No title

APPLIED LINGUISTICS Volume 29 Number 4 December 2008 CONTENTS Articles Contesting ‘Language’ as ‘Heritage’: Negotiation ...

Author: Zuengler | J and Hyland | K

24 downloads 440 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

APPLIED LINGUISTICS Volume 29 Number 4 December 2008 CONTENTS Articles Contesting ‘Language’ as ‘Heritage’: Negotiation of Identities in Late Modernity ADRIAN BLACKLEDGE and ANGELA CREESE Routine Trouble: How Preschool Children Participate in Multilingual Instruction ¨ RK-WILLEŃ POLLY BJO Symmetries and Asymmetries of Age Effects in Naturalistic and Instructed L2 Learning CARMEN MUNÕZ Textual Enhancement of Input: Issues and Possibilities ZHAOHONG HAN, EUN SUNG PARK and CHARLES COMBS

533

555

578 597

The Cultural Productions of the ESL Student at Tradewinds High: Contingency, Multidirectionality, and Identity in L2 Socialization STEVEN TALMY Language Ecology in Multilingual Settings. Towards a Theory of Symbolic Competence CLAIRE KRAMSCH and ANNE WHITESIDE

645

A Register Approach to Teaching Conversation: Farewell to Standard English? ¨ HLEMANN CHRISTOPH RU

672

Form-focused Instruction in Second Language Vocabulary Learning: A Case for Contrastive Analysis and Translation BATIA LAUFER and NANY GIRSAI

694

FORUM The Significance of First Language Development in Five to Nine Year Old Children for Second and Foreign Language Learning ROBERT VANDERPLANK

717

REVIEWS H-Dirksen Bauman, Jennifer Nelson, and Heidi Rose: Signing the Body Poetic. Essays on American Sign Language Literature MARION BLONDEL

723

619

Ivana Markova´, Per Linell, Miche`le Grossen, and Anne Salazar Orvig: Dialogue in Focus Groups: Exploring Socially Shared Knowledge GREG MYERS Simon Gieve and Ines Miller: Understanding the Language Classroom ASSIA SLIMANI-ROLLS Sandra Mollin: Euro-English: Assessing Variety Status. Jennifer Jenkins: English as a Lingua Franca: Attitude and Identity LUKE PRODROMOU

726 728

732

NOTES ON CONTRIBUTORS

737

INDEX TO VOLUME 29

741

Applied Linguistics 29/4: 533–554 ß Oxford University Press 2008 doi:10.1093/applin/amn024 Advance Access published on 21 July 2008

Contesting ‘Language’ as ‘Heritage’: Negotiation of Identities in Late Modernity ADRIAN BLACKLEDGE and ANGELA CREESE University of Birmingham, UK

with TAS° KIN BARAÇ, ARVIND BHATT, SHAHELA HAMID, LI WEI, VALLY LYTRA, PETER MARTIN, CHAO-JUNG WU and DILEK YAG˘CIOG˘LU In this paper we question key terms which appear frequently in discussions of language teaching and learning: ‘language’ and ‘heritage’. The paper draws on empirical data from one of four linked case studies in a larger project funded by the UK Economic and Social Research Council (ESRC), ‘Investigating Multilingualism in Complementary Schools in Four Communities’ (RES-000-23-1180). In our analysis we argue that the relationships between ‘language’ and ‘heritage’, far from being straightforward, are complex in the way they play out in classroom interactions. The data raise a number of questions in our attempts to understand how the linguistic practices of students and teachers in Bengali schools are used to negotiate young people’s multilingual and multicultural identities. First, participants articulate attitudes and values which raise questions about what constitutes ‘language’. Second, participants express views and attitudes, and perform interactional practices, which raise questions about what constitutes ‘heritage’. Our analysis finds that multilingual young people in complementary school classrooms use linguistic resources in sophisticated and creative ways to negotiate subject positions which appear to contest and subvert schools’ attempts to impose upon them ‘heritage’ identities (Creese, A., A. Bhatt, N. Bhojani, and P. Martin. 2006. ‘Multicultural, heritage and learner identities in complementary schools,’ Language and Education 20/1: 23–44).

INTRODUCTION This paper reports part of a research project which investigates how the linguistic practices of students and teachers in Bengali schools in the United Kingdom are used to negotiate young people’s identities. The analysis of data presented here raises questions in two areas. First, participants appeared to differ from each other in their views of what constitutes (a) language, and what ‘counts’ as (a) language. Second, we are led by our participants to consider the relationships between ‘language’ and ‘heritage’, and the role of teaching and learning language(s) in the reproduction of that ‘heritage’. The complexity and sophistication

534

NEGOTIATION OF IDENTITIES IN LATE MODERNITY

of the young people’s responses to the teaching and learning of ‘language(s)’ led us to question by what means ‘the legacy of history is appropriated’ (Bourdieu 2000: 151). Is ‘heritage’ straightforwardly reproduced where the learner is born to linguistic, social, and environmental norms which are typical of urban late modernity, whereas the ‘heritage’ was associated with rural poverty? These questions about social reproduction for young people in the United Kingdom raise broader questions about what constitutes ‘(a) language’, and what counts as ‘heritage’ in late modernity. Before discussing the data, we reflect theoretically on these questions relating to ‘language’ and ‘heritage’.

‘LANGUAGE’ Heller (2007) proposes four sets of concepts in the critical analysis of languages in society. First, she argues that rather than treating notions of ‘community’, ‘identity’, and ‘language’ as though they were natural phenomena, they should be understood as social constructs. Specific or single categorizations therefore cannot be attached to an individual based on their ‘ethnicity’, or ‘language’. Second, Heller refers to the work of Giddens (1984) to consider language as a set of resources which are socially distributed, but not necessarily evenly. The third set of concepts holds that this uneven distribution of resources is the product of political and economic processes, enabling us to ask questions about what linguistic resources are assigned what value, and with what consequences (Gumperz 1982). The final set of concepts considers the discourses which inscribe value (or its lack) to particular linguistic forms and practices. In summary, Heller views language(s) as: sets of resources called into play by social actors, under social and historical conditions which both constrain and make possible the social reproduction of existing conventions and relations, as well as the production of new ones. (Heller 2007: 15) What Heller calls ‘the messiness of actual usage’ (2007: 13) can only be understood in relation to histories, power, and social organization. Conversely, structural analysis must include accounts of actual linguistic practices, which at times may differ from those we might expect. Our multilingual participants’ beliefs, attitudes, and practices in relation to ‘language’ resonated with Garcia’s (2007: xii) account that languages are not hermetically sealed units. The linguistic practices of Garcia’s students in New York bore very little relation to the ‘standard English’ of school texts or the ‘standard Spanish’ that was supposed to be linked to their ‘identity’. Rather, our data suggest, in line with the recent proposition of Makoni and Pennycook (2007), that the notion of languages as separate, discrete entities, and ‘countable institutions’ (2007: 2) is a social construct. Bourdieu (1991) proposed that: language is itself a social artefact invented at the cost of a decisive indifference to differences which reproduces on the level of

ADRIAN BLACKLEDGE and ANGELA CREESE

535

the region the arbitrary imposition of a unique norm. (Bourdieu 1991: 287) Makoni and Pennycook argue for a critical historical account which demonstrates that, through the process of classification and naming, languages were ‘invented’ (2007: 1). They add that, in direct relation with the invention of languages, ‘an ideology of languages as separate and enumerable categories was also created’ (2007: 2). Makoni and Pennycook point in particular to the naming of languages such as ‘Bengali’ and ‘Assamese’ as the construction of ‘new objects’ (2007: 10). Thus languages cannot be viewed as discrete, bounded, impermeable, autonomous systems. Our research participants, all at first glance of the same ‘ethnic’ and ‘linguistic’ group, not only disagreed with each other about what constituted a ‘language’, they also disagreed with each other about where a ‘language’ began and ended, and about the value that could be assigned to a particular set of linguistic resources. Makoni and Pennycook propose that such ‘local knowledge’ is crucial to our understanding of language: We are arguing for an understanding of the relationships between what people believe about their language (or other people’s languages), the situated forms of talk they deploy, and the material effects—social, economic, environmental—of such views and use. (Makoni and Pennycook 2007: 22) This interrelationship between what people believe about language and languages, and the way they access and make use of linguistic resources, provides a further focus to our analysis. If languages are invented, and languages and identities are socially constructed, we nevertheless need to account for the fact that at least some language users, at least some of the time, hold passionate beliefs about the importance and significance of a particular language to their sense of ‘identity’. It is now well established in contemporary sociolinguistics (Harris 2006; Rampton 2006) that one ‘language’ does not straightforwardly index one subject position, and that speakers use linguistic resources in complex, sophisticated ways to perform a range of subject positions, sometimes simultaneously. However, whilst accepting this, May (2001, 2005: 330) argues that ‘historically associated languages continue often to hold considerable purchase for members of particular cultural or ethnic groups in their identity claims’. For some of the people we spoke to in the course of our research, a ‘language’ held powerful connotations in terms of their sense of belonging and selfhood. It is evident from our data that for some people, in some circumstances, ‘particular languages clearly are . . . important and constitutive factor of their individual, and at times, collective identities’ (May 2005: 330). In the context of research on ‘heritage’ language education, it is almost a truism that learning the ‘heritage’ language ‘plays a critical role in the process of children’s identity formation’ (Nicholls 2005: 164). Whilst it is certainly an oversimplification to treat certain languages as ‘symbols’ or ‘carriers’ of ‘identity’, we are obliged

536


to take account of what people believe about their languages, to listen to how they make use of their available linguistic resources, and to consider the effects of their language use—even where we believe these ‘languages’ to be inventions.

‘HERITAGE’ Bourdieu and Passeron (1979) proposed that all teaching implicitly presupposes a body of knowledge, skills, and modes of expression which constitute the heritage of the cultivated classes. In our classroom observations and recordings, and in our participants’ statements in interviews, there was a clear sense that the teaching of ‘language’ was inexorably intertwined with the teaching of ‘heritage’. Many of our participants used the term ‘culture’ to refer to those elements of Bengali/Bangladeshi life and history which they wished to transmit through complementary schooling. In our analysis we interpret this as ‘heritage’, distinguishing ‘heritage’ from ‘culture’. Whereas ‘heritage’ refers to elements of past experience which a group deliberately sets out to preserve and pass on to the next generation, ‘culture’ is ‘reproduced and emerges in people’s activity together—it exists in the processes and resources involved in situated, dialogical, sense-making’ (Rampton 2006: 20). In recent times the scope of definitions of ‘heritage’ has broadened considerably from concern for the preservation of buildings and historical sites to include historical areas, towns, environments, social factors, and ‘intangible heritage’ (Ahmad 2006: 299; Smith 2006: 54). UNESCO’s Convention for the Safeguarding of the Intangible Cultural Heritage defines ‘intangible cultural heritage’ as: The practices, representations, expressions, knowledge, skills—as well as the instruments, objects, artefacts and cultural spaces associated therewith—that communities, groups and, in some cases, individuals, recognise as part of their cultural heritage. This intangible cultural heritage, transmitted from generation to generation, is constantly recreated by communities and groups in response to their environments, their interaction with nature and their history, and provides them with a sense of identity and continuity, thus promoting respect for cultural diversity and human creativity. (UNESCO 2003: Article 2:2) Patrick (2007) points out that appeals for the protection of forms of ‘intangible heritage’ have played an important role in campaigns for language rights. Whether we are dealing with traditional definitions of ‘tangible’ or ‘intangible’ heritage, we are engaging with sets of values and meanings, including emotion, memory, and shared knowledge (Smith 2006). ‘Heritage’ describes sets of shared values and collective memories that are ‘constructed as a ‘birthright’ and are expressed in distinct languages and through other cultural performances’ (Peckham 2003: 1). Pearson and Sullivan (2007: 208) suggest that heritage resources may have a ‘special value to minority groups in the


537

community’, who have a particular interest in their own history. Regardless of the ‘management’ of heritage resources, subordinate groups ‘often choose to mobilise a ‘‘strategic essentialism’’ as a political tool’ (Stanton 2005: 416). From our participants we heard at times that certain sets of linguistic resources were believed to function as threads of association with historic contexts. Sets of resources often come to represent abstract notions such as sense of place, community, or belonging (Smith 2004). ‘Heritage’ can be thought of as the preservation of a potential loss (Peckham 2003: 1), ‘anything that someone wishes to conserve or to collect, and to pass on to future generations’ (Howard 2003: 6). Bourdieu and Passeron (1979: 25) suggest that ‘inheritance always implies the danger of squandering the heritage’. However, it cannot be assumed that the preservation and transmission of ‘heritage’ is straightforward. Simply the process of ‘passing on’ resources will alter them. Tunbridge and Ashworth (1996: 92) argue that there is rarely a simple relationship between a group of people and ‘heritage’ resources: ‘The same piece of heritage can be interpreted and received by different groups in quite different ways’. Rather than being a static entity, ‘heritage’ is a ‘process or performance that is concerned with the production and negotiation of cultural identity, individual and collective memory, and social and cultural values’ (Smith 2007: 2). Heritage as a process of meaning-making may ‘help us bind ourselves, or may see us become bound to, national or a range of sub-national collectives or communities’ (Smith 2006: 66) as particular resources come to act as powerful symbols of, or mnemonics for, the past (Lipe 2007). Smith (2006: 3) proposes that the idea of ‘heritage’ is ‘used to construct, reconstruct, and negotiate a range of identities and social and cultural values and meanings in the present’. She argues that ‘heritage’ is a set of practices involved in the construction and regulation of values, a discourse about negotiation, about using the past, and collective and individual memories, to negotiate new ways of being and to perform identities. People engage with ‘heritage’, appropriate it, and contest it (Harvey 2007). ‘Heritage’ may become a site at which identities are contested rather than imposed unproblematically. That is, those who seek to preserve and pass on certain sets of resources may find that the next generation either rejects imposed subject positions, contests the validity or significance of resources, or appropriates them for other purposes. For Bourdieu ‘heritage’ is reproduced through ‘class’ and ‘education’, in the reproduction of ‘distinction’, ‘an unacquired merit which justifies unmerited attainment, namely heritage’ (Bourdieu and Darbel 1991: 110). Bourdieu (1993: 299) argues that in education there is an assumption of a community of values between pupil and teacher which occurs where the system ‘is dealing with its own heirs to conceal its real function, namely, that of confirming and consequently legitimizing the right of the heirs to the cultural inheritance’. He further argues that: Only when the heritage has taken over the inheritor can the inheritor take over the heritage. And this appropriation of the

538


inheritor by the heritage, the precondition for the appropriation of the heritage by the inheritor (which has nothing inevitable about it), takes place under the combined effect of the conditionings inscribed in the position of inheritor and the pedagogic action of his predecessors, themselves possessed possessors. (Bourdieu 2000: 152) In our data the teaching and learning of ‘heritage’ or ‘community’ languages (Hornberger 2005) act as sites at which ‘heritage’ values may be transmitted, accepted, contested, subverted, appropriated, and otherwise negotiated. These are sites for the negotiation of identities, for the acquisition and performance of sets of linguistic resources which are called into play by social actors under very particular social and historical conditions (Wiley 2005, 2007). These conditions may both constrain and make possible the reproduction of existing conventions and relations, as well as the production of new ones (Heller 2007).

METHODS The research project consisted of four interlocking case studies with two researchers working in two complementary schools in each of four communities. The case studies focused on Gujarati schools in Leicester, Turkish schools in London, Cantonese and Mandarin schools in Manchester, and Bengali schools in Birmingham. The present paper focuses on data collected in and around the Bengali schools in Birmingham. Complementary schools, also known as ‘supplementary schools’, ‘heritage language schools’, or ‘community language schools’, provide language teaching for young people in a nonstatutory setting. Bengali complementary schools in Birmingham are managed and run by local community groups on a voluntary basis, usually in hired or borrowed spaces, with few resources. They cater for children between 4 and 16 years of age, and operate mainly in the evenings and at weekends. The students’ families had migrated from the Sylhet region of Bangladesh. One of the specific aims of the research project was to investigate how the linguistic practices of students and teachers in complementary schools are used to negotiate young people’s multilingual and multicultural identities. Each case study identified two complementary schools in which to observe, record, and interview participants. The classes ran for between two and three hours, either in the evening or at the week-end. After four weeks observing in classrooms using a ‘team field notes’ approach, two key participant children were identified in each school. In the Bengali schools the key participant children were all 10 years old. These children were audio-recorded during the observed classes, and also for 30 minutes before coming to the class and after leaving class. Stakeholders in the schools were interviewed, including teachers and administrators, and the key participant children and their parents. In all we collected 192 hours of audio-recorded interactional data, wrote 168 sets of field notes, made 16 hours of video-recordings, and interviewed 66 key stakeholders.


539

A more detailed account of the methods used to collect documentary and home-based data are outlined in Creese et al. (2008).

‘LOCAL KNOWLEDGE’ ABOUT ‘LANGUAGE’ AND ‘HERITAGE’ The founder and administrator of one of the two schools (school A) made a forceful and emotional statement following an interview question which queried the rationale for teaching Bengali to children in Birmingham: ei bhaashar jonno 1952 te amaar theke dosh haath dure Barkat, Salam maara jaae 5 because of this language in 1952 ten yards away from me Barkat and Salam were killed4 1952 te 5in 19524I was also a student in year 10. From Sylhet to Dhaka was 230 miles we marched there Sylhet to Dhaka 230 miles with slogans. We want our mother language it is a raashtro bhasha 5state language4. How I will forget about my mother language? My brothers gave their life for this language. I will never forget it while I’m alive. (administrator interview)1 Throughout the paper we are mindful of Pavlenko’s (2007: 176) argument that interview or narrative data can not be treated as ‘truth’ or ‘reality itself’. Rather, in line with Pavlenko, we are ‘sensitive to the fact that speakers use linguistic and narrative resources to present themselves as particular kinds of individuals’ (2007: 177). We constantly saw individual participants positioning themselves in relation to the ‘ethnic, linguistic, and cultural loyalties’ (2007: 177) which they chose to emphasize. For the school administrator the ‘mother language’ was a vital symbol of the founding of the Bangladeshi nation. More than 50 years earlier he had witnessed the incident in which the ‘language martyrs’ were killed while demonstrating against the imposition of Urdu as the national language by West Pakistan, and these events seemed to have informed his view that British-born children of Bangladeshi heritage should learn and maintain the Bengali language. The historic incident which marks the Bangladeshi calendar as ‘Ekushey February’, continues to be celebrated as a key moment in the collective memory of the Bangladeshi nation, and in the Bangladeshi community in UK (Gard’ner 2004). One of the senior teachers in the same school argued that learning Bengali was associated with maintaining a knowledge of Bangladeshi ‘roots’: ‘We may have become British Bangladeshi or British Indians but we don’t have fair skin and we cannot mix with them. We have our own roots and to know about our roots we must know our language’. For both of these Bangladeshi-born men, teaching and learning Bengali was an important means of reproducing their ‘heritage’ in the next generation. Many of the students’ parents agreed. One mother typically told us that it was important that her children should be able to speak Bengali because: Bengali is our mother land, where we come from; really we come from Bangladesh. Even if you are born in this country it doesn’t matter, we need to know our mother language first. (parent interview)

540


Asked why it was important to learn the language of the ‘mother land’, she said ‘you need to know your side of the story, where your parents come from, you’ve got to know both from this country and the other one’. For her learning the ‘mother language’ was closely associated with learning about the ‘mother land’, and both represented her ‘side of the story’. We heard an explicit rationale from administrators, teachers, and parents that a key aim of the school was for the children to learn Bengali because knowledge of the national language carried features of Bangladeshi/Bengali ‘heritage’.

TEACHING ‘LANGUAGE’ AND ‘HERITAGE’ The rationale of the schools was put into practice in the classroom through a pedagogy which frequently introduced ‘heritage’ content in the context of teaching Bengali. Here ‘heritage’ included narratives of national belonging, and the introduction of national symbols of Bangladesh. In the first example the teacher (T) engages with historical events in the making of the Bangladeshi nation: T:

S: T:

S: T:

S: T: S: T: S: T:

Bangladesher teen taa national day aache, jaatio dibosh 5Bangladesh has three national days, national events4 National day not national anthem independence day etaa Banglae ki bolbe shaadhinota dibosh 5in Bangla it is shaadhinota dibosh4 Ekushey February shohid dibosh 521st February is shohid dibosh4 aage bolo Ekushey February shohid dibosh 5first say 21st February is shohid dibosh4 ekushey February shohid dibosh er pore aashlo shaadhinota dibosh 5after that comes shaadhinotaa dibosh4 independence day, independence day is not Bangla, it is English. Banglae holo 5in Bangla it is4 shaadhinota dibosh chaabbish-e March 526th March4 shaadhinota dibosh chaabbish-e March 526th March4 lastly nine months we fought against Pakistani collaborator language day language day holo ekushey February. Chaabbish March independence day. Sholoi December, after nine months bijoy dibosh 5victory day4 Pakistani occupied army ke aamraa surrender korchi. 5we made the occupied forces of Pakistan surrender their arms4 Al Badr against our independent war ke aamraa chutaaisi


S: T: S: T:

541

5 we chased them out 4 How many national days in Bangladesh? three Bangladesher jaatio dibosh koiti? 5how many national days in Bangladesh?4 teen ti 5three4 Shaadhinota dibosh ebong bijoy dibosh chilo 1971. Bhasha dibosh chilo 1952. Aar bhaasha dibosh kon din chilo 52. Tokhon amraa choto 5independence day and victory day was in 1971. Language day was in 1952. Language day was 52 when we were young4 Inshaallah eta every day jodi aamraa every day discuss kori taahole bhaalo 5by the grace of God it is good if we discuss this every day4 (classroom recording)

The curriculum content here is strongly nationalistic, and appears to have the aim of instilling in the young language learners an understanding of key dates and events in the making of the Bangladeshi nation. The student (S) seems to have some pre-existing knowledge of the historical context, although she is rather tentative in volunteering this. Here the teacher moves comfortably between Bengali and English, translanguaging within and between sentences, and making his final statement in the common Islamic expression ‘Inshallah’, derived from Arabic, together with Bengali and English. A second common feature of teaching ‘heritage’ in the Bengali complementary school classrooms was the introduction of national symbols associated with Bangladesh. In a typical example, the teacher asked a child to draw the English and Bangladeshi national flags on the whiteboard. After playing the Bangladeshi national anthem on his mobile telephone, the teacher continued: T:

Ss: T:

Ei. Eitaai aamaader jaatio shongeet othobaa national anthem. Ekhon aamaader Bangladesher ko-e ektaa jinish aache jaatio bol-e. 5This, this is our national song or national anthem. Now, we have a few things in Bangladesh which are our national symbols4 jaatio shongeet 5national anthem4 jaatio kobi 5national poet4, jaatio ful 5national flower4 baa jaatio, baa national fol 5or national, or national fruit4 baa national paakhi 5or national bird4 Bangladesher jaatio ful ki? 5What is the national flower of Bangladesh?4 [no response] water lily, water lily, water lily Bangla, water lily, shapla. Etaa aamaader jaatio ful 5shapla. This is our national flower4 (classroom recording)

Here the process of teaching Bengali is intimately interwoven with the process of teaching symbolic representations of Bangladesh, as knowledge of the national/cultural symbols, like knowledge of the Bengali language, comes to represent Bengali ‘heritage’.

