magnets together to hold thegold nail even though we hadthree magnets it wouldnrsquot stick

Text 41 tell us what you found out2 Michelle we found out that the

south and the south donrsquot like tostick together

3 now letrsquosletrsquos start using our scienti clanguage Michelle

4 Michelle the north and the northrepelled each other and the southand the south also repelled eachother but when we put the whenwe put the two magnets in a differentway they they attracted each other

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 265

FIGURE 5

Text 5

Teacher

Turn Students Field Language Field Science

1 letrsquos try this what if I try thenorth pole and the southpole of the magnet whocan tell

2 I want a sentence a nicesentence Carol Ann

3 Carol Ann the north poleand the south pole attract

4 good what if I try thesouth pole of this magnetand the north pole of thatmagnet yes Franccedilois

5 come on a sentence6 Franccedilois the south pole

and the north pole willattract

7 good boy good Franccedilois southalright and letrsquos try the poleof this magnet and thesouth pole of the othermagnet Stephanie

8 Stephanie the south poleand the south pole willre repel

9 my goodness arenrsquot theyspeaking well so I would liketwo ideas that we get fromthis two general ideas whatwe call generalisations hellipwhocan give me something thatwill happen all the time notwhat just happened to us today

10 Gina do you want to try11 If you put the north pole

and the north poletogether em that will not that will repel and ifyou put the south pole andthe south pole togetherthat will repel too

12 good alright that will alwayshappen so wersquoll say

13 south pole and south pole 14 SS repel15 north pole and north pole 16 SS repel17 alright who can give me

something else Jennifer18 Jennifer em the north

pole and the south pole attract19 right they attract each

other north pole andsouth pole attract eachother right

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

266 TESOL QUARTERLY

As in Text 1 the context is a multimodal one with the matrix beingbuilt up on the board providing another way of representing theinformation being constructed through the discourse To show thenature of the mediating work being done by the teacher the transcrip-tion is shown in two elds representing talk about language and talkabout magnets (science)

The teacher talk encapsulates two elds science and language itselfThe relationship between these two in the progression of the discourse isa signi cant factor in the teacherrsquos scaffolding The eld of languagehere involves talk about language that is intended to help studentsunderstand both the concept of a generalisation (see eg the use ofemphasis in Turns 9 and 12) and its wording (ldquoso wersquoll say rdquo) Theteacher shows students how to produce their understandings in moreabstract waysmdashways that are not dependent on a here-and-now contextAs the dialogue progresses and students begin to master the structure ofthe generalisation without help however talk about language progres-sively drops out of the teacherrsquos talk Jenniferrsquos response is a completephrase that though syntactically parallel with the previous responsecontains different information and indicates how she has appropriatedand transferred new learning Scaffolding provided by the talk aboutlanguage is a temporary support and as the layout of Text 5 indicates isprogressively discarded as students show they can produce the targetlanguage alone

Text 5 also illustrates that the pedagogical signi cance of textsmdashhowthey are readmdashcannot be uncovered without a consideration of theirplace within the ongoing discourse that has been built up over time Anyinteractional sequence is simply an excerpt of a much larger piece ofdiscourse namely the total discourse of the subject or topic to datereferred to by Edwards and Mercer (1987) as the long conversation of theclassroom Thus the meaning and nature of a particular interaction canonly be understood in terms of the situational context and ongoing

FIGURE 6

Matrix on Board

Attract Repel

NS

SN

SS

NN

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 267

discourse in which it occurs Though in isolation this text has shades ofa drill and practice it is a far cry from an empty functional drill Ratherit is an example of how within the context of the long conversation inwhich childrenrsquos own observations have been a starting point theteacher mediates for learners ways of recontextualising their personallearning taking a further step toward the language of the sciencecurriculum

CONCLUSIONS AND IMPLICATIONS

Examining the process of mediation across a mode continuum in thisclassroom provided a way of focusing on the building of linguisticbridges between learner language and the target register which must beintegral to any ESL program in which new and unfamiliar ways of usinglanguage are also constructing new content knowledge Describing thesebridges in linguistic terms contributes to an understanding of whatmight constitute effective interaction in such a context In these texts asI have shown teachers mediate language and learning in several waysmode shifting through recasting signaling to the students how they canself-reformulate indicating where a reformulation is needed but hand-ing this task over to the learner and modeling alternative ways ofrecontextualising personal knowledge