542


‘LOCAL KNOWLEDGE’ CONTESTING ‘LANGUAGE’ AND ‘HERITAGE’ Despite these powerful discourses and practices which evidenced the teaching of ‘heritage’ through the teaching of ‘language’, the notion that the discourses and practices of the Bengali schools were homogeneous in their ideological orientation to ‘language’ and ‘heritage’ was not borne out in the data. Rather, what people believed about their language (or other people’s languages), and the situated forms of talk they deployed, revealed divergent and contested views about the value and status of particular linguistic resources. Bengali is the language of education and literacy in Bangladesh, and is characterized by diglossia. The two standard varieties of Bangla are Sadhu Bhasha and Cholit Bhasha and regional varieties include Sylheti, from the north-east of Bangladesh. Sylheti is the variety spoken by the vast majority of Bangladeshi immigrants to Britain. Sylheti is often regarded as a modification of standard Bengali which is not intelligible to the people of other districts in Bangladesh (Hamid 2007). Chalmers (1996: 6–7) observes that Bengali and Sylheti are ‘near enough mutually unintelligible’. Chalmers does acknowledge that Sylheti and Bengali are very closely related and speakers of one language or dialect are often exposed to the other, even though they may not speak them. Whilst Bengali is the literate language of Bangladesh, Sylheti is a vernacular variety. When we interviewed the administrators and teachers in the schools they spoke emphatically about the need for children to learn the standard variety. This was frequently held to be oppositional to Sylheti. One of the school administrators was emphatic that Bengali was ‘completely different’ from Sylheti, and that Sylheti should not be allowed to ‘contaminate’ the standard form. He was concerned that Sylheti forms were beginning to appear in the spelling and grammar of Bengali newspapers in the UK, introducing ‘thousands of spelling mistakes—Bengali newspapers I have seen in many places the spelling was wrong, sentence construction was wrong’. For the administrator non-Standard resources were ‘contaminating the language’. He made this point about the necessity for children to learn Standard Bengali: I am always in favour of preserving languages and all these things. But it doesn’t mean that this should contaminate other languages and give this more priority than the proper one. We have to preserve the proper one first, and at the same time we have to encourage them to you know, use their dialect. But we shouldn’t make any compromise between these two. (administrator interview) This was a strongly articulated argument in the data. The administrator of the other school stated that: ‘Bhasha to bolle Bangla bhasha bolte hobe Sylheti kono bhasha naa’


543

5When you talk about language it means Bengali. Sylheti is not a language4 (school administrator interview) For several respondents ‘Bengali’ constituted a more highly valued set of linguistic resources than ‘Sylheti’, and was regarded as the ‘proper’ language. Pujolar (2007: 78), referring to a different socio-historical context, makes the point that language policy may operate to foster knowledge of some languages, ‘but delegitimise or ignore other languages and other forms of multilingual competence and performance’. Patrick (2007: 127) similarly finds that in arguing in support of a particular language, ‘speakers can be locked into fixed or essentialised notions of identity, ‘‘authenticity’’ and place, which provide no recognition of mobile, postcolonial speakers’. It was clear that for some of our respondents not all linguistic resources were equally valued, and while some sets of linguistic resources were considered to be ‘a language’, others were not. In this sense there was a constant re-invention of ‘language’ on the part of some participants. Those who spoke ‘Sylheti’ were often criticized by ‘more educated’ people who spoke ‘Bengali’. They were characterized by the administrator of one of the schools as members of the ‘scheduled’, or ‘untouchable’ caste: people without rights or resources in the Indian sub-continent: Publicraa ki dibe amar aapne especially bujhben amader desher je shob lok aashche ora kon category lok aashchilo, mostly from scheduled caste, gorib, dukhi krishokra aashchilo. oder maa baba o lekha pora interested naa oder chele meye raa o pora lekha interested naa. Oraa baidhitamolok school jete hoe primary school sholo bochor porjonto jete hoe, ei jonne school jaai. 5What will the public contribute? You [the researcher, Shahela Hamid] especially will understand what type of people came from our country. They belonged to the category of scheduled caste, they are the poor, the deprived, farmers. Their parents were not interested in education nor are the children interested. They go to school because it’s compulsory4 (school administrator interview) Here the Sylheti speakers are referred to as the ‘scheduled caste’. Regarded as the least educated group in society, with no resources of any kind, they are the lowest of the low (Borooah 2005; Kijima 2006; Borooah et al. 2007). Here linguistic features were viewed as reflecting and expressing broader social images of people. Irvine and Gal (2000: 37) suggest that ‘participants’ ideologies about language locate linguistic phenomena as part of, and evidence for, what they believe to be systematic behavioural, aesthetic, affective, and moral contrasts among the social groups indexed’. One of the teachers argued that children should learn Bengali for ‘moral reasons’. Irvine and Gal propose that a semiotic process of iconization occurs, in which linguistic features that index

544


social groups appear to be iconic representations of them, as if a linguistic feature depicted or displayed a social group’s inherent nature or essence. Bourdieu and Darbel (1991: 112) argue that some more powerful groups provide ‘an essentialist representation of the division of their society into barbarians and civilized people’. Here the fact of speaking ‘Sylheti’, rather than ‘Bengali’, appeared to index the Sylheti group in particularly negative terms, despite the relative similarities between the ‘Bengali’ and ‘Sylheti’ sets of linguistic resources. Whilst some speakers in our study considered ‘Sylheti’ to be quite different from ‘Bengali’, others regarded the two sets of resources as indistinguishable. As we have seen, there were several instances of participants commenting on the differences between Sylheti and Bengali in terms of social status and value, but not everyone agreed about the extent to which these sets of linguistic resources were distinct. While the administrator of one of the schools argued that Bengali and Sylheti were ‘completely different’, a student’s mother said they were ‘thoraa different’ 5a little different4, while other parents also held this view, saying they were ‘little bit different thaake 5only4’ and even ‘the same’. Here there was clear disagreement about the nature and extent of the differences between the sets of linguistic resources used by the students’ parents at home, and the literate version of the language taught in the complementary school classrooms. That is, there was disagreement about the permeability of the languages. These differences of perception were likely to be ideological. Those who argued that the ‘languages’ were completely different from each other were speakers of the prestige language, unwilling to allow the lower status language to contaminate their linguistic resources. Those who argued that the ‘languages’ were almost the same as each other were speakers of Sylheti, which was held to index the lower status, less educated group. On many occasions the research participants interactionally evidenced their awareness of differences (perhaps mainly in status and value) between ‘Sylheti’ and ‘Bengali’. There was also an awareness of Bengali as the higherstatus language on the part of teachers (‘I talk posh Bengali, and the children can’t understand me’), students, administrators, and parents. The following example was recorded at the dinner table in the family home of one of the students: Mother: Father: Student: Father: Student: Father:

khitaa hoise? Tanvir, khaibaani saatni? 5what is the matter? Tanvir, would you like some relish?4 aaro khoto din thaakbo 5how many more days is that [voice recorder] going be with you?4 aaro four weeks 5four more weeks4 ( ) No they said any. If you talk all English. . . ginni, oh ginni [calling his wife using a Bengali term of endearment]


Mother:

Father:

545

ji, hain go daakso kheno 5yes, dear why are you calling me?4 Tumaar baabaa shuddho bhasha bolen 5your father is speaking the standard language4 paan dibaa 5can I have some paan4 aapne aamaar biyaai kemne 5how are you my relation?4 (home recording)

Here the Sylheti-speaking parents play the roles of Bengali speakers, adopting the airs and graces which they see as characteristic of the Bengali-speaking group. The terms of endearment used here (‘ginni’, ‘hain go’) are forms of parody (Bakhtin 1973, 1984, 1986), exaggerations beyond common usage, as speakers of Bengali are mocked in discourse which represents an inflated sophistication. This brief interaction is situated in a whole hinterland of language ideological beliefs and practices, as the couple acknowledge differences between Bengali and Sylheti as sets of linguistic resources, and the conditions which differentially provide and constrain access to linguistic resources. In parodic discourse the parents introduce into their own voices the exaggerated voice of the Bengali speaker, and that voice clashes with its host, as ‘Discourse becomes an arena of battle between the two voices’ (Bakhtin 1994: 106). Here the impromptu role-play light-heartedly, but not half-heartedly, ‘parodies another’s socially typical . . . manner of seeing, thinking, and speaking’ (Bakhtin1994: 106). In this section we have seen that for some of our participants, some sets of linguistic resources were very considerably privileged above other, similar sets of linguistic resources. While linguistic resources which were described as ‘Standard’, or ‘proper’, or ‘real’, or ‘book’ Bengali had come to represent the ‘heritage’ of the Bangladeshi nation, sets of resources described as ‘Sylheti’ had come to be associated with the uneducated poor, who were held to be uninterested in schooling, and unmotivated. However, we also saw that these distinctions were contested by others, who denied that clear differences existed, and at times made fun of the assumption that these differences were constitutive of differences in social status. That is, our participants represented disagreements about what constituted (a) language, and about the ideological links between speakers and the sets of linguistic resources which they called into play.

NEGOTIATING ‘LANGUAGE’ AND ‘HERITAGE’ The contested nature of the ideological links between sets of linguistic resources and their assumed associations was frequently made visible in the interactional data recorded in the classroom. Here teaching of ‘heritage’ and ‘language’ became sites at which identities were negotiated in discourse (Blackledge and Pavlenko 2001; Pavlenko and Blackledge 2004). Here ‘negotiation of identities’ is understood as ‘an interplay between reflective positioning, that is self-representation, and interactive positioning, whereby

546


others attempt to position or reposition particular individuals or groups’ (Pavlenko and Blackledge 2004). Negotiations here take place interactionally where particular subject positionings are contested. Negotiable identities refer to all identity options which can be contested and resisted by particular individuals and groups, although of course not all identity positions are equally negotiable. How much room for resistance to particular positioning individuals may have depends on each individual situation, the social and linguistic resources available to participants, and the balance of power relations which sets the boundaries for particular identity options. In the following example a new teacher to the class is perhaps not acquainted with the usual norms and expectations of linguistic behaviour in the classroom: S1: T:

S2: S1: S2: T: S2:

miss why can’t we just go home? Bangla-e maato etaa Bangla class 5speak in Bengali this is Bengali class4 khaali English maato to etaa Bangla class khene 5if you speak in English only then why is this the Bengali class?4 miss you can choose I know English why? because tumi Bangali 5because you are Bengali4 my aunty chose it. She speaks English all the time. (classroom recording)

In this interaction the teacher argues that the language of the classroom should be Bengali, proposing a model of learning which is at odds with the children’s usual experience. One of the students (S1) argues that it should be possible to choose which language to speak in a particular context, and is backed up by her friend (S2). When S2 asks why it is necessary to speak Bengali in class, the teacher says ‘because tumi Bangali’. In this English and Bengali phrase the ideology of the school is summed up in the most succinct terms. Bengali should be spoken, and should be learned, argues the teacher, because the children are Bengali (here ‘Bangla’ refers to the Bengali language, while ‘Bangali’ refers to Bengali national and/or ethnic belonging). Ironically, the teacher uses Sylheti to make her point about speaking Bengali. S2 contests the teacher’s point, and in doing so contests the ideology of the school. Reiterating her argument that it should be possible to choose which language to speak, she cites her ‘aunty’, who has chosen to predominantly speak English. The student’s ‘aunty’, herself of Bangladeshi heritage, is offered as an example of someone who has resisted the notion of ‘one-language-equals-one ethnicity/culture’. For S2 language choice is flexible. For the school, in this example at least, language learning is tied to ethnic and national belonging, and is inflexible. We saw many examples in the classroom of students resisting teachers’ attempts to teach them Bengali. In more than one example students mocked their teachers’ pronunciation of English words. Here, though, the children


547

challenge the teacher’s pronunciation of the name of a new child when she arrives at school: S1: T: S1: T:

S1: S2:

She is coming through the front door Jaara [correcting teacher’s pronunciation] Zahra Tumaader aamaake shikhaate hobe naa 5you all don’t have to teach me4 Ektu chintaa korbaa aamaader theke onaara boishko 5you should think that he is much older than us4 Okay, look Aleha, how do you spell Zahra? Z-a-h-r-a. In school we call her Zahra, in school we call her Zahra. (classroom recording)

Here S1 corrects the teacher’s pronunciation. The different pronunciations of the name are significiant: in Bangladesh /z/ is pronounced as /j/, so the teacher is not pronouncing the name ‘Jara’ incorrectly, but is pronouncing it just as it would be in Bangladesh. The pronunciation of the name which the children use in school is an anglicized version, pronouncing the /z/. The students contest the ‘Bangladeshi’ (‘Sylheti’ and/or ‘Bengali’) pronunciation of the child’s name, and insist on the anglicized version. Here the students appear to use the teacher’s pronunciation of the Bangladeshi name as an opportunity to negotiate a subject position away from the imposed ‘heritage’ identity, and to use available linguistic resources in subtle, nuanced ways to occupy a position which is oppositional to ideologies which rely on the ‘purity’ of the Bengali language. The students’ complex and sophisticated response to the complementary schools’ ‘heritage’ positioning of them was also evident in interviews. In this excerpt two students were talking to the researchers about a drama activity, based on a story of new arrivals from Bangladesh. In talking to the researcher (R) they described this group as ‘freshies’ (cf. Martin et al. 2004): R: S1: R: S1: S2: S1:

R: S2: R:

What do you mean ‘freshie’, what does that mean? freshie as in a newcomer is that bad to say to somebody? yea it’s kind of like a blaze but it’s also a word to describe a new person coming from a different place. it’s not a good thing. it’s kind of both..if you say it as in trying to tease somebody, ‘freshie’, and we say it as in erm trying to say erm, as in they’re newcomers and they come from a different country for the first time could you tell if someone was ‘freshie’? well from Bangladesh it’s not always their skin colour, it’s sometimes how they talk. . . how do you talk ‘freshie’?

548


S1: S2: S1: S2:

it’s kind of like they don’t know that much English. they might just show off in their language but if you ask them a question in English they just. . . they’re like ‘what’, ‘what’, you know they say strange words in their language and if you ask them a question in English they just say ‘what’ in their language. (student interview, School B)

The students negotiate their identity in opposition to that of the newly arrived children, repeatedly referring to ‘their’ language, which they see as different from the language they speak themselves. Here ‘what’ ‘what’ is spoken with an intonation which suggests some confusion on the part of the newly-arrived group. Although the students speak the same language as the new arrivals in daily interactions with their parents, they nevertheless indicate that ‘how they talk’ is one of the defining ways in which the ‘freshies’ are different from them. The students’ linguistic repertoires were wide-ranging. In addition to making use of linguistic resources of English, Sylheti, and Bengali, they watched Hindi films, read the Qur’an in Arabic, and listened to popular contemporary music in varieties of American English, and also Indian and Bengali pop music. For some, listening to contemporary American music and watching DVDs was an important part of the way they viewed themselves. In the following example two siblings were asked what sort of Hindi songs they like: S1: R1: S1: R2: S1: R2: S1: R1: S2: R: S2: S1: R1: S1:

I like Bhangra really? I like Bhangra with rap oh they have all kinds of crossover Bhangra music now don’t they I like rap like Fifty Cent I mean do you like Eminem? yes he’s all right so is that OK? I mean rap and all that is all right? erm yea your dad doesn’t. . .? he doesn’t really erm if it’s in front of him he will shout but erm if we stopped it it’s all right. rap anyway I don’t hear rap at home I might just hear it a bit cos I hear it from my friend’s dad in his cars and everything because is your friend Pakistani or Indian? English (..) I mean Bengali (student interview)

Here ten-year-old S1 associates himself with firstly Bhangra, then ‘Bhangra with rap’, and finally ‘rap like Fifty Cent’. This appears to represent a negotiation of an increasingly risky, or perhaps sophisticated, subject position.


549

Whereas listening to Bhangra music may be regarded as relatively mainstream and conservative, ‘Bhangra with rap’ moves towards an increasingly American pop culture position, and ‘rap like Fifty Cent’ is likely to represent a ‘Gansta Rap’ identity. S1 is happy to be associated with ‘my friend’s dad’, and the researcher assumes that his friend must be of Pakistani or Indian heritage, as it may be surprising for a good Bangladeshi to listen to this kind of music. In the final utterance in this excerpt, S1’s pause between ‘English’ and ‘I mean Bengali’ may suggest that nationality, ethnicity, and even language are not the salient categories for him at this moment—rather, he is more interested in positioning himself as a cool, sophisticated consumer of contemporary, transnational music. In this section we have seen that students and teachers at times used the complementary school classroom as a space in which to negotiate identities. These negotiations were often focused on beliefs, attitudes and values relating to language(s), and played out in sophisticated deployment of linguistic resources (see also Harris 2006). We have also seen that for these students the ‘heritage’ identities which the schools set out to reproduce were often contested in subtle, sophisticated ways, as the students called into play sets of linguistic resources which positioned them as somewhat different from the imposed ‘heritage’ identities of the institution.

DISCUSSION What, then, can we say about negotiations which constitute, and are constituted by, the values, attitudes, beliefs, and practices of ‘language’ in and around these Bengali complementary schools? It is essential to any analysis that ‘the messiness of actual usage’ (Heller 2007: 13) should be understood in relation to histories, power, and social organization. In the course of our research we heard strongly articulated views, from parents and teachers, that Bengali should be taught as a mandatory part of the mainstream school curriculum. In the US context Wiley (2007: 254) refers to ‘the crisis of monolingualist ideology’, which proposes that English alone is of value in society. In British political, media, and other discourses a powerful ideology similarly proposes that minority languages other than English are a negative force in society. Languages which originate in the Indian sub-continent in particular come to be ‘racialized’ in the discourses of elite groups in UK, and associated with social segregation, family breakdown, and even terrorism (Blackledge 2005, 2006, forthcoming 2009). To some extent at least, the complementary schools are ‘safe spaces’ (Creese and Martin 2006: 2) in which young people are able to practise and extend their linguistic repertoires. In doing so they are ‘contesting the historical inequalities that have seen minority languages, and their speakers, relegated to the social and political margins’ (May 2007: 26). However, the process of teaching ‘heritage’ through ‘language’ is complex. First, for our participants the notion of what constitutes a ‘language’ is disputed. For some of the social actors concerned, one set of linguistic resources

550


(‘Bengali’) is heavily endowed with symbolic associations, and becomes a ‘social artefact, invented at the cost of a decisive indifference to differences’ (Bourdieu 1991: 287). This ‘Standard’ set of resources is regarded by some as that which should be ‘preserved’ and kept free from ‘contamination’. This particular set of resources accrues symbolic capital as it is perceived by some social agents as intrinsically superior to some other sets of linguistic resources (Bourdieu 1998: 47), and is ‘invented’ as a ‘language’ which is reified and immutable. The discourse of the school administrators in particular proposes that the ‘language’ should be transmitted to the next generation in its pure and natural form. However, this clear distinction between Standard and non-Standard sets of linguistic resources (‘Bengali’ and ‘Sylheti’) did not attract universal consensus. While the case for the purity of the ‘Standard’ was often argued in institutional discourse, other social actors, especially the parents of the students, and the students themselves, contested this view. Furthermore, we saw that some linguistic features came to be iconic representations of their speakers, as if a linguistic feature displayed the Sylheti group’s inherent nature. The fact of using certain sets of resources (‘Sylheti’), rather than others (‘Bengali’), appeared to index these speakers in particularly negative terms, despite the relative similarities of the two sets of linguistic resources. We saw that the parents of one child mocked the ‘Standard’ resources, and the ideological beliefs which were perceived as accompanying their speakers. In doing so they acknowledged the relations of power at work in the uneven distribution of resources, and the discourses which inscribe value (or its lack) to particular linguistic forms and practices. Second, we saw that the teaching of ‘heritage identities’ (Creese et al. 2006), through nationalist and historical content, was at times contested and subverted by students, in interactions which became sites for the negotiation of identities. We saw examples of language teachers teaching language through ‘heritage’ content with messages which were deeply rooted in Bangladeshi nationalism, invoking features of the collective memory of the nation. Also, we saw the repeated teaching of tangible and less tangible symbols of Bangladeshi heritage, from the national flag and national anthem to symbols such as the national flower, national fish, and national bird. Billig (1995: 174) argues that ‘nationalism’ is produced and reproduced in ‘daily, unmindful reminders of nationhood in the contemporary, established nation-state’, through everyday, ‘banal utterances’. These ‘self-evidences’ are those apparently commonsense misrecognitions which constantly construct and reinforce ideologies (Bourdieu 2000: 181). For the Bangladeshi-heritage group in our study, however, their nationalism was not produced and reproduced in everyday discourses in wider society. Perhaps in the face of the everyday ‘flagging’ of British/English nationalism, their approach to teaching their ‘heritage’ was explicit and often direct. The teachers appeared to impose on the students identities which were associated with Bangladesh and its history. Like the


551

institutional ‘language’ ideology, ‘heritage’ ideology was reified and naturalized. However, the students did not always accept the static, essentialized version of ‘heritage’ which the school was teaching. Howard (2003: 6) propose that ‘things actually inherited do not become heritage until they are recognised as such’. That is, while the teachers and administrators of the schools believed that teaching ‘language’ and ‘heritage’ was a means of reproducing ‘Bengali’ identity in the next generation, the imposition of such identities was often contested and re-negotiated by the students. Their apparent rejection of some ‘heritage’ symbols, their challenge of their teachers’ insistence on the use of ‘Bengali’ in the classroom, and their insistence on the anglicized pronunciation of a Bengali name, all became instances of students negotiating subject positions which contested those imposed by the institution. In the schools there was a perceived institutional need to fill what some teachers and administrators called the ‘cultural gap’ which had been created between the students and their parents. One of the teachers said of the students: to oraa 5they are4 British born, so they need to know Bangladesh, where their parents were born in Bangladesh, what is Bangladesh, where is Bangladesh . . . so many of them never express own self from own self [that] they are Bangladeshi, they always think, they always think they are British. Their mind perform, mind create that they are British. (teacher interview) This teacher was just one of several who argued that the students lacked something in their knowledge and understanding of Bangladesh, and in their sense of themselves as ‘Bangladeshis’. Bourdieu proposed that: The history objectified in instruments, monuments, works, techniques etc. can become activated and active history only if it is taken in hand by agents who, because of their previous investments, are inclined to be interested in it and endowed with the aptitudes needed to reactivate it. (Bourdieu 2000: 151 emphasis in original) It was this very endowment of aptitudes, this activation of history, which appeared to be the raison d’etre of the Bengali complementary schools. For the teachers, the process of teaching ‘language’, and teaching ‘heritage’, had the potential to invest their students with the aptitudes they required to inherit their heritage, because ‘only when the heritage has taken over the inheritor can the inheritor take over the heritage’ (Bourdieu 2000: 152).

CONCLUSION In this paper we have raised questions in relation to understandings of ‘language’ and ‘heritage’. The beliefs and practices of the participants raised a

552


number of questions in our attempts to understand how the language use of students and teachers in Bengali schools in one city in the UK was used to negotiate multilingual and multicultural identities. First, participants articulated attitudes and values which raised questions about what constitutes ‘language’. For some (a) ‘language’ should be preserved and kept free from the contamination of other sets of linguistic resources. For others there was no distinction in practice between resources ideologically framed as legitimate and illegitimate. Second, participants expressed views and attitudes, and performed interactional practices, which raised questions about what constitutes ‘heritage’. While teachers and administrators of the schools believed that teaching ‘language’ and ‘heritage’ was a means of reproducing ‘Bengali’/’Bangladeshi’ identity in the next generation, the imposition of such identities was often contested and re-negotiated by the students, as classroom interactions became sites where students occupied subject positions which were at odds with those imposed by the institutions. These young people were discursively negotiating paths for themselves which were in some ways contrary to the ideologies of the complementary schools, where teachers and administrators held the view that they ought to learn Bengali because to do so was a practice which carried with it knowledge of Bangladeshi history, nationalism, and identity. The young people’s attitudes to their languages, and their multilingual practices, constituted a sophisticated response to their place in the world, as they negotiated subject positions which took them on a path through language ideological worlds constructed by others. The young people were flexible and adaptable in response to their environment, as they negotiated identities which were more complex and sophisticated than the ‘heritage’ positions ascribed to them institutionally.

ACKNOWLEDGEMENT The paper draws on empirical data from one of four linked case studies in a larger, ESRC-funded project, ‘Investigating Multilingualism in Complementary Schools in Four Communities’, ESRC grant RES-000-23–1180.

NOTE 1 Throughout the article we adopt the following transcription conventions: plain font: bold font: italic font: bold font underlined: 5plain font enclosed4 ( )

Sylheti Bengali English other language (e.g. Arabic, Hindi) translation into English speech inaudible


553

REFERENCES Ahmad, Y. 2006. ‘The scope and definitions of heritage: From tangible to intangible,’ International Journal of Heritage Studies 12/3: 292–300. Bakhtin, M. M. 1973. Problems of Dostoevsky’s Poetics. Trans R. W. Rotsel. Ann Arbor, MI: Ardis. Bakhtin, M. M. 1984. Problems of Dostoevsky’s Poetics. Ed. and trans. C. Emerson. Manchester: Manchester University Press. Bakhtin, M. M. 1986. Speech Genres and Other Late Essays. Eds C. Emerson and M. Holquist. Austin, TX: University of Austin Press. Bakhtin, M. M. 1994. ‘Problems of Dostoevsky’s Poetics’ in Morris, P. (ed.): The Bakhtin Reader: Selected Writings of Bakhtin, Medvedev, Voloshinov. London: Arnold. Billig, M. 1995. Banal Nationalism. Sage: London. Blackledge, A. 2005. Discourse and Power in a Multilingual World. Amsterdam: John Benjamins. Blackledge, A. 2006. ‘The racialisation of language in British political discourse,’ Critical Discourse Studies 3/1: 61–79. Blackledge, A. forthcoming 2009. ‘Being English, speaking English: Extension to English language testing legislation and the future of multicultural Britain’ in G. Hogan-Brun, C. Mar-Molinero, and P. Stevenson (eds): Testing Regimes: Critical Perspectives on Language, Migration and Citizenship in Europe. Amsterdam: John Benjamins. Blackledge, A. and A. Pavlenko. 2001. ‘Negotiation of identities in multilingual contexts,’ International Journal of Bilingualism 5/3: 243–59. Borooah, V. K. 2005. ‘Caste, inequality, and poverty in India,’ Review of Development Economics 9/3: 399–414. Borooah, V. K., A. Dubey, and S. Iyer. 2007. ‘The effectiveness of jobs reservation: Caste, religion and economic status in India,’ Development and Change 38/3: 423–45. Bourdieu, P. 1991. Language and Symbolic Power. Cambridge: Polity Press. Bourdieu, P. 1993. Sociology in Question. London: Sage. Bourdieu, P. 1998. Practical Reason. London: Polity Press. Bourdieu, P. 2000. Pascalian Meditations. Cambridge: Polity Press. Bourdieu, P. and A. Darbel. 1991. The Love of Art. European Museums and their Public. London: Polity Press. Bourdieu, P. and C. Passeron. 1979. The Inheritors. Chicago: University of Chicago Press.

Chalmers, R. 1996. Learning Sylheti. London: Centre for Bangladeshi Studies. Creese, A. and P. Martin. 2006. ‘Interaction in complementary school contexts: Developing identities of choice—An introduction,’ Language and Education 20/1: 1–4. Creese, A., A. Bhatt, N. Bhojani, and P. Martin. 2006. ‘Multicultural, heritage and learner identities in complementary schools,’ Language and Education 20/1: 23–44. Creese, A., T. Baraç, A. Bhatt, A. Blackledge, S. Hamid, Li Wei, V. Lytra, P. Martin, C.-J. Wu, and D. Yag˘cıog˘lu-Ali. 2008. Investigating Multilingualism in Complementary Schools in Four Communities. Final Report. Birmingham: University of Birmingham. (RES-000–23–1180). Garcia, O. 2007. ‘Foreword’ in S. Makoni and A. Pennycook (eds): Disinventing and Reconstituting Languages. Clevedon: Multilingual Matters. Gard’ner, J. M. 2004. ‘Heritage protection and social inclusion: A case study from the Bangladeshi community of East London,’ International Journal of Heritage Studies 10/1: 75–92. Giddens, A. 1984. The Constitution of Society. Berkeley: University of California Press. Gumperz, J. 1982. Discourse Strategies. Cambridge: Cambridge University Press. Hamid, S. 2007. Unpublished PhD thesis, ‘A study of language maintenance and shift in the Sylheti community in Leeds. A cross-generational survey of language behaviour, attitudes and perceptions on Sylheti ethnocultural migrants from rural backgrounds’. University of York. Harris, R. 2006. New Ethnicities and Language Use. Basingstoke: Palgrave Macmillan. Harvey, D. 2007. ‘Heritage pasts and heritage presents. Temporality, meaning and the scope of heritage studies’ in L. Smith (ed.): Cultural Heritage. London: Routledge. Heller, M. 2007. ‘Bilingualism as ideology and practice’ in M. Heller (ed.): Bilingualism: A Social Approach. Basingstoke: Palgrave. Hornberger, N. 2005. ‘Heritage/community language education: US and Australian perspectives. The International Journal of Bilingual Education and Bilingualism 8/2&3: 101–8. Howard, P. 2003. Heritage. Management, Interpretation, Identity. London: Continuum. Irvine, J. and S. Gal. 2000. ‘Language ideology and linguistic differentiation’ in P. Kroskrity (ed.): Regimes of Language. Ideologies, Polities and

554


Identities. Santa Fe/Oxford: School of American Research Press. Kijima, Y. 2006. ‘Caste and tribe inequality: Evidence from India, 1983–1999,’ Economic Development and Cultural Change 54: 369–404. Lipe, W. 2007. ‘Value and meaning in cultural resources’ in L. Smith (ed.): Cultural. Heritage. London: Routledge. Makoni, S. and A. Pennycook. (eds) 2007. Disinventing and Reconstituting Languages. Clevedon: Multilingual Matters. Martin, P. W., A. Bhatt, N. Bhojani, and A. Creese. 2004. Final Report on Complementary Schools and their Communities in Leicester. Leicester/ Birmingham: University of Leicester/University of Birmingham. (ESRC R000223949). May, S. 2001. Language and Minority Rights: Ethnicity, Nationalism, and the Politics of Language. London: Longman. May, S. 2005. ‘Language rights: Moving the debate forward,’ Journal of Sociolinguistics 9/3: 319–47. May, S. 2007. ‘Language education, pluralism, and citizenship’ in S. May and N. Hornberger (eds): Language Policy and Political Issues in Education. Encyclopedia of Language and Education, Vol. 1, 2nd edn. New York: Springer. Nicholls, C. 2005. ‘Death by a thousand cuts: Indigenous language bilingual education programmes in the Northern Territory of Australia, 1972–1998,’ The International Journal of Bilingual Education and Bilingualism 8/2&3: 160–77. Patrick, D. 2007. ‘Language endangerment, language rights and indigeneity’ in M. Heller (ed.): Bilingualism: A Social Approach. Basingstoke: Palgrave. Pavlenko, A. (2007). ‘Autobiographic narratives as data in applied linguistics,’ Applied Linguistics 28: 163–88. Pavlenko, A. and A. Blackledge. 2004. ‘New theoretical approaches to the study of negotiation of identities in multilingual contexts’

in A. Pavlenko and A. Blackledge (eds): Negotiation of Identities in Multilingual Contexts. Clevedon: Multilingual Matters. Pearson, M. and S. Sullivan. 2007. ‘Looking after the past. An introduction to the management of heritage places’ in L. Smith (ed.): Cultural Heritage. London: Routledge. Peckham, R. S. 2003. ‘The politics of heritage and public culture’ in R. S. Peckham (ed.): Rethinking Heritage. Culture and Politics in Europe. London: I.P.Tauris. Pujolar, J. 2007. ‘Bilingualism and the nationstate in the post-national era’ in M. Heller (ed.): Bilingualism: A Social Approach. Basingstoke: Palgrave. Rampton, B. 2006. Language in Late Modernity Cambridge: Cambridge University Press. Smith, L. 2004. Archeological Theory and the Politics of Cultural Heritage. London: Routledge. Smith, L. 2006. Uses of Heritage. London: Routledge. Smith, L. 2007. ‘General Introduction’ in L. Smith (ed.): Cultural Heritage. London: Routledge. Stanton, C. 2005. ‘Serving up culture: Heritage and its discontents at an industrial history site,’ International Journal of Heritage Studies 11/5: 415–31. Tunbridge, J. E. and G. J. Ashworth. 1996. Dissonant Heritage. The Management of the Past as a Resource in Conflict. Chichester: Wiley. UNESCO. 2003. Convention for the Safeguarding of the Intangible Cultural Heritage. Paris: UNESCO. Wiley, T. 2005. ‘Discontinuities in heritage and community language education: Challenges for educational language policies,’ The International Journal of Bilingual Education and Bilingualism 8/2&3: 222–9. Wiley, T. 2007. ‘Heritage and community languages in the national language debate,’ The Modern Language Journal 91: 252–4.

Applied Linguistics 29/4: 555–577 ß Oxford University Press 2008 doi:10.1093/applin/amm051 Advance Access published on 10 January 2008

Routine Trouble: How Preschool Children Participate in Multilingual Instruction ¨ RK-WILLEŃ POLLY BJO Linko¨ping University, Sweden This paper examines the turn-by-turn organization of social actions during educational activities at a multilingual preschool in Sweden. Specifically, it focuses on instructional exchanges within two commonplace activities: ‘sharing time’ and ‘Spanish group’. The study builds on earlier research arguing that interactional routines facilitate children’s participation in social activities, and therefore promote language learning. Several instances of interactional trouble are identified and discussed in terms of the teachers’ elaboration of some routine features of these activities, resulting in a mismatch between the teachers’ local aims and the children’s projections of relevant next actions. The analysis further highlights a range of interactional means through which the participants act to come to terms with the trouble. These findings are discussed in terms of the participants’ local concerns as well as the children’s orientations to the routine features of preschool activities. Some educational implications are finally proposed on the basis of these findings.

INTRODUCTION Issues of bilingualism, multiculturalism, and immigrant children’s L2-acquisition have been debated in Sweden for years. The pendulum has swung from the policy of assimilation to the policy of integration of nonnative speakers, emphasizing today the need for ‘good Swedish’ for everyone. A recent report from the Swedish National Agency for Education (prop. 2005) argues that education based on children’s mother tongue promotes their cognitive development and enhances their opportunities for learning. Notably, matters dealing with how bilingualism is accomplished and organized in children’s everyday talk are largely overlooked in the report. However, implications of these arguments may be found in the educational challenges for many schools, as well as for preschools, which need to provide education for children with different language experiences. But how exactly do preschool teachers go about educating young children with diverse language backgrounds, and—perhaps even more importantly—how do preschool children go about participating in various curricular activities? The present paper originates from a larger study conducted in a multilingual preschool in Sweden. The preschool at hand offers daily curricular activities in Swedish, English, and Spanish. Furthermore, all three languages are spoken on a daily basis, in mundane talk. The multilingual approach of the

556

PRESCHOOL CHILDREN AND MULTILINGUAL INSTRUCTION

preschool involves an institutional emphasis on language awareness and language learning. Children’s participation in organized activities in a preschool context may be seen as priming activities for schooling. Preschool is part of the cultural processes that ‘can be thought of as practices and traditions of dynamically related cultural communities across generations’ (Rogoff 2003: 77). However, theories of cultural practice and learning tell us nothing about what goes on in actual preschool activities and the practical actions through which participants take part in, and reflexively construct, educational practice. Hence, I wish to argue that, if our goal is to understand how children achieve interactional competencies required for effective interaction, it is towards such mundane practices we need to direct our empirical efforts. That is to say, if we are to treat culture as a set of intersubjective practices, where the world and people are in constant reflexive involvement (Mehan 1979), the practices must be the focus of our studies of children’s learning. Further, children’s learning, in school as well as in preschool, may be seen as local and practical organizations where knowledge and competence are ‘installed’ (Macbeth 2000), and where educational facts are collaboratively and contingently constituted in interaction. Thus, if we are to provide insight into the nature of learning as embedded in the wider process of schooling, we need to closely attend to the social organization of interaction in educational contexts. Focusing on two local educational activities—‘sharing time’1 and ‘language group’—the present study is based on a view of learning as situated and constituted through social interaction. This perspective on learning originates from Vygotsky’s (1986) work on the relationship between language and the development of thinking and learning, but this view highlights the social, rather than the cognitive nature of learning: Learning is seen as something taking place between people rather than strictly inside the individual mind (Lave and Wenger 1991; Macbeth 2000; Mehan 1979; Rogoff 2003; Sa¨ljo¨ 2000).

INSTRUCTIONAL ROUTINES IN MULTILINGUAL EDUCATION Ethnographic studies of language socialization have often stressed the role of language routines in children’s language development (e.g. Ochs and Schieffelin 1995; Peters and Boggs 1986; Watson-Gegeo and Gegeo 1986). This is no less true for second language learning, and Ervin-Tripp (1986) stresses the role of play in children’s second language development. She points out that play activities show a number of recurrent patterns of action, and proposes that such recurrences may enable children to participate in peer group activities in spite of minimal language resources (cf. Cekaite 2006). Such participation in turn allows them to explore and develop a more nuanced communicative repertoire in their second language. Furthermore, play activities allow children to incorporate action patterns

¨ RK-WILLEŃ POLLY BJO

557

from the classroom into peer group interaction (see also Bjo¨rk-Willeń and Cromdal, in press), hence enabling children to explore some normative patterns of instructional talk between teachers and students. Along somewhat similar lines, Willett (1995) examines the interaction between three 1st-grade second language learners, showing how their talk draws upon routine features of teacher–student interaction. Clearly, drawing on institutionally established routines for accomplishing recurrent, and hence recognizable, activities is an important resource in second language learning. In a study of language socialization of a young Moroccan learner of Italian, Pallotti (1996, 2001) shows how the girl continuously acquired ways of participating in classroom talk by building on the talk of previous speakers within the structure of recognizable activities. By gradually adding new elements to recycled turns she was able to extend her lexical, pragmatic, and wider communicative repertoire. Somewhat similarly, Kanagy (1999) shows how young preschool children with no initial knowledge of the L2 rely on scripted instructional routines for their development of early interactional competence. By highlighting the role of verbal routines in recurring social activities, language socialization studies manage to anchor the process of second language learning within the cultural orders and practices in educational settings (Saville-Troike and Kleifgen 1986; Willett 1995). A distinct line of inquiry into second language learning and instruction is represented by a growing number of interaction analytic studies influenced, in varying degrees, by central ideas within ethnomethodology (e.g. Cromdal 2005; He 2004; Markee 2000; Slotte-Lu¨ttge 2005; Steensig 2001). While clearly distinct with respect to modes and detail of analysis, interaction analysts and researchers into language socialization both share a strong orientation towards the local practices through which second language learning and instruction takes place. For instance, analysing a conversation between a native speaker of English enrolled in an academic course in German and a native speaker of German, Kasper (2004) focuses on the ways through which the participants invoke different language identities by orienting to trouble with the production talk in German. Her analyses demonstrate how the interaction moves back and forth between mundane conversation and institutional (instructional) talk brought off by the non-native speaker’s repair initiations. Kasper’s (2004) study demonstrates how detailed analyses of L2 interaction allow us to treat participants’ identities—for instance, as native or non-native speaker—as products or their own actions, as invoked, displayed, and oriented to in and through the talk. Clearly, conversational repair presents a range of issues relevant to second language acquisition, over and above issues of speaker identity, and conversation analysts have explored a range of repair phenomena in second language educational settings and their relation to the learning affordances set up by different practices of repair in L2 instruction (Juvonen 1989; Kurhila 2001; Lehti-Eklund 2002; Seedhouse 2004; Slotte-Lu¨ttge 2005).

558


Focusing on the organization of different types of instructional activities in L2 education, Seedhouse (2004) highlights the reflexive relationship between the local organization of turns at talk and the curricular focus of the activities in which that talk is embedded. Along this line of reasoning, the specific educational focus of an instructional activity provides learners with cues on how to proficiently take part in the ongoing interaction. Hence, the recurrent and recognizable structure of instructional activities enables L2 learners to meaningfully participate in the institutional business of the L2 classroom despite often very limited language skills. However, this need not imply that L2 students always follow the instructional agenda, and in a series of recent studies of a Swedish immersion classroom, Cekaite and Aronsson (2004, 2005) show how early learners of Swedish draw upon a range of instructional routines across different classroom activities, resulting in humorous ‘time-outs’ from the currently ongoing instruction. Although resulting in a temporary suspension of the ongoing instructional activity, such actions allowed novice learners of Swedish to take part in, and contribute to, the ongoing classroom discourse. These discourse contributions comprised the very early steps in the students’ learning of Swedish—as well as their ability to take part in formal educational events (Cekaite and Aronsson 2004). The studies above highlight, in different ways, the routine aspects of interaction in bilingual and second language education. As such, they form the backdrop for the present analysis which focuses on the different interactional resources through which preschool children participate in two distinct types of multilingual instruction: sharing time and ‘Spanish group’. More specifically, the aim is to investigate the in situ accomplishment of routine instructional activities through which the preschool’s trilingual program is put into daily practice. In a nutshell, the analysis will flesh out some of the interactional means through which preschool teachers and their students talk trilingual education into being.

A MULTILINGUAL PRESCHOOL The present analysis draws on data from fieldwork in a multilingual preschool in a mid-sized Swedish town. The preschool at hand features a multilingual programme, including Swedish, English and Spanish. Swedish is the main language of interaction, but English and Spanish are routinely used for mundane as well as formal educational activities throughout the day.

Multilingual activities Each day at the preschool entailed several daily scheduled activities, such as sharing time, organized group activities (including language groups), outdoor play, meals, as well as a daily nap for the youngest children. In addition, there is also scheduled time for ‘free’ play (cf. Bjo¨rk-Willeń and Cromdal,


559

in press). Common activities, such as sharing time, may involve the use of several of the three languages available at the preschool, with Swedish serving as the common base language, or lingua franca, whereas other activities target one specific language, such as, for instance, the ‘language groups’ are more or less strictly conducted in Spanish or in English. The language group activities, the Spanish group (‘spanskagruppen’) and the English group (‘engelskagruppen’), take place four times a week, directly after sharing time. The curricular aim of the two language groups is to ensure the children’s familiarity with both English and Spanish in a playful manner. In these ‘language groups’, the Spanish teacher typically introduces Spanish vocabulary and simple phrases. Similarly, the bilingual Swedish/English teacher introduces some basics of English communication.

Children and staff There are a total of 36 children at the preschool, including (a) a group of toddlers (12 children) and (b) a group of children 3–5 years of age (24 children). The present study draws on recordings of the latter group. The study involves six children from Swedish-speaking homes, five children from Spanish-speaking homes, five from English-speaking homes and eight children from other language backgrounds (German, French, Arabic, Suryoyo, Latvian, Tamil, and Persian). In several cases, the children speak more than one language at home, for example Spanish and French. The staff members responsible for this group also represent different language backgrounds: Spanish as L1 (Irina), Arabic as L1 (Fatima), English as L2 (Susanne), and Swedish as L1 (Sofia).2 All staff members speak Swedish sufficiently for daily purposes at the preschool, but Susanne and Sofia are the only native speakers of Swedish.

Observations and recordings The children’s everyday activities at the preschool were observed and recorded during a period of 5 months. During data collection, the video camera served as the main tool for recording the interaction (a total of 37 hours). Detailed analyses of situated ‘classroom’ interaction allow us to explore the participants’ methods for accomplishing and participating in educational activities, that is, to investigate participants’ ways of exploiting and reflexively producing contexts for learning.

Taking attendance: teacher improvization as trouble source Sharing time is a daily morning activity, during which children and teachers typically gather in a circle on a big rug at the preschool. One goal with sharing time is to greet each other with ‘good morning’ (in different languages) and to take attendance. Another goal is to sing and talk in the

560


different languages, mostly Swedish, English, and Spanish. It is also a forum for gathering information, for conducting various thematically organised activities, and last but not least, it serves to distribute the morning fruit among the children. Sharing time is always led by one of the teachers, while the other teachers join in, seated among the children as group members, but also supporting the lead teacher. (Please see the online version for a more extensive introduction of the basic attendance taking routines at the sharing time, available online to subscribers at http://applij.Oxfordjournals.org/)

Language choice, repair, and the management of conflicting interactional projects Most of the children start attending the present preschool group at the age of 3 and finish when they are 6 years old. Hence, some of the children are more experienced than others in participating in routine activities such as sharing time. Children’s participation in these daily routines is assumed to result in changes in their knowledge of curricular matters, their social skills, and their discourse competence. During sharing time, the more experienced children are treated as old-timers (Lave and Wenger 1991). The language skills also differ between the children, not only due to age differences, but also to their differing language backgrounds. The latter holds true for the teachers as well. The first two transcripts illustrate the children’s participation in a specific activity that routinely took place during sharing time, namely taking attendance. When the transcript begins, the Swedish teacher Sofia is seated on a small chair within sight of all the children. There is a cotton tapestry hanging on the wall nearby, picturing a train with several trucks in different colours. As part of the tapestry, there is a box containing a number of rag dolls. Each doll represents one individual child. The typical procedure for taking attendance is that the teacher produces a doll from the box, and the particular child or/and the group of children identifies the doll. The child then receives the doll and is expected to put it in one of the trucks on the tapestry. Normally, children are free to pick a truck of their own choice. However, during this particular sharing time event (shown below in Excerpts 1 and 2), it is the teacher who designates the truck for each doll by naming the colour of the truck in English. The routine character of this activity provides for its recognisability for everyone at the preschool. It is not surprising, then, that the interaction between the teacher and the children most of the time runs very smoothly. However, this is not always the case, and the following sections will show in different ways how a local difficulty of interpretation may arise when teachers introduce new elements into routine multilingual activities. The following excerpt shows how the participants’ diverse language skills are interactionally displayed and acted upon within the flow of talk. When the transcript begins the teacher is just producing Gustavo’s doll.


Excerpt 1 (Video 3, 2002–11–12)

561

562


Instead of initially naming a colour, the teacher asks Gustavo (Gusta) what language he prefers: ‘English or Spanish?’ (lines 4–6). She thus turns the issue of language choice into a topic, thereby introducing a new element into the routine format of this activity. During a relatively long pause (line 7), Gustavo, who is standing in front of the teacher, does not respond. Nicolaus, one of the oldest boys in the group, appropriates Gustavo’s proposal that Spanish would be best (line 9). In offering this reply, Nicolaus not only recognises that an answer is expected, but he may be seen to orient to what he believes is Gustavo’s language preference. In so doing, Nicolaus acts upon and displays, partly on behalf of Gustavo, the type of knowledge that the present situation makes relevant. As Macbeth (2000) points out: ‘Classroom discourse is a grammar of activity in and through which students see and find the sense of their lessons, including things like the competence of their fellows . . .’ (p. 42). However, the teacher rejects Nicolaus’ (Nico) contribution by latching on to his utterance and rapidly declaring, in a mildly irritated voice, that she wants Gustavo to make the choice himself (lines 10–11). Nicolaus’ contribution is thus treated as a breach against the local organization of turn allocation. In other words, the teacher openly disqualifies Nicolaus’ contribution, because it is evidently inconsistent with her initiation move (Mehan 1979). However, in the absence of a response from Gustavo, she answers her own question, proffering the colour specification in both English and Spanish (lines 12–13), thus orienting to the absent—but expected— language choice on Gustavo’s behalf. This clearly terminates a failed IRE (Initiative–Response–Evaluation) sequence, as she, in fact, produces the expected response herself. One of the upshots of this move is to stress that an allocated initiating move has a normative feature. It is produced for a unique recipient, who cannot be replaced by other children. Hence, this brief exchange may be seen as a public lesson in the normative workings of the social and educational order of the preschool. However, the lesson at hand is not merely about the forms of participation in sharing time activities—it is also an instance of language instruction in practice. Hence, in answering her own question, the teacher offers both the English and the Spanish terms for the color ‘blue’. What is notable about the turn in line 15 is that teacher Sofia, one of the Swedish teachers at the preschool, pronounces the Spanish word incorrectly: ‘azoll’, and in the subsequent turn, a child immediately initiates a repair sequence by offering the correct pronunciation. In instructional exchanges, such other-initiated other-repair (Schegloff et al. 1977; Schegloff 2000), through which the recipient of talk locates a trouble source and offers a corrected version, are overwhelmingly carried out by the teacher (cf. McHoul 1990). In the second language classroom, repair phenomena may be more complex, as Seedhouse (2004) points out in his discussion of the indexical-reflexive relation between the pedagogical focus of the instructional activity and the organization of repair sequences.