Central to such interactions is the notion of contingency the way anadult judges the need and quality of assistance required by the learneron the basis of moment-to-moment understanding Contingency re-quires interactants to be oriented toward collaboration Contingentdiscourse is anchored within the shared agenda of the participants (vanLier 1996) and requires what Wells (1986) refers to as a rich interpretationof a learnerrsquos attempt to communicate ideas Van Lier refers to the Janus-like quality of such interactions In these examples the discourse looksboth backward to the familiar known or given and forward to thelanguage associated with curriculum learning or the new Such contin-gent interactions evidence an intersubjectivity and sharedness of per-spectives that set up expectancies for what may come next validate(value and respect) the preceding and the following utterances andhelp ensure continued engagement (van Lier 1996 p 184)

The teachersrsquo responses also re ect the ZPD in that they take as astarting point what the student can contribute but extend it by scaffold-ing the language the student will later be expected to use As van Lier(1996) suggests ldquoin order to learn a person must be active and theactivity must be partly familiar and partly new so that attention can befocused on useful changes and knowledge can be increasedrdquo (p 171)The closeness of t between student and teacher contributions offers an

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

268 TESOL QUARTERLY

insight into how far the studentsrsquo ZPD is informing the teacherrsquosresponse As suggested earlier the degree of difference between studentand teacher talk can perhaps more generally illuminate the appropriate-ness of teachersrsquo responses in classroom discourse (especially importantfor ESL students in mainstream classrooms largely driven by particularcurriculum content) Too close a match between teacher and studentcontributions would suggest that students are not being provided with acontext in which learning will occur because they will have no access tounknown language too great a difference may lead to studentsrsquo failing tounderstand the teacherrsquos discourse (see Lemke 1990 for analyses of aphysics classroom where the teacherrsquos and studentsrsquo understandings failto converge)

Exploring the ways in which students and teachers co-constructmeaning also shifts pedagogical questions away from the well-worndebate around traditionalteacher-fronted versus progressivestudent-centred pedagogies toward a focus on the nature of the discourse itselfand its mediating role in the broader knowledge framework of thecurriculum The texts show how language learning is essentially a socialrather than an individual endeavour Meanings are constructed betweenrather than within individuals and are shaped by the social activity inwhich they arise and the collaborative nature of the interaction AsCummins (1996) points out one implication of this view of learning andone that is particularly relevant in an ESL context is that learnersrsquoachievements (and equally educational failure) should not be seen assolely the result of a learnerrsquos innate ability or background but also as themeasure of the nature of the interactions between teacher and learner

Finally the interactions discussed here are not unusual examplesSimilar interactions between teachers and students probably occur dailythroughout hundreds of classrooms without teachers being explicitlyaware of the nature of their responses Much effective teaching mayresult from the intuitive rather than the explicit knowledge that teachershold such knowledge is not necessarily stored in propositional form(Heap 1995) But unless such knowledge is propositionalised by beingarticulated it cannot be re ected on or fed back into the classroom andinto curriculum design One of the strengths of qualitative approachesfor educational research is that they have the potential to recast teachersrsquoinnate understandings as educationally usable propositions theorisingfrom practice can lead to usable theory for future teacher development(van Lier 1994 p 338)

Teacher-student interaction in the content-based ESL classroom isone area in which such research would be pro table Further researchneeds to focus on analysing linguistically the mechanisms through whichteachers mediate between the language of their students and thelinguistic demands of the school curriculum The model of language-in-