563

In a sense then, we are dealing here with an instance of public language instruction in which the roles have been reversed. What makes this reversal possible is the specific domain of knowledge relevant to the situation at hand, namely knowledge of Spanish phonology. This allows some of the children in the group to display for the whole group a higher degree of competence than that held by the teacher. Upon the repair initiating turn in lines 16–17, the teacher latches on to the child’s correction, and produces a phonological repair of the word azul, followed by a repetition (lines 18–19), hence displaying a recognition of her own mispronunciation and the need for its remedy. Overlapping with the Swedish teacher’s self-repetition, the Spanish teacher Irina, produces the correct form of ‘azul’ in line 20. Given Irina’s authority as the Spanish teacher, this action can be seen as a validation of the repair sequence in lines 16–19, but it also seems to provide the public norm for the pronunciation of the colour term, as evidenced in line 21, in which a child silently repeats the word, thus orienting to Irina’s pronunciation as instructionally implicative. At the same time, Gustavo is slowly moving—doll in hand—from the teacher towards the tapestry with the train trucks where the dolls are to be placed. The Swedish teacher then repeats her earlier instruction (preserving the bilingual format, as well as the erroneous pronunciation; line 23) to Gustavo to place his doll in the blue truck. Gustavo rebukes this instruction, suggesting in Swedish that he prefers the grey truck: ‘nej jag vara gra˚ (.) den’ (‘no I be the grey (.) this one’; lines 25–27). Gustavo hereby shows an agenda that differs from the Swedish teacher’s, in that he intends to place his doll in the same grey truck as Irina, the Spanish teacher, has placed her doll earlier during this sharing time event. By claiming a right to chose a different truck than the one suggested by the Swedish teacher, Gustavo can be seen to act upon the standard organisation of the attendance taking routine, according to which it is up to each child to pick a truck for her/his doll. Gustavo’s choice of language at this particular point of interaction is interesting: by producing his reply in one of his weaker languages (as displayed by the incorrect verb inflection in line 25, as well as the teacher’s orientation to his language preference in line 23), Gustavo manages to decline the language choice options proffered by the teacher in the initial phase of this interaction. Thus, his disaffiliation with the teacher’s instruction is displayed both at the level of the content of talk as well as his choice of language (for a study of children’s use of code-switching in the construction of oppositional stances, see Cromdal 2004). The above analysis targets some difficulties arising from the organization of instruction during sharing time. It could of course be argued that some of the educational difficulties may be related to children’s varying experience of instructional events. Some children already know the routines, whereas others are still exploring the territory. However, we have seen that in spite of his ‘old-timer’ position, Nicolaus’ attempt to supply a noticeably absent reply to a preallocated teacher question (line 9) is treated as a violation of the

564


norms governing turn-taking in formal events at the preschool. The teacher’s handling of such norm infractions involves efforts to restore the institutional order—to bring the interaction back to a routine activity. This practice then becomes in itself a lesson for the children in the interactional workings of schooling in practice. However, in the current extract, the problems arise with the teacher’s deviation from the attendance taking routine. In particular, three nonstandardized elements were oriented to by the children: the Swedish teacher’s insistence on choosing the truck for the children, her delegating the issue of her own language choice to Gustavo, and her notably erroneous use of Spanish. The first deviation relates to Sofia’s attempt to integrate the routine activity of taking attendance with L2 instruction: her naming the colour of the target truck in English allows the students to practise and display their passive colour vocabulary. However, the teacher’s instructional project is clearly at odds with Gustavo’s local concerns, and we learn towards the end of the transcript that he insists on picking his own truck. The disparity of Gustavo’s and the teacher’s interactional projects has a bearing on the second deviation from the attendance taking routine, namely the teacher’s delegating her own language choice to Gustavo. Here, Gustavo’s noncompliance with the teacher’s request to display his language preference for this exchange (line 7) may be accounted for in terms of his reluctance to letting Sofia choose the truck for him. Finally, Gustavo’s refusal to go along with the teacher’s instructional project has profound implications for the subsequent interaction, as it leaves the issue of his language preference unresolved. Accordingly, the teacher chooses to specify the target truck in both English and Spanish, the latter resulting in a side-sequence in which her erroneous pronunciation in Spanish was collaboratively repaired.

Some collaborative techniques for accomplishing multilingual activities In the previous extract, it was shown how the teacher’s attempt to integrate the attendance taking activity with L2 instruction resulted in an observable deviation from the routine organization of the interaction during this activity. In this section, I will further explore some distinct, interactionally deployed techniques that the teachers employ to accomplish language instruction during sharing time. The excerpt below took place within the same sharing time event as the previous extract and further illustrates how colour terms are used as part of taking attendance.

Excerpt 2 (Video 3, 2002–11–12)


565

566


In rough outline, Sofia shows the ‘Henrik doll’ and Henrik picks it up. She then indicates that he should place the doll in the orange truck by saying orange three times (lines 4–5). However, rather than completing the expected course of action—placing the doll in the orange truck—Henrik keeps staring at the multicoloured trucks on the tapestry (line 9). It is impossible to tell whether he is hesitating because he does not recognize the English colour term or because he is just working out which of the two orange trucks to choose. However, the teacher seems to act upon the former interpretation—that he does not know the English colour term for orange—and repeats the colour term, first in Swedish, then in English (line 10). Latching on to the teacher’s last ‘orange’, Henrik ultimately puts his doll in one of the orange trucks (line 11). Note also that there is a small gap in Sofia’s utterance (between lines 4 and 10) where Irina, the Spanish teacher, tells Gustavo to pay attention. Her silent voice and choice of language suggests that this is a side activity, addressed only to Gustavo. Moreover, her use of Spanish contrasts with the current ‘official’ activity, and this choice of constitutive language may be a way of accomplishing a parallel activity, rather than disrupting Sofia’s role taking. On the contrary, it is clear that Irina is also cooperating with Sofia, as well as directing Gustavo’s attention to the main activity (line 8). Therefore, the gap Sofia leaves before she finishes her invitation to Henrik may be seen as a smooth interactional device, allowing Irina to finish her admonition and Sofia to regain full attention. Sofia’s way of repeating the colour term ‘orange’ in English several times is an instructional technique typical of language teaching (Chaudron 1988). In line 10, she also provides the Swedish equivalent and then translates it back into English. This turn shape, comprising repetition interspersed


567

with translation (repetition–translation–repetition), was recurrently observed at the preschool and may be seen as a teaching strategy with distinct cognitive overtones: in the present extract, it provides a way of making sure that Henrik and the other children have understood the word in a ‘correct’ manner (Lemke 1990).4 Notice that the teacher is using the local availability of several languages to deal with a problem of potential incomprehension on Henrik’s part, while Irina may be seen as using this availability to produce a non-distributive turn in the midst of the change. In lines 12–13, after Sofia has provided feedback to Henrik, it is Anna’s turn to place her doll. Anna is a 4-year-old girl who speaks both Swedish and Spanish at home; this is her second year in the group. The teacher says ‘brown’ in English using a soft and low voice, and simultaneously shows the ‘Anna doll’ (lines 12–13). In the next line, Anna takes her doll and, in accordance with the activity routine, puts the doll in a brown truck. After a positive evaluation of Anna’s action, Sofia proceeds to ask Anna to teach her the Spanish word for brown. Her request may be seen as an explicit educational strategy, where she highlights the situated import of the request as an educational routine: ‘kan du la¨ra mig det’ (‘can you teach me that’) (Mercer 1995). The utterance also construes Anna as more skilled in Spanish than the Swedish teacher (lines 16–18). Anna, however, does not respond; instead she looks down when passing Sofia on her way back to her place among the other children (line 19). In the absence of immediate compliance from Anna, Sofia begins to vocalize the Spanish word for brown. However, her vocalization of the first phoneme is incorrect: ‘na’ (line 18), and Sofia displays her own uncertainty by breaking off mid-word and turning to look at the Spanish teacher (lines 20–21). At the same time, Anna orients to the mispronunciation by looking up at Sofia with a swift headshake. Fixing her gaze at the Spanish teacher, Sofia produces a hesitant ‘na¨?’ (‘no?’, line 24), thus displaying her inability to produce the correct Spanish colour term. In this way, Sofia’s actions comprise a self-initiated other-repair (Schegloff et al., 1977), and in line 25, Irina begins to produce the correct term. However, rather than completing the correction, Irina uses a word-search format (Goodwin and Goodwin 1986) ‘ma::’, managing in this way to solicit the engagement of several children (Anna included), who produce the correct Spanish term for the colour brown (lines 26 through 27–29). Following up on this, Irina elaborates on the Spanish colour term, presenting a synonymous alternative, cafe´ (line 30). Receiving this as news, Sofia enthusiastically picks up on the latter term and presents the children with a mnemonic device, pointing out that coffee is in fact brown. Several children follow suit by repeating the term (lines 35 and 38). Hence, by highlighting the relation between the colour brown and its referent on behalf of the children group, Sofia regains the initiative as head teacher in the current activity.

568


This example shows that the teachers are using different educational techniques such as repetition, attention getting, guidance, elaboration, association, etc. (cf. Mercer 1995) to accomplish language instruction while taking attendance. It also shows how, in pursuing a multilingual instructional agenda, the Swedish teacher initially relies on the children’s cooperation in producing locally appropriate terms in Spanish. Failing that, she appeals to the expertise of the assisting Spanish teacher, who readily manages to accomplish what the Swedish teacher could not: to solicit Spanish talk from the children. This episode then clearly shows the children’s local sensitivity to matters of language choice and recipient design. Ultimately, the children seem to act upon an understanding of the teachers’ language preferences, such that they will respond to the Swedish teacher’s talk in Swedish or English, depending on local circumstances. Their use of Spanish, however, is largely restricted to exchanges with the Spanish teacher. Clearly, the children are facing a range of concerns related to their participation in a multiparty activity, which potentially includes more than one teacher. This involves making sense of the routine features of the unfolding activity and the local organization of different languages in multilingual instruction. Hence, the children’s successful participation in such events involves knowing the other participants’ language skills as well as designing their actions in accordance with this knowledge. Finally, I have argued that a potential trouble source for the children’s participation in this sharing time episode has to do with the institutional hybridity of the event, resulting from the Swedish teacher’s attempts to integrate two routine yet distinct activities: taking attendance and language instruction. Against this backdrop, let us now move on to consider one of the frequently occurring ‘language group’ activities, which explicitly focuses on aspects of language instruction and learning.

TEACHER FEEDBACK AS TROUBLE SOURCE IN THE SPANISH GROUP In moving from the daily sharing time activity of taking attendance to the activities of the Spanish group, I will highlight some features of teacher feedback in language instruction. In particular, activities in which the teacher questions the children for specific words in the target language entail children’s expectations for routine feedback.

Making sense of teacher feedback In the Spanish group, as well as during sharing time, there is a formal organization of talk, especially when the focus is on learning issues. This implies that the teachers present the topics and initiate the joint activities


569

and that almost any deviation from this pattern is handled as repairable (McHoul 1978). Part of the normative organization of the Spanish group is that the children are expected to participate in the activities using Spanish. To be successful and fully recognized by the teachers, the children must discover and adapt to this routine. However, the instructional procedure is not always easy to penetrate for the children, because teachers do not always stick to the routinized organization of the events. The final excerpt shows how children work at figuring out how to produce the answer that the teacher expects. (Another example of teacher feedback in the Spanish group is available online to subscribers at http://applij.Oxfordjournals.org/) Excerpt 3 shows the opening of a teacher-led play event, involving small beanbags. The bag distribution procedure that we will explore in the excerpt below involves the introduction of a physical activity. Here, the children are expected to use the bags for different purposes, such as throwing them up in the air and catching them, or walking around the room with a beanbag on their head, hand, foot, back, etc. These exercises are accompanied by music. The excerpt begins when the teacher shows the first beanbag.

Excerpt 3 (Video 3A, 2002–09–26)

570



571

This episode shows how the children respond to and coordinate their talk with their teacher during the Spanish activity. During this introduction part, Irma is very eager to get the first beanbag, jumping up and down saying: ‘den vill jag ha’ (‘I want that one’; line 2). The teacher gives Irma the bag (line 5) and then asks explicitly about its colour (line 6). Note that the teacher is using Swedish to present children with the procedural information for this activity. Thus, on the level of conversational organization, the question in line 6 results in a switch into Spanish. Irma aligns with the teacher’s language choice in line 15 and this exchange sets off the Spanish activity, the normative feature of which seems to be naming the colour of the beanbag in Spanish. In the next sequence the teacher shows a red bag, and Anna correctly identifies its colour in Spanish. What we may interpret as a confirmation of Anna’s answer in line 20, when the teacher repeats ‘rojo’, may also be interpreted as a collectively addressed turn made for all children. The repetition of rojo is followed by a translation to Swedish, where the teacher identifies rojo as the Spanish word for red. We see that the teacher orients to Swedish, which works as a lingua franca, when she provides explanations and sometimes even when she asks

572


questions. Nevertheless, the routine being established here requires that the children answer in Spanish. Hence, we see how the children continue to answer in Spanish, despite the teacher’s occasional switches to Swedish, thus orienting to the ‘Spanishness’ of the activity (lines 20 and 26). In other words, the children orient to perceived normative expectations regarding the choice of language in this instructional event. In spite of this, the relation between producing the correct answer and receiving a beanbag, which is the fundamental evaluative move in this activity, is not very clear for the children. For example, Anna produces the right answer in line 20, without receiving a bag, which is instead given to Evelina (line 22). However, as the turn moves to Michael in lines 30–33, the event takes a rather different direction, as he repeatedly fails to provide the answer the teacher expects. How can we account for his failure? An intuitively appealing answer might be that Michael’s failure to provide the Spanish word for blue has to do with his limited language proficiency. However, upon closer inspection of the unfolding organization of this event, the picture looks rather different, and we see that Michael is very consistent in his way of answering. At first, when Anna answers in Swedish ‘bla˚’ (‘blue,’ line 34), Michael follows suit, producing the Swedish colour term (line 35). Then, when Anna and Henrik, in lines 39–42 and 46–48, produce the ‘expected’ answer twice without the teacher producing the feedback to establish its correct status, Michael consistently does not follow suit. Instead, he tries to come up with an alternative, testing English in line 44, and finally settling for the French word for the colour blue in line 50, hence avoiding giving the answer in Spanish. I would thus argue that Michael’s failure to produce the expected answer is not related to his limited proficiency in Spanish. Rather, it should be seen as a result of the unfolding interaction, in which the teacher fails to furnish the expected feedback by acknowledging the ‘correct’ answer that has been repeatedly produced by several other children. Whatever the teacher’s reason for sidestepping the expected routine (not awarding Anna a beanbag in line 20 and not accepting the correct answer from Anna and Henrik in lines 39–42 and 46–48), the publicly visible result is an inconsistent orchestration of this educational activity, which, as has been argued, puts Michael in an interactional cul-de-sac. In light of this, it is interesting to see the teacher’s subsequent line of action, in which she simply provides the colour terms herself, distributing the bags to the remaining children (lines 58–62). The well-known routine is thereby suspended, and we can only speculate as to whether this is her way of getting around the impasse and getting on with the physical activity. It clearly brings about a radical change in the children’s way of participating, in that they no longer have to tackle issues of language choice and other concerns about turn organization. They simply have to catch their beanbags. This excerpt thus shows how children rely for their successful participation in structured instructional events on the recognizable, routine features of Spanish group activities. Having grasped the procedural routines of


573

language instruction, children design their actions in line with the expected—or projected—trajectory of interaction, and deviations from the expected course of action may comprise a source of interactional trouble.

CONCLUDING DISCUSSION The overall aim of this paper has been to highlight the in situ accomplishment of educational activities in a multilingual preschool. Two different types of activities have been explored: sharing time, led by a Swedish preschool teacher, and Spanish group, conducted by the Spanish teacher. As the analyses show, a fundamental concern for the children participating in these activities has to do with the routine organization of interaction, which provides for the predictability of relevant next actions. The significance of children’s expectations becomes clearly observable in the trouble arising with the teachers’ improvising, elaborating, or otherwise deviating from the established, routine performance of these activities. This general point may seem trivial, as several studies have argued for the educational upshots of activity routines in second language learning (e.g. Ervin-Tripp 1986; Kanagy 1999; Pallotti 2001; Willett 1995), highlighting the predictability of expected action trajectories as a resource for the children’s participation in educational practice. However, the interactional workings of activity routines—their local relevance for the children’s moment-to-moment accomplishment of participation—has not been widely explored in studies of multilingual or second language education (but see Cekaite 2007; Seedhouse 2004). The analyses, then, contribute to the understanding of multilingual instruction as a contingently sensitive practice, and it does so by examining, in some degree of interactional detail, the trouble arising with instructional routines being somehow modified or suspended, as well as the management and resolution of such trouble. One type of children’s procedural expectations highlighted in the data has to do with the teacher’s language preferences, as oriented to and acted upon by the children. Here, the analyses have shown how the Swedish/English teacher’s attempts at doing Spanish instruction resulted in elaborate repair work, initiated by the children (Excerpt 1). As a possible result of this experience, we have seen how the teacher later on consults the expert Spanish knowledge of one of the children (Excerpt 2). This reversal of the institutional roles and relationships between teacher–child constitutes another alien element in the registry taking activity, resulting in some procedural trouble, which was ultimately resolved through the authority and expertise of the Spanish teacher. Another normative feature of instructional exchanges at the preschool has to do with teacher feedback. More specifically, children overwhelmingly expected their correct responses to be assessed in positive terms by the

574


teacher and I proceeded to show in Excerpt 3 how the lack of teacher feedback following upon several correct answers from the group raised some acute problems for Michael in predicting the relevant response to the teacher’s prompts. Let us consider some educational implications of these findings. The above analyses of the interactional trouble arising with the teacher’s deviations from the expected course of events suggests that children monitor the unfolding activity for its routine features, and further, that they use the predictability of the routine to participate in the interaction. This is not to suggest that teachers should constantly carry out instruction according to a fixed format. For instance, the first two excerpts showed a somewhat modified procedure for registry taking, according to which the teacher would temporarily take over the task of nominating the truck in which to place the student dolls. This modification as such did not counter—or fail— the children’s expectations to the extent that they were not able to participate in the activity. This would suggest then that children’s interactional competence is flexible enough to allow for novel elements to be introduced into routine activities. However, such variations may need to be clearly introduced, so as to enable children to predict the expected trajectory of the interaction. The analyses therefore allow us to conclude that whereas young children need the routines to participate in preschool business, they show enough interactional skill to follow the teachers’ modifications of recurring activities, provided that the procedural upshots of these elaborations are made clear. Finally and in conclusion, the focus has been on children’s trouble in ‘reading’ some educational routines, particularly in those cases where the teachers would depart from the expected course of action. This might seem ironic, as our data show that some of the teachers’ modifications and elaborations of the standard activities are explicitly designed to enhance children’s engagement and active participation in the educational practice at the preschool. However, I want to point out that the repair exchanges through which routine-related procedural problems are handled provide opportunities for the children’s early learning of foreign languages. As Goodwin (2000) puts it: if a child grows up in an ideal world where she heard only wellformed sentences she would not learn to produce sentences herself because she would lack the analysis of their structure provided by events such as repair. (Goodwin 2000: 5) Clearly, such learning is inseparably embedded in the children’s development of social and interactional skills. Thus, recognizing and remedying interactional trouble may be a crucial way of learning to become a competent participant in social life, including the context of organized multilingual activities in preschool. At present, we are far from understanding in any


575

sufficient detail the educational activities taking place in multilingual preschools and the upshots of these local practices for children’s early learning. Hence, the present work contributes to our understanding of very young bilinguals’ ways of managing commonplace social activities in educational settings—a hitherto largely overlooked line of inquiry in Swedish research on bilingualism.

ACKNOWLEDGEMENTS I am very grateful to Karin Aronsson, Jakob Cromdal, Michael Tholander, Asta Cekaite and Mathias Broth for their support and many helpful comments. Financial support from the Committee for Educational Research of the Swedish Research Council is gratefully acknowledged.

TRANSCRIPTION KEY [ Underlining CAPITALS

(3) (.) (( text )) : 4word5 5word4 ¼ ! #" (. . .) (text) (x) (xx) he he text text text text

Square brackets mark the start of overlapping speech Signals emphasis; the extent of underlining within individual words locates emphasis Marks speech that is obviously louder than surrounding speech Quieter speech Pauses measured in seconds Micropause Transcriber’s comments Prolongation of preceding vowel Quicker speech Slower speech Immediate ‘latching’ of successive talk Utterance interrupted or ebbed away Highlights a particular feature discussed in the text Indicates falling or rising intonation in succeeding syllables(s) Talk has been omitted Uncertain interpretation Inaudible word or words Laughter Swedish Spanish English Translation into English

576


NOTES 1 ‘Sharing Time,’ which may also be known as ‘Circle Time’ or ‘Show and Tell,’ is an umbrella activity that is widely practised throughout the Western world in both L1 and L2 primary classrooms and in preschools. Normally the children sit in a circle and the focus is on creating a sense of community and developing social and interactional skills (Yazigi and Seedhouse 2005). 2 All personal names are pseudonyms.

3 ‘Azoll’ is the teacher’s pronunciation of the Spanish word ‘azul’ (blue). Because Swedish is a non-phonetic language, I have used standard written language throughout the transcripts, except when pronunciation has an impact on the interaction, such as in the case of ‘azoll.’ 4 Lemke (1990) shows how thematic patterns are communicated using the strategy of repetition-with-variation to explore semantic relations for a wider understanding.

REFERENCES Bjo¨rk-Willeń, P. and J. Cromdal. In press. ‘When education seeps into ‘‘free play’’. How preschool children accomplish multilingual education,’ Journal of Pragmatics. Cekaite, A. 2006. Getting started. Children’s participation and language learning in an L2 classroom. Linko¨ping Studies in Arts and Science, 350. Linko¨ping: Linko¨ping University. (Diss.) Cekaite, A. 2007. ‘A child’s development of interactional competence in a Swedish L2 Classroom,’ The Modern Language Journal 91/1: 45–62. Cekaite, A. and K. Aronsson. 2004. ‘Repetition and joking in children’s second language conversations: Playful recyclings in an immersion classroom,’ Discourse Studies 6/3: 373–92. Cekaite, A. and K. Aronsson. 2005. ‘Language play, a collaborative resource in children’s L2 learning,’ Applied Linguistics 26/2: 169–91. Chaudron, C. 1988. Second Language Classrooms: Research on Teaching and Learning. Cambridge: Cambridge University Press. Cromdal, J. 2004. ‘Building bilingual oppositions: Code-switching in children’s disputes,’ Language in Society 33: 33–58. Cromdal, J. 2005. ‘Bilingual order in collaborative word processing: On creating an English text in Swedish,’ Journal of Pragmatics 37: 329–53. Ervin-Tripp, S. 1986. ‘Activity structure as scaffolding for children’s second language learning’ in J. Cook-Gumperz, W. Corsaro, and J. Streeck (eds): Children’s World and Children’s Language. Amsterdam: Mouton de Gruyter.

Goodwin, C. 2000. ‘Practices of seeing: Visual Analysis: An Ethnomethodological Approach’ in T. van Leeuwen, and C. Jewitt (eds): Handbook for Visual Analysis. London: Sage Publications. Goodwin, M. and C. Goodwin. 1986. ‘Gesture and coparticipation in the activity of searching for a word,’ Semiotica 62/1–2: 51–75. He, A. W. 2004. ‘CA for SLA: ‘‘Arguments from the Chinese Language Classroom’’,’ Modern Language Journal 88/4: 568–82. Juvonen, P. 1989. ‘Repair in second-language instruction,’ Nordic Journal of Linguistics 12: 183–204. Kanagy, R. 1999. ‘Interactional routines as a mechanism for L2 acquisition and socialization in an immersion context,’ Journal of Pragmatics 31: 1467–92. Kasper, G. 2004. ‘Participant orientations in German conversations-for-learning,’ Modern Language Journal 88/4: 551–67. Kurhila, S. 2001. ‘Clients or languages learners— being a second language speaker in institutional interaction’ in R. Gardner and J. Wagner (eds): Second Language Conversations. London: Continuum. Lave, J. and E. Wenger. 1991. Situated Learning. Legitimate Peripheral Participation. Cambridge: Cambridge University Press. Lehti-Eklund, H. 2002. ‘Dictoglossamtal som o¨vningsform i svenska spra˚kbadsklasser’, in M. Saari (ed.): Samtal och Interaktion (Conversation and Interaction). Meddelande fra˚n institutionen fo¨r nordiska spra˚k och nordisk litteratur vid Helsingfors universitet Vol. B22. Helsinki: Helsinki University.


Lemke, L. J. 1990. Talking Science: Language, Learning and Values. New Jersey: Ablex Corporation Norwood. Macbeth, D. 2000. ‘Classrooms as installations. Direct instructions in early grades’ in S. Hester and D. Francis (eds): Local Educational Order Ethnomethodological Studies of Knowledge in Action. Amsterdam/Philadelphia: Benjamins. Markee, N. P. P. 2000. Conversation Analysis. Mahwah, NJ: Lawrence Erlbaum. McHoul, A. 1978. ‘The organisation of turns at formal talk in the classroom,’ Language in Society 7: 183–213. McHoul, A. 1990. ‘The organisation of repair in classroom talk.’ Language in Society 19: 349–77. Mehan, H. 1979. Learning Lessons, Social Organization in the Classroom. Cambridge, MA: Harvard University Press. Mercer, N. 1995. The Guided Construction of Knowledge. Talk amongst Teachers and learners. Clevedon-Philadelphia-Adelaide: Multilingual Matters. Ochs, E. and B. Schieffelin. 1995. ‘The impact of language socialization on grammatical development’ in P. Fletcher and B. MacWhinney (eds): The Handbook of Child Language. London: Blackwell. Pallotti, G. 1996. ‘Towards an ecology of second language acquisition: SLA as a socialization process’ in E. Kellerman, B. Weltens, and T. Bongaerts (eds.): Proceedings of EUROSLA 6, Toegepaste Taalwetenshap in Artikelen, 55. Pallotti, G. 2001. ‘External appropriations as a strategy for participating in intercultural multiparty conversations’ in A. Di Luzio, S. Gu¨nthner, and F. Orletti (eds): Culture in Communication. Amsterdam: Benjamins. Peters, A. and S. Boggs. 1986. ‘Interactional routines as cultural influences upon language acquisition’ in Schieffelin and Ochs (eds): Language Socialization across Cultures. Cambridge: Cambridge University Press. Proposition 2005/06:2 (2005). Ba¨sta spra˚ket—en samlad spra˚kpolitik (The best language—a collected language politics). Stockholm: Utbildnings- och kulturdepartementet (Swedish National Agency for Education).

577

Rogoff, B. 2003. The Cultural Nature of Human Development. New York: Oxford University Press. Saville-Troike, M. and J. Kleifgen. 1986. ‘Scripts for school: Cross-cultural communication in elementary classrooms,’ Text 6: 207–21. Schegloff, E. A. 2000. ‘When ‘‘Others’’ Initiate Repair,’ Applied Linguistics 21/2: 205–43. Schegloff, E. A., G. Jefferson, and H. Sacks. 1977. ‘The preference for self-correction in the organisation of repair in conversation,’ Language 53: 361–28. Schieffelin B. and E. Ochs (eds): 1986. Language Socialization across Cultures. Cambridge: Cambridge University Press, pp. 80–96. Seedhouse, P. 2004. The Interactional Architecture of the Language Classroom: A Conversation Analysis Perspective. Oxford: Blackwell. Slotte-Lu¨ttge, A. 2005. Jag vet int va de heter pa˚ svenska‘‘: Interaktion mellan tva˚spra˚kiga elever och deras la¨rare i en enspra˚kig ˚ bo Akademi klassrumsdiskurs. A˚bo, Finland: A University Press. (Diss.) Steensig, J. 2001. ‘Some notes on the use of conversation analysis in the study of bilingual interaction’ in J. N. Jørgensen (ed.): Multilingual Behaviour in Youth Groups: Scandinavian studies in the use of two or more languages in group conversations among children and adolescents. Copenhagen: The Danish University of Education. Sa¨ljo¨, R. 2000. La¨rande i praktiken. Ett sociokulturellt perspektiv (Learning in practice. A sociocultural perspective). Stockholm: Prisma. Watson-Gegeo, K. A. and D. W. Gegeo. 1986. ‘Calling-out and repeating routines in Kwaraáe children’s language socialization’ in Schieffelin and Ochs (eds): Language Socialization across Cultures. Cambridge: Cambridge University Press. Willett, J. 1995. ‘Becoming first graders in an L2: An ethnographic study of L2 socialisation,’ TESOL Quarterly 29/3: 473–503. Vygotsky, L. S. 1986. Thought and Language (trans. A. Kozulin). Cambridge, MA: Harvard University Press. Yazigi, R. and P. Seedhouse. 2005. ‘ ‘‘Sharing time’’ with young learners,’ TESL-EJ 9/3: 1–26.

Applied Linguistics 29/4: 578–596 ß Oxford University Press 2008 doi:10.1093/applin/amm056 Advance Access published on 31 January 2008

Symmetries and Asymmetries of Age Effects in Naturalistic and Instructed L2 Learning CARMEN MUNÕZ University of Barcelona The effects of age on second language acquisition constitute one of the most frequently researched and debated topics in the field of Second Language Acquisition. Two different orientations may be distinguished in age-related research: one which aims to elucidate the existence and characteristics of maturational constraints on the human capacity for learning second languages, and another which purports to identify age-related differences in foreign language learning, often with the aim of informing educational policy decisions. Because of the dominant role of theoretically-oriented ultimate attainment studies, it may be argued that research findings from naturalistic learning contexts have somehow been hastily generalized to formal learning contexts. This paper presents an analysis of the symmetries and asymmetries that exist between a naturalistic learning setting and a foreign language learning setting with respect to those variables that are crucial in the discussion of age effects in second language acquisition. On the basis of the differences observed, it is argued that the amount and quality of the input have a significant bearing on the effects that age of initial learning has on second language learning. It is also claimed that age-related studies in foreign language learning settings have yielded significant findings that contribute to the development of an integrated explanation of age effects on second language acquisition.

The distinction between naturalistic language learning and foreign language learning is often ignored in discussions of maturational constraints in second language acquisition. However, it is the contention of this paper that the differences in the amount and quality of the respective input of the two learning settings may have a significant influence on the effects that the age of initial learning has on the outcome of second language learning.1 In simple terms, naturalistic second language learning may be characterized as learning through immersion in the second language environment, whereas foreign language learning may be characterized as formal learning in the classroom. To expand, a typical foreign language teaching/learning situation in the schools of many countries may be characterized by means of some or all of the following features: (i) instruction is limited to 2–4 sessions of approximately 50 minutes per week;2 (ii) exposure to the target language during these class periods may be limited in source (mainly the teacher), quantity (not all teachers use the target language as the language of

CARMEN MUNÕZ

579

communication in the classroom) and quality (there is a large variability in teachers’ oral fluency and general proficiency);3 (iii) the target language is not the language of communication between peers;4 and (iv) the target language is not spoken outside the classroom.5 While a binary distinction serves the aim of this paper of pinpointing symmetries and asymmetries between second language learning in the two different settings, it goes without saying that the situation varies from country to country. Likewise, although the paper will be focused on the two opposite ends of the distinction, it is expected that the discussion will be relevant to second language learning that takes place in school immersion situations, in which input and use of the target language may be said to be limited but to a much lesser extent. In order to avoid confusion, in this paper the term second language learning will be used in its general sense to refer to both learning that occurs in a naturalistic setting and learning that takes place in a foreign, instructed or formal setting. When referring to the former situation, the term naturalistic learning will be used, in opposition to the terms foreign, formal, and instructed, which will be used in reference to the latter situation. However, the terms learning and acquisition will be used as synonyms. Having drawn this first general distinction, we move now to a distinction that is specific to the field of age effects on second language acquisition—that between rate of learning and ultimate attainment—and which is drawn from an extended review of research findings made almost three decades ago (Krashen et al. 1979). These early research findings, which have been corroborated by subsequent research, revealed that (i) older children, adolescents, and adults generally make more rapid progress in the first stages of the second language acquisition process than younger children, particularly in morphosyntactic aspects, and that (ii) the younger a L2 learner is when the L2 acquisition process begins, the more successful that process will be, that is, the more possibilities s/he will have of attaining native-like proficiency. Therefore, the distinction identifies two different agerelated advantages: an initial rate advantage on the part of older learners over younger learners, and an ultimate attainment advantage of younger starters over older starters. Evidence from the latter comes from studies of naturalistic second language acquisition, the so-called immigrant studies, in which both phenomena are widely observed. Studies of formal learning in foreign language settings to date have only provided evidence of the former advantage, that is, the older learners’ initial rate advantage, in situations in which older starters proceed faster than younger starters with the same amount of instruction time (Garcıá Mayo and Garcıá Lecumberri 2003; Munõz 2006b). The initial rate advantage is also observed in situations in which older learners begin instruction in the foreign language later than younger learners. In these cases, as Krashen et al. (1979: 579) note, the common finding is that children who start learning the foreign language later eventually catch up to those who begin earlier.

580

SYMMETRIES AND ASYMMETRIES OF AGE EFFECTS

It has been claimed that, from a theoretical point of view, the crucial notion in the distinction is ultimate attainment. The superior ultimate attainment of younger starters is seen as evidence of the Critical Period Hypothesis (CPH) (Lenneberg 1967) and, in general, for maturationally-based explanations of the biological constraints on second language acquisition. What is important for a theory of second language acquisition—according to Long (2005: 291)—is not short-term differences in performance, but rather long-term differences in capacity for acquisition, that is, ultimate attainment. Although the term ultimate attainment6 has often been used as a synonym for nativelike proficiency, Birdsong (2004, 2006) points out that the term properly refers to the final product of L2 acquisition, whether this is nativelike attainment or any other outcome, that is, irrespective of the degree of approximation to the native grammar.7 Although it is difficult to dissociate the two terms because of their strong association in the literature in this area, it is certainly desirable both methodologically and conceptually. In fact, more and more voices in the field of SLA now argue against the traditional view of nativelikeness and its characterization, claiming that a bilingual or multilingual competence cannot properly be compared with a monolingual competence (see Grosjean 1989, 1998; Cook 2002; Birdsong 2005, for a recent discussion of the nativelikeness standard in CPH studies). It could certainly be claimed that people who have acquired two languages simultaneously from birth constitute a better yardstick for comparison than monolingual native speakers. In the case of foreign language acquisition, where the proficiency levels attained are known to fall very short of nativelike proficiency, the association of ultimate attainment and nativelikeness is clearly inappropriate. More importantly, not even the concept of final product can be transported or adopted. The first reason for this is that the definition of the final product in ultimate attainment studies may be said to entail the cessation of learning (this is why we need a long period of residence to ascertain that L2 users are not still learning, as discussed below). As in studies of fossilization, or permanent stabilization (see Long 2003), a necessary methodological requirement seems to be that cessation of learning appears in spite of optimal learning conditions (see, among others, Selinker and Lamendella 1979; Han 2004; Han and Odlin 2005; Nakuma 2005). The requirement of the presence of optimal learning conditions (including input that is neither quantitatively nor qualitatively limited) is not met in foreign language settings. Failing this requirement, in the case of classroom learning the attribution of low achievement solely to starting age does not seem to be justified. In the same vein, the case for the crucial influence of intensity of exposure in second language learning has been recently reinforced by neurolinguistic research that has obtained evidence of the crucial role played by intensity of exposure in the neural representation of multiple languages (Perani et al. 2003).

CARMEN MUNÕZ

581

Returning to the generalization concerning the higher ultimate attainment of younger learners in a naturalistic language setting, we can say that this implies that the initial advantage of older learners must be progressively reduced to allow the younger learners’ eventual advantage to emerge. In a naturalistic immersion setting, research by Snow and Hoefnagel-Ho¨hle (e.g. 1978) provided evidence of the beginning of the process by which differences in test results between older and younger learners were being reduced in a period of approximately 1 year. When comparing children, adolescents, and adults who had been in the L2 community for 6 months, significant age differences were noticed in favour of the adolescent and adult learners. However, by the second and third tests, after 4- to 5-month intervals, the younger learners had begun to catch up with and, in some tests, overtake the older learners. These findings reinforced the generalization concerning a higher eventual attainment of younger starters in a naturalistic learning setting but, as observed by Harley (1998) twenty years later, no evidence has yet been found that the generalization of the higher eventual attainment of younger starters in naturalistic settings can be automatically extended to foreign language settings. This remains the case today. In other words, no evidence has yet been found that the proficiency of younger formal learners eventually becomes higher than that of older formal learners after the same amount of instruction. It has been suggested that the reason why younger beginners have not been observed to overtake older beginners is that, in school conditions, there is less exposure to the target language and the experience is less intensive (Ellis 1994; Singleton 1995). That is to say, because of the differences in input in both learning contexts, the period of time needed for younger formal learners must necessarily be much longer than that needed in a naturalistic learning setting. It then follows that the older starters’ initial advantage as well as the younger starters’ eventual advantage must imply longer periods than in a naturalistic situation. Following this line of reasoning, Singleton and Ryan (2004) argue that the study by Florander and Jensen (1969) offers a small amount of evidence that the same phenomenon of progressive reduction of older L2 learners’ lead over younger learners also occurs in formal instructional settings. In that study the older learners’ lead diminishes after 320 hours, although it still remains significant. This may indeed constitute evidence of the progressive reduction of the older learners’ advantage, which has been observed recently with respect to some language dimensions and after a longer period (more than 700 hours of instruction; see Munõz 2006b), but it does not yet in itself constitute proof that younger learners will necessarily outperform older learners in the longer term. Indeed, if the older learners’ advantage is due mainly to their superior cognitive development, as is claimed in Munõz (2006c), no differences in proficiency are to be expected when differences in cognitive development also disappear with age. To summarize, there is still a lack of empirical evidence to date confirming that, after the initial stages

582


of foreign language learning, younger starters overtake older starters in school settings. On the other hand, even in studies where instructional time is not constant and younger starters have had greater exposure than older starters, the former do not succeed in showing a significant advantage. In this respect, Harley (1998: 27) notes that no explanation has yet been provided for why in school settings ‘the additional time associated with an early headstart has not been found to provide more substantial long-term proficiency benefits’. In contrast, studies comparing early and late immersion students in Canadian immersion programmes have observed long-term benefits, particularly in oral communication. While it is true that late immersion students have been observed to catch up in aspects of proficiency such as reading comprehension and writing skills (Lapkin et al. 1980; Swain 1981; Cummins 1983; Harley 1986; Turnbull et al. 1998) and that early learners have sometimes failed to demonstrate the dramatic long-term advantage that was expected of them,8 it can be argued that the massive amount of exposure provided by school immersion has had a significant influence on the results. That is, an early starting age produces long-term benefits when associated with greater time and massive exposure, as in immersion programmes, but not when associated with limited time and exposure, as in typical foreign language learning classrooms.

THE REQUIRED LENGTH OF EXPOSURE In ultimate attainment studies, the length of time of residence (LoR) is the operationalization of the learners’ amount of exposure between age of acquisition (AoA) or initial age of learning and testing time. In foreign language learning studies, amount of exposure is operationalized either as the number of hours of instruction or as courses of instruction, although equating hours of exposure with hours of instruction is a gross generalization. Researchers have speculated about the length of time of residence or immersion in the L2-speaking environment that is required for a learner to have reached ultimate attainment in naturalistic learning settings. Importantly, the effects of the LoR variable appear to be limited to an initial period, after which they diminish or disappear. Krashen et al. (1979) suggest that length of residence plays little or no role after the first 5–10 years. This is in line with DeKeyser’s (2000: 503) argumentation that a minimum of 10 years should be required in ultimate attainment studies in order to ensure that it is ultimate attainment rather than rate effects that are being picked up. Likewise, using self-estimations of L2 proficiency from an extremely large sample of census data in the United States, Hakuta et al. (2003) found no evidence of the effect of LoR on English in individuals who had 10 or more years of US residence. Stevens (2006) claims, however, that large-scale empirical analyses have uniformly shown that the positive relationship between LoR and measures of English language proficiency among adult

CARMEN MUNÕZ

583

immigrants in the United States extends well beyond 5 or 10 years of residence. This period of 10 (or even more) years would be a proxy—using Birdsong’s (2004) words—for ultimate attainment, or L2 end state, given the difficulty of determining when the end state has been reached. Because the issue of reaching the end state, as discussed in the previous section, loses its relevance in second language learning in a foreign language setting, deprived by definition of unlimited input, no conceptual symmetry can be traced here. In contrast, the length of instructional time is an important matter of contention in relation to the alleged long-term benefits of an early start, namely the higher eventual attainment of younger starters. The issue then is to identify the long-term in foreign language learning, that is, the amount of instructional time that may be considered sufficient for the study of long-term benefits. Unsurprisingly, studies that have aimed to observe the long-term effects of starting age in a foreign language classroom setting have looked for them at the end or close to the end of secondary education, although the long-term benefits sought have not been found (e.g. Oller and Nagato 1974; Burstall 1975). More recently, the BAF (Barcelona Age Factor) project has compared groups of learners after approximately 200, 400, 700, and 800 hours (Munõz 2006b; Nave´s 2006). Other recent studies of foreign language learning have used similar periods (from 600 to 800 hours) in their longest-term comparisons (Garcıá Mayo and Garcıá Lecumberri 2003). Identifying the long term at the end of secondary education, the natural limit of most students’ foreign language learning process, has external validity because it represents the longest possible term for the majority of the population and enables the determination of relevant educational implications. In addition, it marks the point beyond which keeping a tight control of variables becomes more and more difficult. In this respect, studies whose informants are college students who began foreign language education in elementary or infant school might be credited with studying longer-term effects. In fact, the amount of exposure time observed may be more equivalent to the amount known to be required for younger starters to overtake older starters in naturalistic language acquisition (18 years, according to Singleton’s (1989) estimations on the basis of the findings made by Snow and Hoefnagel-Ho¨hle (1978, 1979)). However, the drawback of these studies is that they may depend too heavily on selfreports, and they are therefore deprived of reliable information about the individual learning trajectories. They also inevitably contain significant interindividual variability in exposure time in and out of the classroom. To complete this account of time requirements and age effects, it is also pertinent to note that experts in the area of L2 speech acquisition have suggested that simple estimations of LoR may not always constitute a valid index of L2 input. For example, the study by Flege and Liu (2001) shows that L2 proficiency increases with LoR only if the L2 speaker is participating in social settings, such as schools, in which they receive a substantial amount of input from native speakers of the L2. Thus, very young children who do

584


not yet attend school may not be exposed to the L2 immediately after arrival, while some immigrants may reside in L1 ghettos, or may not have a job outside the home that requires use of the L2. In addition, immigrants are likely to hear different dialects of the L2, compatriots from the home country who speak the L2 with a similar foreign accent, and individuals from other L1 backgrounds with different L1-inspired foreign accents (Flege, in press). From his review of previous research that has revealed small or non-existent LoR effects (and significant age of arrival (AoA) effects), Flege infers that LoR may provide a valid index of L2 input only for immigrants who receive a substantial amount of native-speaker input. Hence, the quality in the input to which all learners are exposed must also be examined alongside the quantity of that input or LoR. In a foreign language learning setting, learners do not usually receive native-speaker input and, when they do, it is not substantial. Moreover, learners are not exposed to the target language during all hours of instruction, and the exact proportion of lesson time varies according to the different educational systems, schools, and teachers. As a result, it may be argued that the length of instruction alone may not be a valid index of L2 input either, and that the quantity and the quality of the input to which instructed learners are exposed must also be examined.

INITIAL AGE OF LEARNING AS THE BEGINNING OF SIGNIFICANT EXPOSURE Further asymmetries are revealed when considering the initial point of learning. It has been argued that estimations of LoR should take as the initial point the beginning of significant exposure.9 In naturalistic L2 acquisition studies, this corresponds to the age of acquisition, also referred to as age of onset. Birdsong (2006: 11) notes that age of acquisition is understood as the age at which learners are immersed in the L2 context, typically as immigrants.10 This landmark is separated from the age of first exposure, which can occur in a formal schooling environment, visits to the L2 country, extended contact with relatives who are L2 speakers and so forth. Although age at immigration is taken as a proxy for age at onset of second language learning in studies of L2 acquisition among immigrants (see Stevens 2006), the age of acquisition may not coincide with the age upon arrival, because residence does not guarantee exposure to and use of the L2, as seen above. Therefore, precise estimations of the amount of L2 input are crucial because it is often observed to be confounded with age of acquisition, since early learners have typically used the L2 longer than late learners (Flege et al. 1999). The development of detailed questionnaires that measure amount of L2 (and L1) input (e.g. Flege and MacKay 2004) is an important recent contribution. In studies of instructed foreign language acquisition, initial age of learning is equated with the age at which instruction begins. In order to decide whether this is ever significant exposure it is wise to examine what is meant by

CARMEN MUNÕZ

585

the term: significant exposure is widely conceived as immersion in the L2 context, which provides learners with a variety of contexts of use and interaction. In other words, it is considered that learners have significant exposure when they are able to carry out a variety of speech acts over a wide range of situations and topics, and to participate in social settings effectively dominated by the L2 (see Stevens 2006: 681). Following this argument, and because foreign language learners are deprived of significant exposure of this sort, initial age of foreign language learning may be equated with the age at which insignificant exposure begins. As in studies in naturalistic settings, however, it is also methodologically important to measure precisely the amount of L2 input because, in this type of setting, initial age of learning may be confounded with amount of exposure when groups of early and late learners with different amounts of accumulated instruction hours are compared. When learners move from one kind of language learning setting to the other, it may be possible to identify an age of first significant exposure that is different to the age of first insignificant exposure. As noted above, naturalistic learners may have had some instruction in the target language before moving to the L2-speaking community. In ultimate attainment studies there has usually been no role for prior insignificant exposure, and the age of initial exposure in classroom contexts has not usually been found to be strongly predictive. For example, in the study by Johnson and Newport (1989) the learners’ experience of formal learning before arrival in the L2-country did not correlate significantly with the test scores; see also Birdsong and Molis (2001). In contrast, the study by Urponen (2004) found that the age at which learning of English as a foreign language began prior to arrival in the L2-context, a significant predictor of ultimate attainment, was in fact better than age upon arrival. A different type of change of language learning setting occurs when learners in regular school programmes with limited exposure later join immersion programmes, as in the study by White and Genesee (1996) in which French-speaking Canadians who had studied English as a compulsory school subject for at least the period of secondary education joined an English-medium university. White and Genesee (1996: 242–3), arguably (see Long 2005), took this latter point in time as the age of onset, on the basis that this was the first significant exposure to the target language. A change of language learning setting may also take place for only a limited period of time, as when foreign language learners spend a few weeks in the target language community, usually attending language classes and sometimes living with a host family. In such circumstances, the stay-abroad experience may provide significant exposure even if only for a short period of time, and in some cases the experience is repeated periodically over the years. Clearly, research is needed that looks into the impact that these short stays make on learners’ L2 proficiency in relation to their initial age of learning (with limited exposure) and their age at the beginning of significant exposure.