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

MEDIATING LANGUAGE LEARNING 269

context central to systemic linguistics offers a way to do this because ofits focus on the relationship between context meaning and language itsconcern with how meanings are made and its complementarity with asociocultural perspective on teaching and learning Register analysis forexample can show how teacher-student discourse mediates shifts in eld(topic) tenor (the relationships constructed between teacher and stu-dent) and mode (the textual resources whereby everyday languagebecomes less dependent on the immediate situation) Analysis of class-room language that draws on this model has the potential to lead tomore explicit and linguistically oriented descriptions or reconceptuali-sations of constructs of SLA such as comprehensible input and learneroutput (and the relationship between them) negotiation and recastingand will help describe more precisely the relationship between classroominteractions and language development Most important it will groundsuch research within the dynamic social context in which ESL schoollearning is played out In addition studies such as the one described inthis article suggest that teacher education courses might usefully paymore attention to developing teachersrsquo understandings of the role ofdiscourse in mediating learning The kind of analysis included hereoffers one means of achieving this

THE AUTHOR

Pauline Gibbons teaches postgraduate TESOL courses at University of TechnologySydney She has worked as a teacher educator in Hong Kong Laos the UnitedKingdom Iran and South Africa among other locations Her research interests arein ESL pedagogy and the role of classroom discourse in L2 development

REFERENCES

Appel G amp Lantolf J (1994) Speaking as mediation A study of L1 and L2 recalltasks The Modern Language Journal 78 437ndash452

Baynham M (1993) Literacy in TESOL and ABE Exploring common themes OpenLetter 2(2) 4ndash16

Biber D (1986) Spoken and written textual dimensions in English Resolving thecontradictory ndings Language 62 384ndash414

Brilliant-Mills H (1993) Becoming a mathematician Building a situated de nitionof mathematics Linguistics and Education 5 301ndash334

Brooks F (1992) Communicative competence and the conversation course A socialinteraction perspective Linguistics and Education 4 219ndash246

Cazden C (1988) Classroom discourse The language of teaching and learning Ports-mouth NH Heinemann

Christie F (1995) Pedagogic discourse in the primary school Linguistics andEducation 7 221ndash242

Collier V (1989) How long A synthesis of research in academic achievement in asecond language TESOL Quarterly 23 509ndash531

270 TESOL QUARTERLY

Cummins J (1984) Bilingualism and special education Issues in assessment and pedagogyClevedon England Multilingual Matters

Cummins J (1988) From multicultural to anti-racist education In T Skutnabb-Kangas amp J Cummins (Eds) Minority education From shame to struggle (pp 126ndash157) Clevedon England Multilingual Matters

Cummins J (1996) Negotiating identities Education for empowerment in a diverse societyOntario California Association for Bilingual Education

Cummins J (2000) Language power and pedagogy Bilingual children in the crossreClevedon England Multilingual Matters

Davis K (1995) Qualitative theory and methods in applied linguistics TESOLQuarterly 29 427ndash453

Derewianka B (1990) Exploring how texts work Sydney Australia Primary EnglishTeaching Association

Donato R (1994) Collective scaffolding in second language learning In J Lantolfamp G Appel (Eds) Vygotskian approaches to second language learning research (pp 33ndash56) Norwood NJ Ablex

Donato R (2000) Sociocultural contributions to understanding the foreign andsecond language classroom In J Lantolf (Ed) Sociocultural theory and secondlanguage learning (pp 27ndash50) Oxford Oxford University Press

Driver R (1994) The fallacy of induction in science teaching In R Levinson (Ed)Teaching science (pp 41ndash48) London Routledge

Edwards D amp Mercer N (1987) Common knowledge The development of understandingin the classroom London Methuen

Eggins S (1994) An introduction to systemic functional linguistics London PinterEngerstrom Y amp Middleton D (Eds) (1996) Cognition and communication at work

Cambridge Cambridge University PressFloriani A (1993) Negotiating what counts Roles and relationships texts and

contexts content and meaning Linguistics and Education 5 241ndash273Gibbons P (1998) Classroom talk and the learning of new registers in a second

language Language and Education 12 99ndash118Gibbons P (2001) Learning a new register in a second language In C Candlin amp

N Mercer (Eds) English language teaching in its social context (pp 258ndash270)London Routledge

Gibbons P (in press) Bridging discourses in the ESL classroom London ContinuumHall J (1995) (Re)creating our worlds with words A sociohistorical perspective of

face-to-face interaction Applied Linguistics 16 206ndash232Hall J (1998) Differential teacher attention to student utterances The construction

of different opportunities for learning in the IRF Linguistics and Education 9287ndash311

Halliday M (1985) An introduction to functional grammar London Edward ArnoldHalliday M (1993) Towards a language-based theory of learning Linguistics and

Education 5 93ndash116Halliday M amp Hasan R (1985) Language context and text Geelong Victoria

Australia Deakin University PressHalliday M amp Martin J (1993) Writing science Literacy and discursive power London

Falmer PressHammersley M (1994) Introducing ethnography In D Graddol J Maybin amp

B Stierer (Eds) Researching language and literacy in social context (pp 1ndash17)Clevedon England Multilingual Matters

Hammond J (1990) Is learning to read and write the same as learning to speak InF Christie (Ed) Literacy for a changing world (pp 26ndash53) Victoria AustraliaACER Press

MEDIATING LANGUAGE LEARNING 271

Heap J (1995) The status of claims in ldquoqualitativerdquo educational research CurriculumInquiry 25 271ndash292

Heras A (1994) The construction of understanding in a sixth grade classroomLinguistics and Education 5 275ndash299

Kozulin A (1998) Psychological tools A sociocultural approach to education CambridgeMA Harvard University Press

Kramsch C (1993) Context and culture in language teaching Oxford Oxford Univer-sity Press

Kress G (1982) Learning to write London Routledge amp Kegan PaulLantolf J P (Ed) (2000) Sociocultural theory and second language learning Oxford

Oxford University PressLee C amp Smogarinsky P (2000) Vygotskian perspectives on literacy research Cam-

bridge Cambridge University PressLemke J (1990) Talking science Language learning and values Norwood NJ AblexLin L (1993) Language of and in the classroom Constructing the patterns of social

life Linguistics and Education 5 367ndash409Lyster R (1998) Recasts repetition and ambiguity in second language classroom

discourse Studies in Second Language Acquisition 20 51ndash58Maley Y Candlin C Crichton J amp Koster P (1995) Orientations in lawyer-client

interviews Forensic Linguistics 2 42ndash45Martin J (1984) Language register and genre In F Christie (Ed) Children writing

A reader (pp 21ndash30) Geelong Victoria Australia Deakin University PressMartin J (1990) Literacy in science Learning to handle text as technology In

F Christie (Ed) Literacy for a changing world (pp 79ndash117) Hawthorn VictoriaAustralia ACER Press

Martin J amp Veel R (1998) Reading science London RoutledgeMaybin J Mercer N amp Stierer B (1992) Scaffolding learning in the classroom In

K Norman (Ed) Thinking voices The work of the national oracy project (pp 186ndash195) London Hodder amp Stoughton

McCormick D amp Donato R (2000) The discourse of teacher questions asscaffolding in an integrated ESL classroom In J K Hall amp L S Verplaeste (Eds)Second and foreign language learning through classroom interaction (pp 183ndash201)Mahwah NJ Erlbaum

McKay P Davies A Devlin B Clayton J Oliver R amp Zammit S (1997) Thebilingual interface project report Canberra Australia Department of EmploymentEducation Training amp Youth Affairs

Mercer N (1994) Neo-Vygotskian theory and classroom education Clevedon EnglandMultilingual Matters

Mercer N (1995) The guided construction of knowledge Talk amongst teachers andlearners Clevedon England Multilingual Matters

Moll L (1994) Vygotsky and education Instructional implications and applications ofsociohistorical psychology Cambridge Cambridge University Press

Newman D Grif n P amp Cole M (1989) The construction zone Working for cognitivechange in school Cambridge Cambridge University Press

Ohta A (1995) Applying sociocultural theory to an analysis of learner discourseLearner-learner collaborative interaction in the zone of proximal developmentIssues in Applied Linguistics 6 93ndash121

Ohta A (1999) Interactional routines and the socialisation of interactional style inadult learners of Japanese Journal of Pragmatics 31 1493ndash1512