586


Turning now to examine the relationship between age of acquisition and second language outcomes, in naturalistic studies there is a wide consensus about the importance of initial age of learning on second language achievement (for recent reviews of experimental studies, see Birdsong (2005) and DeKeyser and Larson-Hall (2005)). For many researchers who defend the existence of a maturationally defined critical or sensitive period, a necessary though not sufficient condition for nativelikeness (see Hyltenstam and Abrahamsson 2003) is to begin learning the second language during the sensitive period (see also Flege et al. 1999). For example, in Long’s (1990) formulation of the maturational state hypothesis, it is explicitly stated that even beginning to learn (my emphasis) the second language during that period is sufficient. In contrast, in foreign language learning studies the consensus concerning the relationship between age of acquisition and second language outcomes is that older starters have a faster rate of learning (e.g. Cenoz 2002, 2003; Garcıá Lecumberri and Gallardo 2003; Garcıá Mayo 2003; Lasagabaster and Doiz 2003; Munõz 2003, 2006c; Nave´s et al. 2003; Perales et al. 2004; A´lvarez 2006; Miralpeix 2006; Mora 2006; Torras et al. 2006). As seen above, no evidence exists that an early start in foreign language learning leads to higher proficiency levels after the same amount of instructional time, and even younger starters with more instructional time have often failed to show a particularly substantial advantage in terms of long-term proficiency benefits. To discuss this asymmetry one can turn to the formulation of the maturational state hypothesis above. It can certainly be argued that the condition that it is sufficient to begin learning the second language during the sensitive period could only be relevant to foreign language learning if, during the privileged period, children have access to significant exposure, which marks the beginning of acquisition. Similar considerations have led researchers to argue that the CPH is irrelevant to formal language acquisition (Patkowski 1994, 2003; Lightbown 2000; DeKeyser and Larson-Hall 2005: 101). DeKeyser’s argument (e.g. 2000, 2003) is based on an interpretation of the CPH according to which it applies only to implicit language acquisition (acquisition from mere exposure to the language). Children necessarily learn implicitly, DeKeyser argues, and implicit learning requires massive input. But in regular school programmes children are not provided with the massive amounts of input which constitute a necessary condition for triggering the implicit learning mechanisms that characterize their language learning capacity (2000: 520). Further, while a lack of significant exposure in the classroom deprives young learners of the possibility of using implicit learning mechanisms, the older learners’ superior cognitive development allows them to make more efficient use of faster, explicit learning mechanisms (Munõz 2001, 2006c). To conclude, turning again to the maturational state hypothesis and its applicability to foreign language learning settings, it is arguable that age at the initial point of learning has as much influence in that context as in

CARMEN MUNÕZ

587

naturalistic learning settings. In fact, with low intensity of exposure, foreign language learning proceeds in a slow cumulative way through multiple and discontinuous short amounts of limited exposure. Therefore, in a lowintensity or distributed programme, the whole age range that extends during the period in which foreign language learning takes place may have a stronger influence on the process and the outcome of second language learning than the initial point of learning.

LEARNERS’ CHRONOLOGICAL AGE Learners’ chronological age at time of testing (AT) is a biographical variable that has drawn the attention of a number of researchers, most of whom have been critical of the methodologies used in studies of the effects of age on second language acquisition. It has been argued that AT may impact on ultimate attainment by confounding with cognitive factors, education, and other background variables (Bialystok and Hakuta 1999). Hakuta et al. (2003) attribute the steady decline across the life-span (from age 5 to 60) in their census data to cognitive declines of older learners, which ‘gradually erodes some of the mechanisms necessary for learning a complex body of knowledge, such as a new language’ (2003: 31). The learning abilities that are relevant to language learning include, according to Hakuta et al. (2003), the ability to learn paired associates (Salthouse 1992), the ability to encode new information (Craik and Jennings 1992) and to recall detail as opposed to gist (Hultsch and Dixon 1990). In a similar vein, Birdsong (2006) has highlighted the difference between the effects of aging on L2 learning (i.e. effects in the temporal-associative areas of the brain, particularly) from the effects of maturation on L2 learning, that is, the probability of nativelike ultimate attainment if learning begins before the end of the sensitive period. Other researchers, however, have argued against the claim that the confound between age of onset or age on arrival and age at testing may partly explain the inverse correlations found between AoA and test scores in some ultimate attainment studies. Such a claim would diminish the strength of the effects of maturational constraints. Specifically, DeKeyser (2000) reports a study in which the correlation between test scores and AoA is still significant with age at test partialled out, whereas that between test scores and AT is non-significant. A new perspective on the interplay of these variables has been brought to the field of SLA by Stevens’s (2006) discussion of the linear dependency of age at immigration, length of residence, and age at testing. The linear dependency among trios of variables has been extensively discussed by demographers in relation to what is known as the age–period–cohort problem or conundrum (see e.g. Wilson and Gove 1999). This linear dependency is expressed by Stevens (2006: 672) by means of an equation, which can be rewritten as AT ¼ LoR þ AoA. A methodological consequence is that the age–length–onset problem cannot be unambiguously addressed by correlational analyses,

588


and hence it requires a more careful consideration of the concepts and conceptual processes that these three variables index. Stevens (2006) uses Johnson and Newport’s (1989) landmark study to illustrate how difficult it is to reach unambiguous conclusions when investigating the relationship between age at arrival and L2 proficiency among adult immigrants. Although that classic study considers only age at arrival and length of residence, Stevens argues that the interdependence between age, length of residence, and age at onset entails a negative relationship between chronological age and L2 proficiency, which provides grounds for some ambiguity about the conclusion concerning the importance of age at onset, interpreted as the only predictive factor. In addition, Stevens claims that the negative relationship may be explained by the fact that the participants in Johnson and Newport’s study were faculty and students at a university; the older participants in the study were more likely to be faculty and the younger ones to be students. The types of experience the two groups have with the language may be very different and much more intense and demanding in the case of students than for the faculty, which may explain, at least in part, the relationship between age at immigration and success in L2 acquisition found in the study (see also Bialystok and Hakuta 1994, 1999). In instructed foreign language learning, AT may also be seen as confounding with age of onset, but in this case it has a negative effect on the performance of the youngest learners in comparison with older learners in school settings, and thus contributes to the positive relationship between L2 proficiency and older age of learning. Specifically, AT may have an impact on test-taking skills, favouring older children, adolescents, and adults over younger children when taking language tests. In consequence, the older learners’ higher scores in cognitively-demanding tests might be attributable not only to a language proficiency advantage over younger learners, but also to their superior cognitive development, which helps them achieve a better understanding of the task in comparison with younger learners as well as choose better strategies for accomplishing the learning task. This highlights the importance of the type of tasks that can be used with different-age learners in order to ensure that the effects of AT are not confounded with AoA effects (see Munõz 2006c). Chronological age is not just an indicator of biological processes associated with senescence; it is also, as illustrated in the discussion above, an excellent indicator of life-cycle stage, strongly associated with motivations and opportunities to speak and to maintain or improve proficiency in an L2 (Stevens 2006: 684). In instructed foreign language learning chronological age may also be seen as associated with attitudes towards learning the language, the teacher, and the classroom. A confound between initial age of learning and chronological age may also be suspected in studies that analyse learner motivation in relation to age. To illustrate, it may be argued that adult beginners who choose to enrol on a foreign language course may possess stronger motivation than younger learners following compulsory education

CARMEN MUNÕZ

589

courses, as it has been argued that students in late immersion programmes may be more strongly motivated than students in early immersion programmes (Turnbull et al. 1998). On the other hand, it has often been reported that younger school learners have a more positive attitude towards a foreign language than older school learners, as shown by their answers to questionnaires on motivation, and that this is a definite advantage of an early start (e.g. Hawkins 1996; Blondin et al. 1998). However, the young learners’ positive attitude may be seen as associated with chronological age rather than, or as well as, age at onset. It is well known that young children are more eager to please the adult than pubescent and adolescent learners and, consequently, the former may be more inclined than the latter to respond positively to motivation-related questions. Thus, the relatively common finding that younger starters have a more positive attitude towards learning a second language than older starters may be a result of their chronological age (they are younger) when they answer the motivation questionnaire rather than or in addition to their earlier start. This is clearly an empirical question since evidence could be provided by studies that asked attitudinal questions to learners with different AoA when they were the same chronological age. It was mentioned above that the amount of formal education at AT has been observed to be a very good predictor of second language learning among immigrants. Strong evidence has been provided by the large-scale study using census data in US conducted by Hakuta et al. (2003). Flege and coworkers have also considered the amount of education in the host country in their studies of L2 speech learning. For example, in the study conducted by Flege et al. (1999), AoA effects disappeared when education was controlled. In a more recent study, Flege et al. (2005) observe the influence on foreign accent of variables that were confounded with AoA and find that AoA was correlated with length of residence in Canada, which was correlated with years of education in the host country. In instructed foreign language learning, the type and characteristics of formal education may have a strong influence on learners’ proficiency levels. For one thing, in many countries, socioeconomic background may determine a child’s initial age of foreign language learning through the choice of schools (privately- or state-funded) as well as the amount of extracurricular exposure learners can enjoy (in the form of extracurricular lessons, technological devices in the home, and studies abroad, mainly). More importantly, the educational level of parents has a significant influence on children’s foreign language learning success (see for example the PISA report in Europe 2003), which suggests a parallel with the role played by the formal education of immigrant learners in studies of naturalistic L2 acquisition. Because children of parents from different socioeconomic backgrounds and with different educational levels are not homogeneously distributed in all types of school, the effects of AoA may be confounded with the effects of socioeconomic and sociocultural variables which result in varying availability of curricular and extracurricular exposure to the foreign language (Munõz 2006a).

590


CONCLUSION This paper has presented an analysis of symmetries and asymmetries that exist between a naturalistic learning setting and a foreign language learning setting with respect to those variables that are crucial in the discussion of age effects in second language acquisition. The analysis has highlighted important differences that have largely been ignored in such discussion; the reason being that due to the higher theoretical relevance granted to ultimate attainment studies in the field of SLA, the findings of classroom studies have been interpreted in the light of the assumptions and priorities of the former. One notable asymmetry is derived from the different status of ultimate attainment or final product in the two types of learning setting. Studies of age effects on naturalistic second language learning compare younger and older starters in terms of the final product of their second language learning process, bounded by a minimum length of time (i.e. at least 10 years of residence). In contrast, age-related studies in a foreign language setting, in which the conditions for reaching the end state are not met, compare the gains of younger and older learners after different lengths of time with an emphasis on long-term benefits. In such studies, long-term benefits are often temporally bounded by the educational system itself, which does not usually require foreign language learning beyond secondary education. In consequence, the amount that can be learned in such a limited period and with such limited exposure—that is, rate of learning—becomes a crucial concern. In relation to the long-term benefits of an early start, it is important to underline the absence of empirical evidence to date confirming that younger starters overtake older starters in school settings with the same amount of input. Nevertheless, the generalization has been accepted in the field of SLA and extended beyond the field that, as occurs in a naturalistic learning setting, younger starters will overtake older starters in the long run. Likewise, it has been tacitly agreed that when younger starters show better outcomes than older starters in situations in which the former have had a greater amount of exposure or instruction, this is due to the learners’ earlier starting age and not to the greater instruction time to which they have had access due to their earlier start, ignoring this methodological flaw in variable control. Noticeable asymmetries are also found in the amount of exposure between the two types of learning setting. While in naturalistic second language learning settings input is generally unlimited, in foreign language settings input is, by definition, limited and it is usually distributed in very small doses. Therefore, it may be hard to find equivalent amounts of exposure, and research that takes into account such long periods of time (i.e. 18 years of uninterrupted instruction) is difficult to conduct with methodological rigour. Similarly, it has been pointed out in this paper that while in naturalistic second language learning studies initial age of learning is taken to be the age of first significant exposure, foreign language learning studies irremediably take as the initial age

CARMEN MUNÕZ

591

of learning the age of first insignificant exposure. A consequence of these considerable differences in amount and type of exposure is that a foreign language learning setting cannot provide the optimal learning conditions (input that is neither quantitatively nor qualitatively limited) that constitute the methodological requirement for establishing the cessation of learning. Notwithstanding, the generalization has again been made that the comparatively lower levels of language proficiency in classroom settings are due to starting age, and that they will automatically be improved by lowering the initial age of learning, disregarding the crucial role played by intensity of exposure in language learning. In summary, for a number of years now in the field of SLA there has been a tacit acceptance of sweeping generalizations of findings from natural settings to classroom settings that have not been upheld by research into the latter.11 An inferential leap has been made in the assumption that learning age will have the same effect on learners in an immersion setting as on students of a foreign language, when the latter are exposed to only one speaker of that language (the teacher) in only one setting (the classroom) and for only limited amounts of time (Munõz 2006b). However, recent studies conducted in foreign language settings have clearly illustrated the role of input and exposure in the equation: an early start leads to success but only provided that it is associated with enough significant exposure. Not only have findings been generalized; the aims and priorities of research in naturalistic settings have also been extended to research in classroom settings. As has been discussed in this paper, the aim of age-related classroom-based research cannot be the same as that of ultimate attainment studies, namely, to provide ‘evidence for the existence, scope or timing of maturational constraints on the human capacity for learning second (including foreign) languages’ (Long 2005: 288). This recognition frees age-related classroom research from endeavours which strictly belong to the area of CPH studies and provides it with a clearer focus. Similarly, legitimate priorities for research in foreign language learning settings that should have implications for guiding age-related classroom research may be: (i) to determine the amount of input that is required for an early start to be effective in promoting language learning; (ii) to focus on the relative gains of different-age pupils with different types of time distribution, that is the advantages of intensive learning programmes that may be plausible in schools; (iii) the distinction between short-term and long-term benefits of starting at different ages; and (iv) the comparative study of the learning rate of different-age learners, which should inform educators about what to expect after 4, 8, or n years of foreign language instruction from different-age learners. To conclude, the field of SLA should now be mature enough to recognize that it is ultimately the findings of studies in actual classrooms that are most relevant to decisions concerning the time and timing of second language instruction. It is these findings that have yielded a more complete picture of the factors that lead to success in second language learning, providing

592


evidence of the essential role of input, and thus contributing to the development of an integrated explanation of age effects on second language acquisition.

ACKNOWLEDGEMENTS This work was supported by research grants HUM2004–05167 and 2005SGR00778. The author thanks Teresa Nave´s, two anonymous reviewers, and the editors for their comments on the manuscript.

NOTES 1 The term second language will be used in this paper in a general sense to include all its different aspects (from phonology to discourse) notwithstanding the differences in the way they are learned and the speed and degree of success with which they are learned. 2 For example, in Europe the average number of foreign language classes is between 30 and 50 in primary education, and a little over 90 hours in secondary education (see Eurydice 2005). This amounts in many countries to around 800 hours by the end of compulsory education. 3 For example, in a study assessing pupils’ skills in English in eight European countries (Bonnet 2002) it is observed that ‘in most countries it is basically the teacher who does the talking’ (p. 93) and that teachers’ use of the target language varies among countries: only 15% of the Spanish teachers state that they always speak in English in their lessons, whereas 40% of teachers from Norway, Sweden, or Denmark usually do so. 4 For example, in the study referred to in note 4 (Bonnet 2002) only 8.5% of pupils state that they work in groups and talk in the target language most of the time.

5 Variation among countries is large and in the case of English the situation is more correctly depicted as a gradient or a cline (Berns 1990). See also Bonnet (2002). 6 Other terms used more or less interchangeably in the literature on naturalistic L2 learning are end state, final state, steady state, and asymptote (see Birdsong 2006: 11). 7 On the other hand, and on the basis of their recent research findings, Abrahamsson and Hyltenstam (2006) argue that nativelikeness alone may not be enough to assess the CPH, and claim that language aptitude seems crucial for reaching nativelikeness. 8 Cummins (1980, 1981) and Cummins and Swain (1986) argued that older learners show greater mastery of L2 syntax, morphology and other literacy-related skills, such as vocabulary and reading comprehension, due to their greater cognitive maturity. 9 MacWhinney (2005: 136) qualifies it as serious exposure. 10 In that respect, Long (2005) argues that estimations of LoR should take as the initial point the beginning of significant exposure even in cases in which periods spent in the L2 environment are interrupted or discontinuous, such as when learners spend different periods of residence in the L2 community.

CARMEN MUNÕZ

11 This is not exactly parallel to the divide between pure and applied research in SLA (see, for example, Sharwood-Smith 2004: 5), although some of the features of the latter may be reminiscent of it, such as the

593

direct use of results from basic theoretical L2 research in dealing with applied questions generated in instructional language learning situations. I thank an anonymous reviewer for the suggestion.

REFERENCES Abrahamsson, N. and K. Hyltenstam. 2006. ‘The role of language learning aptitude for near-native L2 attainment.’ Paper presented in the Joint AAAL and ACLA/CAAL Conference, Montreál, 17–20 June. ´ lvarez, E. 2006. ‘Rate and route of acquisition in A EFL narrative development at different ages’ in C. Munõz (ed.): Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters, pp.127–55. Berns, M. 1990. ‘ ‘‘Second’’ and ‘‘foreign’’ in second language acquisition/foreign language learning: a sociolinguistic perspective’ in B. Van Patten and J. F. Lee (eds): Second Language Acquisition–Foreign Language Learning. Clevedon: Multilingual Matters. Bialystok, E. and K. Hakuta. 1994. In Other Words: The Science and Psychology of Second-language Acquisition. New York: Basic Books. Bialystok, E. and K. Hakuta. 1999. ‘Confounded age: Linguistic and cognitive factors in age differences for second language acquisition’ in D. Birdsong (ed.): Second Language Acquisition and the Critical Period Hypothesis. Mahwah, NJ: Erlbaum, pp. 161–81. Birdsong, D. 2004. ‘Second language acquisition and ultimate attainment’ in A. Davies and C. Elder (eds): The Handbook of Applied Linguistics. Malden, MA: Blackwell Publishers, pp. 82–105. Birdsong, D. 2005. ‘Nativelikeness and nonnativelikeness in L2A research,’ International Review of Applied Linguistics 43/4: 319–28. Birdsong, D. 2006. ‘Age and second language acquisition and processing: A selective overview,’ Language Learning 56/1: 9–49. Birdsong, D. and M. Molis. 2001. ‘On the evidence for maturational constraints on second-language acquisition,’ Journal of Memory and Language 44: 235–49. Blondin, C., M. Candelier, P. Edelenbos, R. Johnstone, A. Kubanek-German, and

T. Taeschner. 1998. Foreign Languages in Primary and Pre-School Education. London: CILT. Bonnet, G. (ed.) 2002. The Assessment of Pupils’ Skills in English in Eight European Countries. Project commissioned by the European Network of Policy Makers for the Evaluation of Education Systems. http://cisad.adc.education.fr/reva/pdf/ assessmentofenglish.pdf. Burstall, C. 1975. ‘Primary French in the balance,’ Foreign Language Annals 10/3: 245–52. Cenoz, J. 2002. ‘Age differences in foreign language learning,’ ITL Review of Applied Linguistics 125–42, 135–6. Cenoz, J. 2003. ‘The influence of age on the acquisition of English’ in M. P. Garcıá Mayo and M. L. Garcıá Lecumberri (eds): Age and the Acquisition of English as a Foreign Language. Clevedon: Multilingual Matters, pp. 77–93. Cook, V. 2002. ‘Background to the L2 user’ in V. Cook (ed.): Portraits of the L2 User. Clevedon: Multilingual Matters, pp. 1–28. Craik, F. I. M. and J. M. Jennings. 1992. ‘Human memory’ in F. I. M. Craik and T. A. Salthouse (eds): The Handbook of Aging and Cognition. Hillsdale, NJ: Erlbaum, pp. 51–110. Cummins, J. 1980. ‘The cross-lingual dimensions of language proficiency: Implications for bilingual education and the optimal age issue,’ TESOL Quarterly 14/2: 175–87. Cummins, J. 1981. ‘Age on arrival and immigrant second language learning in Canada. A reassessment,’ Applied Linguistics 11/2: 132–49. Cummins, J. 1983. ‘Language proficiency, biliteracy and French immersion,’ Canadian Journal of Education 8/2: 117–38. Cummins, J. and M. Swain. 1986. ‘Bilingualism in Education’. London: Longman. DeKeyser, R. 2000. ‘The robustness of critical period effects in second language acquisition,’ Studies in Second Language Acquisition 22: 499–533.

594


DeKeyser, R. 2003. ‘Implicit and explicit learning’ in C. J. Doughty and M. H. Long (eds): Handbook of Second Language Acquisition. London: Blackwell, pp. 313–48. DeKeyser, R. and J. Larson-Hall. 2005. ‘What does the critical period really mean?’ in J. F. Kroll and A. M. B. de Groot (eds): Handbook of Bilingualism: Psycholinguistic Approaches. Oxford: Oxford University Press, pp. 88–108. Ellis, R. 1994. The Study of Second Language Acquisition. Oxford: Oxford University Press. Eurydice. 2005. Key Data on Teaching Languages at School in Europe. European Commission. Flege, J. in press. ‘Give input a chance!’ in M. Young-Scholten and T. Piske (eds): Input Matters. Clevedon: Multilingual Matters. Flege, J. and S. Liu. 2001. ‘The effect of experience on adults’ acquisition of a second language,’ Studies in Second Language Acquisition 23/4: 527–52. Flege, J. and I. MacKay. 2004. ‘Perceiving vowels in a second language,’ Studies in Second Language Acquisition 26: 1–34. Flege, J., G. Yeni-Komshian, and S. Liu. 1999. ‘Age constraints on second language learning,’ Journal of Memory and Language 41: 78–104. Flege, J., I. MacKay, and S. Imai. 2005. ‘What accounts for ‘‘age’’ effects on overall degree foreign accent?’ Manuscript submitted for publication. Florander, J. and M. Jensen. 1969. Skoleforsøg i engelsk 1959–1965. (School experiments in English 1959–1965). Stencilled. Denmark’s Institute of Education. Cited in Ekstrand (1976) ‘Age and length of residence as variables related to the adjustment of migrant children with special reference to second language learning’ in G. Nickel (ed.) Proceedings of the Fourth International Congress of Applied Linguistics. Volume 3. Stuttgart: Hochschulverlag. Reprinted in S. Krashen, R. Scarcella and M. Long (eds). 1982. Child–Adult Differences in Second Language Acquisition. Rowley, MA: Newbury House. Garcıá Lecumberri, M. L. and F. Gallardo. 2003. ‘English FL sounds in school learners of different ages’ in M. P. Garcıá Mayo and M. L. Garcıá Lecumberri (eds): Age and the Acquisition of English as a Foreign Language. Clevedon: Multilingual Matters, pp. 115–35. Garcıá Mayo, M. P. 2003. ‘Age, length of exposure and grammaticality judgements in the acquisition of English as a foreign language’ in M. P. Garcıá Mayo and M. L. Garcıá Lecumberri (eds): Age and the Acquisition of English as a Foreign

Language. Clevedon: Multilingual Matters, pp. 94–114. Garcıá Mayo, M. P. and M. L. Garcıá Lecumberri (eds). 2003. Age and the Acquisition of English as a Foreign Language. Clevedon: Multilingual Matters. Grosjean, F. 1989. ‘Neurolinguists, beware! The bilingual is not two monolinguals in one person,’ Brain and Language 36: 3–15. Grosjean, F. 1998. ‘Studying bilinguals: Methodological and conceptual issues,’ Bilingualism: Language and Cognition 1: 131–49. Hakuta, K., E. Bialystok, and E. Wiley. 2003. ‘Critical evidence: A test of the Critical Period Hypothesis for second language acquisition,’ Psychological Science 14: 31–8. Han, Z.-H. 2004. Fossilization in Adult Second Language Acquisition. Clevedon: Multilingual Matters. Han, Z.-H. and T. Odlin. 2005. ‘Introduction’ in Z.-H. Han and T. Odlin (eds): Studies of Fossilization in Second Language Acquisition. Clevedon: Multilingual Matters, pp. 1–20. Harley, B. 1986. Age in Second Language Acquisition. Clevedon: Multilingual Matters. Harley, B. 1998. ‘The outcomes of early and later language learning’ in M. Med (ed.): Critical Issues in Early Second Language Learning. Scott Foresman-Addison Wesley, pp. 26–31. Hawkins, E. 1996. ‘The early teaching of modern languages. A pilot scheme’ in E. Hawkins (ed.): Thirty Years of Language Teaching. London: CILT, pp. 155–64. Hultsch, D. F. and R. A. Dixon. 1990. ‘Learning and memory in aging’ in J. E. Birren and K. W. Schaie (eds): Handbook of the Psychology of Aging 3rd edn. San Diego, CA: Academic Press, pp. 259–74. Hyltenstam, K. and N. Abrahamsson. 2003. ‘Maturational constraints in SLA’ in C. Doughty and M. Long (eds): The Handbook of Second Language Acquisition. Malden, MA and Oxford, UK: Blackwell, pp. 539–88. Johnson, J. and E. Newport. 1989. ‘Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language,’ Cognitive Psychology 21: 60–99. Krashen, S. 1981. Second Language Acquisition and Second Language Learning. Oxford: Pergamon. Krashen, S., M. Long, and R. Scarcella. 1979. ‘Age, rate, and eventual attainment in second language acquisition,’ TESOL Quarterly 13/4: 573–82.