Ohta A (2000) Rethinking interaction in SLA Developmentally appropriateassistance in the zone of proximal development and the acquisition of L2

272 TESOL QUARTERLY

grammar In J P Lantolf (Ed) Sociocultural theory and second language learning (pp51ndash78) Oxford Oxford University Press

Oliver R (1995) Negative feedback in child NS-NNS conversation Studies in SecondLanguage Acquisition 17 459ndash481

Painter C (1984) Into the mother tongue A case study of early language developmentLondon Pinter

Painter C (1998) Learning through language in early childhood London CassellPica T (1988) Interlanguage adjustments as an outcome of NS-NNS negotiated

interaction Language Learning 38 45ndash73Pica T (1994) Research on negotiation what does it reveal about second language

learning conditions processes and outcomes Language Learning 44 493ndash527Pica T Holliday L Lewis N amp Morgenthaler L (1989) Comprehensible output

as an outcome of linguistic demands on the learner Studies in Second LanguageAcquisition 11 63ndash90

Snow C (1986) Conversations with children In P Fletcher amp P Garman (Eds)Language acquisition Studies in rst language development (pp 363ndash375) CambridgeCambridge University Press

Swain M (1985) Communicative competence Some roles of comprehensible inputand comprehensible output in its development In S Gass amp C Madden (Eds)Input in second language acquisition (pp 235ndash253) Cambridge MA NewburyHouse

Swain M (1995) Three functions of output in second language learning In G Cookamp B Seidlhofer (Eds) Principle and practice in applied linguistics Studies in honour ofH G Widdowson (pp 125ndash144) Oxford Oxford University Press

Swain M (2000) The output hypothesis and beyond Mediating acquisition throughcollaborative dialogue In J P Lantolf (Ed) Sociocultural theory and second languagelearning (pp 97ndash114) Oxford Oxford University Press

Swain M amp Lapkin S (1995) Problems in output and the cognitive processes theygenerate A step towards second language learning Applied Linguistics 16 371ndash391

Toohey K (2000) Learning English at school Identity social relations and classroompractice Clevedon England Multilingual Matters

van Lier L (1988) The classroom and the language learner Harlow England Longmanvan Lier L (1994) Forks and hope Pursuing understanding in different ways

Applied Linguistics 15 328ndash346van Lier L (1996) Interaction in the language curriculum Awareness autonomy and

authenticity London Longmanvan Lier L (2000) From input to affordance In J P Lantolf (Ed) Sociocultural

theory and second language learning (pp 245ndash259) Oxford Oxford University PressVygotsky L (1978) Mind in society The development of higher psychological processes

Cambridge MA Harvard University PressVygotsky L (1981) The genesis of higher mental functions In J Wertsch (Ed) The

concept of activity in Soviet psychology (pp 144ndash188) White Plains NY SharpeVygotsky L (1986) Thought and language Cambridge MA MIT PressWebster A Beveridge M amp Reed M (1996) Managing the literacy curriculum

London RoutledgeWells G (1985) Language development in the pre-school years Cambridge Cambridge

University PressWells G (1986) The meaning-makers Children learning language and using language to

learn Portsmouth NH HeinemannWells G (1993) Reevaluating the IRF sequence A proposal for the articulation of

MEDIATING LANGUAGE LEARNING 273

theories of activity and discourse for the analysis of teaching and learning in theclassroom Linguistics and Education 5 1ndash37

Wells G (1996) Using the tool-kit of discourse in the activity of learning andteaching Mind Culture and Language 3 74ndash101

Wells G (1999) Dialogic inquiry Towards a sociocultural practice and theory ofeducation New York Cambridge University Press

Wells G amp Chang-Wells G (1992) Constructing knowledge together Classrooms ascenters of inquiry and literacy Portsmouth NH Heinemann

Wong-Fillmore L (1985) When does teacher talk work as input In S Gass ampC Madden (Eds) Input in second language acquisition (pp 17ndash50) Rowley MANewbury House

Wood D Bruner J amp Ross G (1976) The role of tutoring in problem solvingJournal of Child Psychology and Psychiatry 17 89ndash100