CARMEN MUNÕZ

Lapkin, S., M. Swain, J. Kamin, and G. Hanna. 1980. Report on the 1979 evaluation of the Peel County late French immersion program, grades 8, 10, 11 and 12. Unpublished report, University of Toronto, OISE. Lasagabaster, D. and A. Doiz. 2003. ‘Maturational constraints on foreign-language written production’ in M. P. Garcıá Mayo and M. L. Garcıá Lecumberri (eds): Age and the Acquisition of English as a Foreign Language. Clevedon: Multilingual Matters, pp. 136–60. Lenneberg, E. H. 1967. Biological Foundations of Language. New York: Wiley. Lightbown, P. 2000. ‘Classroom second language acquisition research and second language teaching,’ Applied Linguistics 21/4: 431–82. Long, M. 1990. ‘Maturational constraints on language development,’ Studies in Second Language Acquisition 12: 251–85. Long, M. 2003. ‘Stabilization and fossilization in interlanguage development’ in C. Doughty and M. Long (eds): The Handbook of Second Language Acquisition. Oxford: Blackwell, pp. 487–536. Long, M. 2005. ‘Problems with supposed counterevidence to the Critical Period Hypothesis,’ International Review of Applied Linguistics 43/4: 287–317. MacWhinney, B. 2005. ‘Emergent fossilization’ in Z.-H. Han and T. Odlin (eds): Studies of Fossilization in Second Language Acquisition. Clevedon: Multilingual Matters, pp. 134–56. Miralpeix, I. 2006. ‘Age and vocabulary acquisition in English as a foreign language’ in C. Munõz (ed.): Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters, pp. 89–106. Mora, J. C. 2006. ‘Age effects on oral fluency development’ in C. Munõz (ed.): Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters, pp. 65–88. Munõz, C. 2001. ‘Factores escolares e individuales en el aprendizaje formal de un idioma extranjero’ in S. Pastor and V. Salazar (eds): Estudios de Lingu¨ı´stica. Universidad de Alicante, Anexo 1: 249–270. Munõz, C. 2003. ‘Variation in oral skills development and age of onset’ in M. P. Garcıá Mayo and M. L. Garcıá Lecumberri (eds): Age and the Acquisition of English as a Foreign Language. Clevedon: Multilingual Matters, pp. 161–81. Munõz, C. 2006a. ‘Investigating age-related differences in SLA: Methodological issues’. Paper presented in the Symposium Doing SLA Research: Theoretical, Methodological and Ethical Issues. AESLA Conference, Madrid, 31 March.

595

Munõz, C. (ed.) 2006b. Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters. Munõz, C. 2006c. ‘The effects of age on foreign language learning: The BAF Project’ in C. Munõz (ed.): Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters, pp. 1–40. Nakuma, C. K. 2005. ‘Attrition: easy questions, difficult answers’ in Z.-H. Han and T. Odlin (eds): Studies of Fossilization in Second Language Acquisition. Clevedon: Multilingual Matters, pp. 21–34. Nave´s, T. 2006. ‘The long-term effects of an early start on writing in foreign language learning.’ Unpublished doctoral disertation. University of Barcelona. Nave´s, T., M. R. Torras, and M. L. Celaya. 2003. ‘Long-term effects of an earlier start. An Analysis of EFL written production’ in S. Foster-Cohen and S. Pekarek (eds): Eurosla-Yearbook, vol. 3. Amsterdam: John Benjamins, pp. 103–30. Oller, J. and N. Nagato. 1974. ‘The long-term effect of FLES: An experiment,’ Modern Language Journal 58: 15–19. Patkowski, M. S. 1994. ‘The critical period hypothesis and interlanguage phonology’ in M. Yavas (ed.): First and Second Language Phonology. San Diego: Singular, pp. 205–21. Patkowski, M. S. 2003. ‘Laterality effects in multilinguals during speech production under the concurrent task paradigm: Another test of the age of acquisition,’ International Review of Applied Linguistics 41/3: 175–200. Perales, S., M. P. Garcıá Mayo, and J. M. Liceras. 2004. ‘The acquisition of English sentential negation by bilingual (Basque/Spanish) children: The age factor in an institutional setting.’ Paper presented at the Annual Meeting of the American Association of Applied Linguistics, Portland, OR, May 1–4. Perani, D., J. Abulatebi, E. Paulesu, S. Brambati, P. Scifo, S. F. Cappa, and F. Fazio. 2003. ‘The role of age of acquisition and language usage in early, high-proficient bilinguals: An fMRI study during verbal fluency,’ Human Brain Mapping 19: 170–82. PISA. 2003. Learning for Tomorrow’s World—First Results from Pisa 2003. OECD Publishing. Salthouse, T. A. 1992. Mechanisms of Age–Cognition Relations in Adulthood. Hillsdale, NJ: Erlbaum. Selinker, L. and J. Lamendella. 1979. ‘The role of extrinsic feedback in interlanguage fossilization: A discussion of ‘‘rule fossilization: A tentative model’’’, Language Learning 29/2: 363–75.

596


Sharwood-Smith, M. 2004. Second Language Learning: Theoretical Foundations. London: Longman. Singleton, D. 1989. Language Acquisition: The Age Factor. Clevedon: Multilingual Matters. Singleton, D. 1995. ‘A critical look at the Critical Period Hypothesis in second language acquisition research’ in D. Singleton and Z. Lengyel (eds): The Age Factor in Second Language Acquisition. Clevedon: Multilingual Matters, pp. 1–29. Singleton, D. and L. Ryan. 2004. Language Acquisition: The Age Factor 2nd edn. Clevedon: Multilingual Matters. Snow, C. and M. Hoefnagel-Ho¨hle. 1978. ‘The critical period for language acquisition: Evidence from second language learning,’ Child Development 49: 1114–28. Snow, C. and M. Hoefnagel-Ho¨hle. 1979. ‘Individual differences in second language ability: A factor analytic study,’ Language and Speech 22: 151–62. Stevens, G. 2006. ‘The age-length-onset problems in research on second language acquisition among immigrants,’ Language Learning 56/4: 671–92.

Swain, M. 1981. ‘Time and timing in bilingual education,’ Language Learning 31: 1–16. Torras, M. R., T. Nave´s, M. L. Celaya, and C. Pe´rez-Vidal. 2006. ‘Age and IL development in writing’ in C. Munõz (ed.): Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters, pp. 156–82. Turnbull, M., S. Lapkin, D. Hart, and M. Swain. 1998. ‘Time on task and immersion graduates’ French proficiency’ in S. Lapkin (ed.): French Second Language Education in Canada: Empirical Studies. Toronto, ON: University of Toronto Press, pp. 31–55. Urponen, M. I. 2004. Ultimate Attainment in Postpuberty Second Language Acquisition. Boston University: Boston. White, L. and F. Genesee. 1996. ‘How native is near-native? The issue of ultimate attainment in adult second language acquisition,’ Second Language Research 12/3: 233–65. Wilson, J. A. and W. R. Gove. 1999. ‘The age–period–cohort conundrum and verbal ability: Empirical relationships and their interpretation,’ American Sociological Review 64: 287–302.

Applied Linguistics 29/4: 597–618 ß Oxford University Press 2008 doi:10.1093/applin/amn010 Advance Access published on 6 May 2008

Textual Enhancement of Input: Issues and Possibilities ZHAOHONG HAN, EUN SUNG PARK, and CHARLES COMBS Teachers College, Columbia University The input enhancement hypothesis proposed by Sharwood Smith (1991, 1993) has stimulated considerable research over the last 15 years. This article reviews the research on textual enhancement of input (TE), an area where the majority of input enhancement studies have aggregated. Methodological idiosyncrasies are the norm of this body of research. Seven major issues appear to be limiting the generalizability of the findings and holding up further progress in the understanding of the efficacy of TE for learning: (1) noticing and/or acquisition; (2) TE and comprehension; (3) simultaneous or sequential processing; (4) TE and the nature of the enhanced form; (5) TE and prior knowledge; (6) TE and input flood; and (7) TE and overuse. The existing research has nonetheless offered some important insights that future research should seek to build on.

Over the last 15 years, there has been a proliferation of studies on ways to work with, and more importantly, against second language (L2) learners’ natural, meaning-exclusive tendency for input processing (VanPatten 1996, 2004). The studies were guided by a number of theoretical and/or pedagogical proposals, which include, but are not limited to, Sharwood Smith’s (1991) ‘input enhancement’, Long’s (1991) ‘focus on form’, Gass’ (1988) model of second language acquisition (SLA), and VanPatten’s (1991) ‘processing instruction’. The empirical findings so far have largely been inconclusive. In this article, we review the research conducted within the framework of Sharwood Smith’s input enhancement, delving, in particular, into 21 studies involving textual enhancement (TE, hereafter). Our goal is to explore and examine underlying issues, with a view to providing an interpretation of the conflicting findings, taking stock of insights, and identifying directions for future research.

WHAT IS INPUT ENHANCEMENT? The question that has served as an impetus for Sharwood Smith’s (1991, 1993) proposal of input enhancement can be formulated as follows: Why is it that L2 learners typically appear to ignore a vast mass of evidence and continue, obstinately, to operate with a system that is in contradiction with the target norms as manifest in the input? (adapted from Sharwood Smith 1993: 168)

598

TEXTUAL ENHANCEMENT OF INPUT

For Sharwood Smith, the answer to the question is multi-faceted. First, L2 learners, in general, lack sensitivity to grammatical features of target language input. Consequently, even when a large amount of input is available in their learning environment, they may not benefit much from it. Second, certain grammatical features in the input to which the learners are exposed are inherently non-salient, and hence their presence often escapes the learners’ attention. Third, learners’ first language (L1) may act as a hindrance to their ability to notice certain linguistic features in the input. Thus, failure to benefit from input may arise from a combination of lack of noticing ability on the learner’s part and poor input characteristics such as lack of perceptual salience or ‘noticeability’ (Schmidt 1990). Accordingly, Sharwood Smith hypothesizes that a way to stimulate input processing for form as well as meaning, and therefore language learning, is through improving the quality of input. Specifically, he proposes input enhancement, an operation whereby the saliency of linguistic features is augmented through, for example, TE (e.g. color-coding, boldfacing) for visual input, and phonological manipulations (e.g. oral repetition) for aural input. The underlying assumption is that noticing is a prerequisite for intake (Sharwood Smith 1981; Gass 1988; Schmidt 1990). Input salience, Sharwood Smith further suggests, can be created by an outsider (e.g. a teacher) or by an insider (i.e. the learners themselves). Learners possess their own natural learning and processing mechanisms which can, in and of themselves, generate input enhancement—so-called ‘internally generated input enhancement’, which may or may not coincide with ‘externally generated input enhancement’ (as by a teacher or researcher). The learner’s mind, as Sharwood Smith sees it, is not singular or global, but rather modular in character; the learner has many minds—to use his term—vis-a`-vis different linguistic domains and subsystems. Consequently, when exposed to externally enhanced input, learners (a) may or may not notice it, or (b) may notice it partially, contingent on whether or not they are ready for it or how much overlap there is between externally and internally generated salience. A mismatch may, therefore, arise ‘between the intentions lying behind teacher or textbook generated enhancement of the input and the actual effect it comes to have on the learner system’ (Sharwood Smith 1991: 130). Hence, ‘whether the enhanced input will ultimately trigger the relevant mental representation is . . . an empirical question’ (p. 120). Precisely, it is this question that has spurred a considerable amount of empirical research, to which we will turn in the next section. To facilitate an understanding of the empirical research, a further note on externally created salience is in order. Sharwood Smith (1991) posits two variables, elaboration (i.e. duration) and explicitness (i.e. metalinguistic depth), to allow permutations of a multitude of input enhancement strategies. Thus, one-time use of underlining would count as a non-elaborate, non-explicit strategy; one-time rule presentation would be a non-elaborate yet explicit strategy; and so forth. Sample strategies given in Sharwood Smith (1993)

ZHAOHONG HAN, EUN SUNG PARK, and CHARLES COMBS

599

include explicit discussion of the form, metalinguistic description, explicit or implicit error correction, and TE—the focus of the present article (for a review of first language research on TE, see Lorch 1989). Whatever its elaboration and explicitness, input enhancement is premised on the notion that learners must comprehend what they read or hear before their attention can be drawn to form within the input (cf. Gass 1988; VanPatten 1996, 2004; Wong 2005).

RESEARCH ON TEXTUAL ENHANCEMENT A survey of the L2 literature since the 1990s yielded 21 empirical studies of TE,1 most of which adopted a comparative approach whereby TE was pitted against another strategy, such as rule presentation (Doughty 1991; Alanen 1995), output production (Izumi 2002), or exclusively meaning-based communicative teaching (Leeman et al. 1995). The studies were typically controlled experiments under meaning-oriented task conditions, with a pretest–posttest design, the treatment period varying between 15 minutes (e.g. Alanen 1995) and 2 weeks (e.g. J. White 1998), with the posttest usually following immediately. On most occasions, the studies targeted one or more morphosyntactic elements2 such as locative suffixes in a semi-artificial form of Finnish (Alanen 1995), English relative clauses (Izumi 2002), and Spanish preterit versus imperfect forms (Jourdenais et al. 1995). These features were chosen largely based on considerations of difficulty, learnability, semantic content, communicative value, perceptual salience, and natural occurrence and frequency. Participants were mostly adult second language learners in the United States, and the sample size varied between 14 (Jourdenais et al. 1995) and 259 participants (Lee 2007). Some studies involved learners with little prior knowledge of the target form (Alanen 1995), and some had learners with considerable prior knowledge (Doughty 1991). The mode of exposure varied from reading printed texts (Alanen 1995) to reading computer-mediated texts (Doughty 1991). Although all studies employed TE, few did so singly. Rather, the majority of the studies employed additional means to augment the effect of TE, such as ‘an explicit mention to the learner to attend to the highlighted form’ (Izumi 2002: 543), a memory-based recall task (Williams 1999), or activation of prior knowledge (Shook 1994). Another notable feature of the studies is that the post-instructional measurements employed were mostly form-oriented. Examples are: sentence completion (Izumi 2002), grammaticality judgement (Alanen 1995), sentence combination (Doughty 1991), sentence production (Doughty 1991), sentence interpretation (Izumi 2002), picture-cued oral sentence completion (Doughty 1991), picture-cued written production (Jourdenais et al. 1995), multiplechoice (Bowles 2003), fill-in-the-blanks (Leow 2001), and error correction (Lee 2007). In most cases, statistical significance served as a prime, if not the only, indicator of whether or not TE was effective.

600


The 21 studies have yielded a spectrum of effects for TE, as sampled below (for a complete listing, see Appendix): 1 2

3

‘TE promotes noticing of target L2 form and has an effect on learners’ subsequent output’ (Jourdenais et al. 1995: 208). ‘Input enhancement failed to show measurable gains in learning despite the documented positive impact of enhancement on the noticing of the target form items in the input’ (Izumi 2002: 542). ‘The results indicated no significant benefit of TE over unenhanced input for (1) the amount of reported noticing of Spanish present perfect or present subjunctive forms, (2) learners’ intake of the forms, or (3) learners’ comprehension of the reading passage’ (Leow et al. 2003: 12).

Thus, in some studies TE was highly effective in that it led to noticing as well as acquisition; in some it was moderately effective in that it led to noticing but not to acquisition; but in others, it did not appear to be effective. This lack of congruence in the findings is arguably a natural consequence stimulated by the numerous methodological idiosyncrasies characterizing the individual studies, some of which have been noted above. In the main, the studies differ in the following aspects: (1) employing simple versus compound enhancement; (2) employing isolated words versus sentences versus discourse as stimuli; (3) enhancing a meaning-bearing versus a nonmeaningful form; (4) employing learners with or without prior knowledge of the target form; (5) enhancing the target form many versus one or a few times; (6) using a longer versus a shorter text; (7) employing a single versus multiple short sessions over an extended period of time; (8) enhancing one form versus multiple forms; (9) providing (or not) comprehension support prior to the treatment; and (10) providing (or not) explicit instruction on what to focus on prior to the treatment. By way of illustration of (10), in Shook (1999), the following instructions were given to both the experimental groups and the control group: Read the following article through so that you understand the information presented. You will be asked to recall the information from the article, so focus on comprehending the passage. (Shook 1999: 47) Here the researcher explicitly oriented his participants to reading for meaning. In so doing, he was hoping to see if the groups exposed to the visually enhanced input would simultaneously focus on the forms as well. Such practice contrasted with that of Leeman et al. (1995) and J. White (1998) where there were explicit instructions to the experimental group to focus on the forms while reading the enhanced material. Given the wide array of differences, the studies are not comparable, hence the difficulty of making any extrapolations. Each difference contributes a piece of divergence, essentially a variable, to what overall appears to be a confusing picture.


601

UNDERLYING ISSUES The methodological idiosyncrasies, however, were not random. Rather, they were underlain by conceptual differences over, inter alia, these issues: (1) noticing and/or acquisition; (2) TE and comprehension; (3) simultaneous or sequential processing; (4) TE and the nature of the enhanced form; (5) TE and learner prior knowledge; (6) TE and input flood; and (7) TE and overuse. Further advances in the general understanding of TE are to depend on how well these issues are to be resolved.

Noticing and/or acquisition This issue concerns what to expect of TE. The assumption that appears to have underpinned the design of most of the studies is, as Izumi (2002) put it, a two-step logic: First, the perceptual salience created by highlighting the input will draw the learner’s attention to the highlighted forms. Second, once the first step is successful, learning of the attended form will occur based on the premise that attention is what mediates input and intake. (Izumi 2002: 567–8) However, in reality, the first step has more or less been eschewed in the TE research, as many have noted. Jourdenais et al. (1995), for example, observed that ‘while [the] effectiveness of input enhancement has been evaluated in terms of subsequent acquisition of the target forms, few attempts have been made to investigate whether enhanced input is processed differently by learners’ (p. 183). Similarly, Leow et al. (2003) pointed out that ‘most of the studies . . . did not methodologically measure learners’ noticing while exposed to the experimental L2 data’ (p. 2). Hence, the issue of internal validity arises. One does not have to look beyond the abstracts of the 21 studies to notice a test bias in assessing the efficacy of TE: The majority of the studies solely invoked so-called acquisition measures for pre- and post-tests (cf. Wong 2003). Thus, most researchers have more or less equated the efficacy of TE with its ability to generate acquisition, where acquisition is associated mainly with improved accuracy in production. There are, nevertheless, studies, though few in number, that have sought to measure noticing and acquisition—using one or more tasks (e.g. Alanen 1995; Jourdenais et al. 1995; Leow 2001; Izumi 2002; Bowles 2003; Park 2004). For these researchers, the primary function of TE is to serve as a priming device for learners’ noticing of features in input, whose corollary may, then, be that what is noticed translates into acquisition (cf. Leow et al. 2003). This understanding is largely in accord with the tenor of the input enhancement hypothesis. ‘Noticing a form is a prerequisite for its acquisition’ (Leeman et al. 1995: 222), yet it is not enough, Sharwood Smith points out, for learners to just notice critical features in the input and/or to detect the anomaly in

602


their output. What is crucial is whether they can act upon the noticed features, and this would depend on whether or not they are able to perceive, store, and apply the salient information. Accordingly, it is hoped that input enhancement will spark a chain of cognitive processes initiated by noticing.3 Whether these processes can be, and how much time is needed for them to be, set in motion and completed has yet to be empirically ascertained. This issue itself in need of independent research, many studies appear to have operated on the assumption that if TE has any effect, it should show up instantly in learners’ production. This is aptly seen in the practice of providing a short treatment and an immediate, output-based posttest to measure acquisition (e.g. Alanen 1995; Jourdenais et al. 1995). Just as it is problematic to assume, without establishing, that TE draws learners’ attention to the target form, which, in turn, leads to further processing of it, it is equally problematic to expect instant learning to arise from a single (typically, short) treatment session of TE (cf. Wong 2005). Both create validity and reliability concerns. In the latter case, notably, because the treatment does not allow learners time to pursue deeper processing of any of the attended input; changes (or lack thereof) shown on the posttest can be evidence of anything but learning. Although it may be a truism that nothing in input can become intake without noticing (Schmidt 1990, 2001; see, however, Truscott 1998), the TE literature offers important evidence showing that noticing is not a guarantee for acquisition. For example, Izumi (2002) found that more externally induced noticing was not correlated with more learning, concluding that noticing may not lead to acquisition. Similarly, Williams (1999) noted that ‘not everything that is registered by the senses is . . . encoded in long-term memory’ (p. 2). Likewise, Leeman et al. (1995) observed that ‘forms may be noticed perceptually, but not linguistically’ (p. 219; cf. Doughty 1998). In a nutshell, enhanced forms may attract attention but may fall short of further processing. Taking the possible pitfalls into consideration, Sharwood Smith (1991), in his critical elaboration of input enhancement, cautions: Externally generated input enhancement does not automatically imply the internalization of that enhancement by the learner . . . Input enhancement should take into account a modular view of the learner as a set of systems: signaling information to the learner is, in effect, signaling to one or more of many separate knowledge systems. Input enhancement may . . . work in ways unforeseen by the enhancer. (Sharwood Smith 1991: 131; emphasis added) Implied here is (a) that not all the cognitive effects triggered by TE will have external manifestations, and (b) that the efficacy of TE is not as much controlled by the researcher as by the learner (cf. Bardovi-Harlig and Reynolds 1995; Wong 2005). As we will argue later, the efficacy of TE is, in part, a function of the learner’s prior knowledge (or lack thereof) and of the nature of the linguistic element enhanced.


603

Evidence that learners are active controllers of the input enhancement process can be found in Leeman et al. (1995), reporting that ‘not all of the enhancements were noticed by all the participants’ and ‘at least some learners prioritize meaning over form despite the emphatic instructions they received to consider both’ (p. 248). What such findings suggest is that learners have their own agenda to pursue, an agenda that may not be susceptible to instructional guidance (cf. Allwright 1984). In sum, it appears that short-term TE research—both in terms of the duration of the treatment and the interval between the treatment and the posttest—should concentrate on measuring its effect at the level of noticing and refrain from measuring acquisition (cf. Shook 1994; Overstreet 1998), which we define here as mastery of form–meaning–function relations.

TE and comprehension As indicated earlier, the input enhancement proposal is premised on comprehension, its underlying thrust being to prompt occasional metalinguistic attention for the ultimate benefit of balanced development in comprehension and acquisition (cf. Doughty 2004; Wong 2005; Kim 2006). It thus follows that any validation of its efficacy should investigate comprehension as well as noticing and, in the case of longer-term studies, acquisition. However, most of the studies have overlooked the need to measure comprehension (cf. Leow 1997; Wong 2003; Lee 2007). Thus, in spite of accumulated evidence suggesting that TE can promote noticing of some linguistic features, it remains unclear whether or not it has simultaneously created a trade-off with comprehension.4 Preliminary evidence from Overstreet (1998), substantiated recently by Lee (2007), suggests that input enhancement may detract from learners’ attention to meaning. This should, accordingly, raise a red flag for future TE researchers. In most of the existing studies, researchers have made a conscious effort to establish a meaning-oriented environment. Some had their participants answer comprehension questions before the onset of the treatment (Alanen 1995), some provided vocabulary assistance (Williams 1999), some had their participants write recall summaries, often in their L1, of the content of the texts they had read during the treatment (Izumi 2002), and some even took care not to make the enhancement so salient that it would distract students from reading (J. White 1998).5 Still, the literature has not yet witnessed any study with a rigorous design for measuring comprehension. The most concerted effort yet has been seen in trying to obtain a single summative score from the participants’ recall summaries, as in Doughty (1991) and Izumi (2002), or putting participants through a post-treatment comprehension test, as in Leow et al. (2003). Inasmuch as these practices did not include a pre- and post-treatment comparison of comprehension, they shed little light on the effect of TE, if any, on comprehension. A more robust design to measure the latter, therefore, should treat comprehension as a dependent

604


variable to be measured both before and after the TE treatment, at withingroup and/or between-group levels. When measuring comprehension in relation to TE, it also appears that the focus should be on measuring local, as opposed to global comprehension, an insight from Leow (1997). Local comprehension here concerns only the part of the text that contains the enhanced form, and hence is arguably a more valid indicator of the influence, if any, of TE on comprehension; it also appears to be a good predictor of global comprehension, as demonstrated in Leow (1997). Other studies such as Shook (1999) and Wong (2003), however, have provided evidence that the effect of TE is positively felt at the level of local comprehension, but not at the level of global comprehension, thereby suggesting a dissociation between local and global comprehension. Future research controlling for local and global comprehension should shed further light on this issue.

Simultaneous or sequential processing Many studies had as their underlying assumption that a focus on form integrated into an overall focus on meaning is beneficial to L2 acquisition (Lightbown 1991; Long and Robinson 1998). However, there was little consistency among them with regard to when such integration should occur. According to information processing theory, (a) processing of information is selective; (b) individuals can process two different types of information simultaneously and effectively only if the processing of one of the information types is automatized and requires little, if any, conscious attention; and (c) simultaneous processing of two different types of information that are not automatized can lead to inadequate processing of either or both types of information, and to a ‘trade-off’ effect (cf. Skehan 1996; VanPatten 1996; Han and Peverly 2007; see, however, Robinson 2003). From this theory it follows that sequential processing will be more effective than simultaneous processing of meaning and form. In other words, learners might benefit more from having their attention first directed to meaning decoding and then to grammatical encoding than to both tasks at the same time. The research on TE has employed both kinds of design, although most of the studies probed simultaneous processing. As an illustration, in Alanen (1995), participants in the TE group were invited to read two short passages in which the learning targets were italicized. At the outset of their reading, they were explicitly instructed to ‘try to understand its meaning’ (p. 271). Thus, an experimental condition was created to stimulate an intentional focus on meaning and an incidental focus on form. Such is also the thrust of Shook (1994, 1999). In Shook’s (1999) study, 73 participants were randomly assigned to three input conditions: (a) textually enhanced, (b) textually enhanced plus emphatic instruction,6 and (c) textually unenhanced. They subsequently performed a free written recall task in which they wrote down anything they could recall from the passages they had read. The recall task


605

sought to measure both comprehension and intake. The results were revealing: on the one hand, the unenhanced group recalled the most idea units, the enhanced group recalled fewer idea units, while the enhanced plus instruction group recalled the fewest, thus suggesting that the unenhanced group comprehended the reading passages better than both the enhanced and the enhanced plus instruction group. On the other hand, in terms of intake of the target items, the following picture emerged: participants who received TE recalled more tokens than did those who received TE plus instructions, while those who received no enhancement recalled the least tokens, thus evidencing a trade-off between comprehension and intake. The difference between simultaneous and sequential processing lies not just in the temporal order of meaning-based and form-based processing of input, but in that one calls for incidental processing and the other intentional processing. In other words, the researchers who implemented a simultaneous processing design assumed that in processing the input for meaning, the processing of the enhanced forms would occur incidentally, as a byproduct, an assumption similar, though not identical, to that underlying Krashen’s Input Hypothesis. On the other hand, the researchers who had a sequential design assumed that with the processing of input for meaning happening first, attentional resources could be freed up and reallocated to the processing of form, an assumption in line with the fundamental tenets of cognitive theory of information processing.7 Although research on TE to date has not seen much of a sequential design, Doughty (1991) has offered some useful preliminary evidence. In her study, 20 participants, randomly assigned to one of the three groups, (a) the meaningoriented TE group (MOG), (b) the rule-oriented enhancement group (ROG), and (c) the control group (COG), first had to undertake a comprehension task to achieve global comprehension of the content of a reading passage prior to receiving the treatment. Thus, the study roughly had a sequential arrangement for a focus on meaning and a focus on form, followed, then, by a more fine-tuned sequential arrangement during the treatment. The treatment itself comprised a series of lessons, one per day over a 10-day period, carried out as follows: during each lesson, the participants in the MOG received chunks of the reading passage aided first by lexical and semantic explanation and then by TE, and those in the ROG received the same chunks of text enhanced by a presentation of rules concerning the target item. In contrast, those in the COG, as expected, received the same chunks of texts with no enhancement whatsoever. Each lesson ended with the participants writing recall summaries based on their comprehension of the texts. Results from the post-treatment tasks indicated that the MOG and the ROG did almost equally well and better than the COG, with respect to acquisition of the target feature, but the MOG displayed an advantage over both the ROG and the COG in terms of comprehension. Despite the fact that this study did not employ a pre- and a post-test task to measure and compare the participants’ comprehension, the TE employed did

606


not produce a trade-off, as evident in a twofold fact. First, the MOG and the ROG did equally well on the acquisition measures; and, second, the ROG and the COG did equally poorly on comprehension measures. If the TE had affected comprehension, the COG should have achieved greater comprehension than the ROG, because they were not ‘distracted’ by any sort of TE. What, then, explains the ‘double’ gains in comprehension and acquisition by the MOG, but not by either of the other groups? The answer is in fact simple: this group received assistance in both meaning and form processing, notably, sequentially. This combined and sequential treatment may have aided both comprehension and acquisition. While future empirical research should attempt to ascertain which grammatical items would benefit most from a combined meaning–form enhancement, evidence from studies by Doughty (1991) and Izumi (2002) suffices to motivate the understanding that when meaning is clarified before a focus on form, TE is likely to stimulate even growth in comprehension and acquisition. Izumi (2002) assessed the effects of an output-based approach versus those of a TE approach to integrating a focus on form with a focus on meaning. Sixty-one university level ESL learners served as participants in this study, and were divided into four groups: (a) the þoutput –enhancement group, (b) the þoutput þenhancement group, (c) the –output –enhancement group, and (d) the –output þenhancement group. The experimental design largely followed that of Doughty (1991). Thus, it began by having the participants perform content-based processing (i.e. skimming the text for overall comprehension) before subjecting them to an extended period of focus-on-form treatment, hence, similarly, a sequential design. The treatment procedures for the enhanced and unenhanced groups, although bearing some similarities to the Doughty study, are nevertheless different in two main respects: First, the participants were asked to take notes while reading chunks of the reading passage, a means to measure noticing. Second, exposure to TE was not preceded by a lexical and semantic explanation of the target item; instead, the participants were directed through instructions to focus on meaning throughout the treatment period. Results showed that TE was able to promote noticing without taxing comprehension. Whereas the two studies mentioned above have incidentally revealed an advantage for sequential over simultaneous processing, a study by VanPatten (1990) has purposely investigated the feasibility of simultaneous processing of meaning and form, albeit within a different theoretical framework8 than input enhancement. Two hundred and two university students of Spanish, representing three levels of proficiency, were randomly assigned to one of the four experimental conditions: (a) attention to meaning alone, (b) simultaneous attention to meaning and an important lexical item, (c) simultaneous attention to meaning and a grammatical functor, or (d) simultaneous attention to meaning and a verb form. Results showed that ‘[C]onscious attention to form in the input competes with conscious attention to meaning, and, by extension, that only when input is easily understood can learners attend to form as part


607

of the intake process’ (p. 296). The study thus provides direct evidence that processing input for form presupposes that the meaning of the input has been comprehended,9 and that the former, when occurring, is sequential to the latter (see, however, Wong 2001). This finding is replicated in the TE research. Overstreet (1998) and Lee (2007) both noticed a similar form– meaning trade-off. Apart from responding to the limited information processing capacity that characterizes all learners, sequential processing has an additional putative advantage, namely that it matches learners’ predilection for processing input first for meaning (Barry and Lazarte 1995; VanPatten 1996, 2004). Shook (1994, 1999) and Leeman et al. (1995) observed that some of their participants focused exclusively on meaning, in defiance of TE and the researchers’ admonition to focus on both meaning and form. In a case like this, sequential processing may take learners beyond such natural processing tendency through redirecting their attention to form. The importance of the latter aspect cannot be overstated, as research has shown that even when learners are able to reallocate attentional resources to form-based processing as a function of attainment of comprehension, they do not automatically do so (see, e.g. VanPatten 1990; Leeser 2004). Interestingly, however, J. White (1998) attributed this phenomenon to the fact that the input had been comprehended, noting that ‘the forms may not have been novel enough to attract the learners’ attention to the extent that was predicted’ (p. 103). Hence, there is an intriguing circular problem here: Does comprehension facilitate or hinder attention to form? Further research is clearly needed to tease out the intricacies between the two.

TE and the nature of the enhanced form A now well-established tenet in SLA research is that not all linguistic elements are created equal. Over the years, researchers have posited a number of parameters for describing inter-structural differences. They include, but are not limited to, communicative value, formal complexity, functional complexity, semantic load, perceptual saliency, and underlying rules. VanPatten (1996), for example, distinguishes between linguistic features of high, medium, or low communicative value, based, in turn, on their semantic value and structural redundancy. Accordingly, English -ing is deemed of high communicative value in that it carries semantic value and, syntactically, is non-redundant. In comparison, English 3rd person -s is of low communicative value, because it is structurally redundant. In addition to functional differences, linguistic features may differ according to their underlying rules. DeKeyser (1994) notes that some linguistic features follow categorical rules (e.g. English plurals), and some probabilistic rules (e.g. English irregular past tense verbs). In the instructed SLA literature, evidence abounds that not all forms are equal in terms of the effectiveness of instructional activities (Williams and Evans 1998). Form type can mitigate the effect of instruction through their

608


mutual interaction as well as interaction with other variables such as task demands, learner proficiency, prior knowledge, familiarity, content complexity, and availability of pragmatic information. In the TE literature, two studies have specifically investigated the nature of the target form and its interaction with different enhancement strategies. One study is Leow et al. (2003) in which two structures, the Spanish present perfect and present subjunctive, were investigated in connection with the presence or absence of TE. Analyses of the participants’ think-aloud protocols and their performance on an intake-oriented recognition task revealed more noticing and learning of the present perfect than of the present subjunctive. Of importance to note is that the structural effect overrode that of TE: both the enhanced and the unenhanced group noticed one structure better than the other. Similar findings are reported in Shook (1999) examining the free written recalls produced by participants in the wake of TE in relation to the Spanish present perfect and relative pronouns. Both studies thereby lend empirical support in favor of Sharwood Smith’s notion of learner-created salience. In light of these findings, it is likely, then, that the positive effects of TE, as reported in some of the studies, have in part derived from the fact that the researchers happened to have chosen linguistic elements that are susceptible to this type of intervention, although it is not entirely clear at the present time what determines the susceptibility (cf. Wong 2003). Current speculations include form–meaning/form–function transparency (Wong 2005) and the relationship between cognitive processing and grammatical elements (Izumi 2002). While future research should test these speculations, it will equally be worthwhile to build on and extend the existing research by replicating Leow et al. (2003) and Shook (1994, 1999), and by investigating other forms, ideally from other languages as well. If future research indeed demonstrates that only some forms are amenable to TE, then instruction utilizing this type of strategy should selectively target certain forms as opposed to any forms indiscriminately. The question of what forms to (or not to) enhance is a complex one, and the existing evidence is meager but suggestive. Preliminary evidence from Wong (2003), corroborated by Shook (1999), showed that enhancing a nonmeaningful form does not lead to better intake. On the other hand, as noted above, results from Leow et al. (2003; see also Leow 2001) and Shook (1994, 1999) revealed that learners are able to notice, on their own, meaning-bearing forms, suggesting, therefore, that there is no need to enhance them (see, however, Leow 1997; Shook 1999 for a different interpretation). This reasoning would seem logical were it not for the results of Leow (1997) comparing the effects of TE and text length on comprehension and intake. One of the findings of the study, although not statistically significant, illustrates a meaningful effect for enhancement on the participants’ comprehension, while no such effect was discerned for their intake. Pending future, independent verification, this finding underscores that comprehension and intake are different entities (cf. Færch and Kasper 1986; Sharwood Smith 1986),


609

inasmuch as enhancing the same linguistic element can produce a differential impact on comprehension and intake (cf. Wong 2003, 2005).

TE and learners’ prior knowledge One approach to tackling the susceptibility issue is through examining variables that can interact with the target form to modulate the effect of TE. Learners’ prior knowledge is one such factor. Evidence thereof can be found in studies that employed participants with or without prior knowledge of the target form (e.g. Jourdenais et al. 1995 vs. Leow et al. 2003). Research to date has offered three main findings on the relationship between TE and learners’ prior knowledge. First, simple enhancement (i.e. TE alone) is more effective for learners with some prior knowledge of the form in question (Park 2004) than for learners without (Alanen 1995). Second, simple enhancement may induce noticing (i.e. low-level awareness, following Schmidt 1990) but not understanding (i.e. high-level awareness) in learners with little prior knowledge (Shook 1994); however, it may incite understanding as well as noticing in learners with some prior knowledge (Lee 2007). Third, compound enhancement (i.e. TE in combination with other attention-getting strategies such as corrective feedback) is more effective than simple enhancement in inducing noticing, and further processing of, the target form in both types of learners (e.g. Doughty 1991; Leeman et al. 1995; Williams 1999). Given the nature of simple versus compound enhancement, these findings are logical. According to Sharwood Smith (1991, 1993) and as Robinson (1997) has demonstrated, TE is more an implicit than an explicit attention-focusing device. As such, its underlying purpose may not always be transparent, even to learners with some prior knowledge of the target form. Several participants in J. White’s (1998) study reported, via a post-treatment debriefing questionnaire, that they were not certain about the purpose of the enhanced forms. For learners with little prior knowledge of the target form, TE could be even more of a puzzle. Participants in the Leow (2001) study provided the following comments in their think-aloud protocols: ‘I don’t know what that is (means)’, ‘I don’t understand these underlined ones’, ‘I am not sure’, ‘I don’t know why this is underlined’, etc. (p. 502). Likewise, as shown in the protocols from Alanen (1995), some participants who reported noticing the use of italics had not considered a reason for its use. Thus, the fact that TE alone (i.e. simple enhancement) may fail to have a noticeable impact on learners’ comprehension and intake, as reported in Leow (1997, 2001) and Leow et al. (2003), might be due to the learners having no prior knowledge of the target form. Had those learners been exposed to compound enhancement, they would likely have demonstrated superior intake to that shown in the comparison group (i.e. who received unenhanced input).

610


TE and input flood Most studies in the TE literature exhibited varying degrees of conflation of TE and input flood, both being options of input enhancement. In the case of input flood, salience is created of the target of instruction by virtue of (usually, artificially engineered) frequency. As an illustration, compare Leow et al. (2003) with Shook (1999). In Leow et al. (2003), two passages were used as stimuli, one having 222 words and the other 227 words. In each passage there were 10 tokens of a target form, giving the frequency ratios of 1/22 for one passage and 1/23 for the other—meaning in one text, one out of every 22 words was an exemplar of the target form, and one out of 23 words in the other. Shook (1999), on the other hand, also utilized two written texts as stimuli, one containing 185 words and the other 217 words, but there were only 6 instances of the target item in each text, yielding frequency ratios of 1/30 words for one text and 1/36 words for the other. Thus, the two studies differ in the amount of exposure to the target forms created for the participants: There is more of an ‘input flood’ in Leow et al. (2003)10 than in Shook (1999). Neither, however, matches the degree of input flood found in J. White’s quasi-experimental study where the students were provided numerous instances of the target form over an extended period of time. These variations notwithstanding, problems may arise when input flood is used along with TE to control for the experimental text and at the same time to manipulate the control text—the so-called ‘unenhanced text’. For one thing, the results from the experimental group can be contaminated with ambiguity. For another, input flood may cancel out the effects of TE. As J. White (1998) noted in her study, ‘benefits resulting from the experimental treatment conditions were due to increased exposure through input flood of target forms and not to any other kinds of enhancement’ (p. 103). The available evidence, albeit limited, induces the hypothesis that input flood may have a greater attention-getting capacity than TE. Future research should examine this further, including ascertaining, in the case of input flood, how many exemplars would be optimal (Wong 2005).

TE and overuse Several studies have provided evidence that learners receiving TE oversupplied the target form; that is, their production of the target form extended to non-obligatory contexts. Close inspection of the experimental procedures underlying these studies indicates that not only was the relevant input typographically enhanced, but there was input flood as well, which, as mentioned, is itself a form of enhancement. Hence, the target forms were double-enhanced. In the Jourdenais et al. (1995) study, in which participants showed overuse of the target form, the stimulus text of 204 words contained 10 instances of the Spanish imperfect and 18 of the preterit, each enhanced


611

in three typographic ways (i.e. underlining, changing the font, bolding/ shadowing). Similarly, in Alanen (1995), participants’ overgeneralization of -lla and -ssa could be linked to the over-salience of the target forms which were not only highlighted but also exhibited a high frequency in the two stimulus texts. In one text of 87 words, they appeared 6 versus 5 times, and in the other text of 98 words, 7 against 8 times, as compared with the rest of the target forms which each made between 2 and 4 appearances in either stimulus text. Alanen herself recognized that ‘these errors could have been caused by instruction itself (Selinker 1972)’ (p. 289). Further evidence of overuse can be found in Leeman et al. (1995) where overgeneralizations may be attributed to the fact that the researchers overemphasized the target forms through enhancing them in and out of the classroom, including providing corrective feedback whenever possible and giving explicit instructions, in order to attract the learners’ attention. Clearly, over-enhancing the target form can be counterproductive. When the target form is unnaturally frequent and excessively salient in the input, it hurts, rather than facilitates, learners’ processing of meaning as well as form. A case in point is Overstreet (1998), where participants in the enhancement groups not only received an input flood via a short text, but each instance of the target form was highlighted in three ways: (a) font enlargement, (b) underlining, and (c) boldface or shadowing. Results obtained from measures of recognition, comprehension, and production revealed the absence of a positive effect for TE on the intake of the target forms but the presence of a negative effect on comprehension. Indeed, underlying the reported negative effect of TE, it may not only have been the fact that there was too much salience, but also that more than one form was enhanced in the same input text.11 The coalescence of these two factors might have confused, rather than enlightened, the learners. Enhanced as such, input may trigger learners’ noticing but may not lead to much understanding, as seen in Jourdenais et al.’s study. In sum, finding the right balance appears difficult but necessary between frequency and saliency (cf. Barcroft 2003). If TE is purported only to induce noticing, then, of course, the more frequent the target forms are in the input, the more salient they are likely to be (however, see Lorch 1989), and the more salient they are, the more likely it is that they will be noticed. This reasoning has driven most of the empirical studies to date. Nonetheless, it seems to be time to switch gears and investigate a different question: if TE is intended to stimulate higher-level awareness of form–meaning connections, to what extent should input be enhanced so as not to produce aberrant noticing to the detriment of acquisition?

CONCLUSION As Wong (2003) has noted, ‘The contribution of [TE] to SLA . . . is presently not clear’ (p. 18). The available literature has provided conflicting findings on

612


its efficacy. Taken at face value, these findings may suggest that TE is either helpful or unhelpful. Such an understanding is, however, the least desirable: from a theoretical standpoint, it neither validates nor invalidates the theory, and from a practical standpoint, little can be extrapolated from the ambiguous findings to teaching in the classroom. Although the jury is still out on the efficacy of TE for learning, the research to date has nonetheless produced a number of valuable and, importantly, testable insights, such as the following:

Simple enhancement is capable of inducing learner noticing of externally enhanced forms in meaning-bearing input. Whether or not it also leads to acquisition depends largely on whether the learner has prior knowledge of the target form. Learners may automatically notice forms that are meaningful. Simple enhancement is more likely to induce learner noticing of the target form when sequential to comprehension than when it is concurrent with comprehension. Simple enhancement of a meaningful form contributes to comprehension. Simple enhancement of a non-meaningful form does not hurt comprehension. Simple enhancement is more effective if it draws focal rather than peripheral attention. Compound enhancement is more likely to induce deeper cognitive processing than simple enhancement, possibly to the extent of engendering ‘overlearning’.

Future research should seek to test the robustness of, and/or substantiate, these insights. This effort will lead researchers into ascertaining psycholinguistic contingencies under which TE is able to trigger a chain of cognitive processes that result ultimately in acquisition. En route, researchers will gain a better understanding of the key modulating variables such as learners’ prior knowledge, the nature of the target form, learner processing capacity, and frequency of the enhanced element. They will perforce strive to operationalize the type and process of attention and memory adequately and rigorously in order to find out how what is noticed is further processed, retained, and eventually becomes transferable. Moreover, they will likely be able to determine, and even set up a benchmark for, ‘the extent to which learner attention should be directed’ (Doughty 2004: 187). Current research has provided evidence showing that compound enhancement is not always necessary (see Williams and Evans 1998; Berent et al. 2007). Furthermore, future research on TE must include a longitudinal perspective. SLA researchers at large have paid excessive lip service to the need for carrying out longitudinal studies, but have not yet been able to break the status quo (cf. Ortega and Iberri-Shea 2005). This is particularly surprising for research on the effects of instruction, since it is here—more than anywhere else—where study after study has conceded that its methodological


613

shortcoming is the lack of a longitudinal perspective. Non-longitudinal studies of effects of instruction have an inherent bias in favor of explicit instruction (for discussion, see Doughty 2004). Explicit instruction is known to be quick at provoking instant yet superficial changes (see, e.g. L. White 1991) in behavior, in contrast to implicit instruction whose impact on learning—especially in the domain of morphosyntax—may take time to build and be felt (see, e.g. Robinson 1996; Hulstijn 1997; Doughty 2004; Long 2007; Mackey and Goo 2007). A related issue that researchers at large have overlooked, but that TE researchers should examine, is the scope of form–meaning connection (FMC) in acquisition.12 If FMC concerns linking one surface form with one meaning, then explicit (in particular, short and intensive) instruction might be more effective than implicit instruction in drawing learner attention and provoking changes in representation and behavior. However, if FMC entails connecting a form with its multi-faceted meaning in a variety of contexts, protracted implicit instruction might prove superior to explicit instruction (cf. Han 2007). Clearly, unless the time issue is taken into account, any comparison between explicit and implicit instructional strategies, TE included, may risk being invalid and misleading.

ACKNOWLEDGEMENTS We are grateful to the anonymous reviewers for their insightful comments and helpful suggestions. Any errors are exclusively our own.

APPENDIX General findings from 21 studies on textual enhancement 1

2

3 4 5

‘[A]ttention to form, either via detailed analysis of structure or highlighting of target language (TL) structures in context, promotes acquisition of interlanguage (IL) grammar, but that only the latter comes hand-in-hand with comprehension of input’ (Doughty 1991: 431). ‘A significant main effect was found for the attention condition across all of the data, indicating that subjects whose attention was drawn to the grammatical items gained more linguistic information about the items than the subjects whose attention was not called to the items. Also, the type of information called to the input was shown not to be a significant factor’ (Shook 1994: 79). ‘Visual enhancement seemed to have a facilitating effect on the learners’ recall and use of the targets, especially those of locative suffixes’ (Alanen 1995: 259). ‘Textual enhancement promotes noticing of target L2 form and has an effect on learners’ subsequent output’ (Jourdenais et al. 1995: 208). ‘Only participants in the Focus on Form group demonstrated significantly improved accuracy and suppliance of the target forms, thus suggesting that it is

614

6

7

8

9

10

11

12

13

14

15

16


possible to increase accuracy within a content-based instructional setting’ (Leeman et al. 1995: 217). ‘Results revealed a significant main effect for text length on readers’ comprehension but not on intake, no significant main effect for input enhancement on either comprehension or intake, and no significant interaction’ (Leow 1997: 151). ‘The knowledge acquired during incidental and enhanced learning is also partially memory-based, but the enhanced learners in particular show evidence of the development of a generalizable rule-based representation. Access to this is slow and effortful, in contrast to decisions based on memory for prior experiences’ (Robinson 1997: 242). ‘Even though learners in all three treatment groups improved in their ability to use his and her in an oral communication task, the findings did not support the hypotheses of this study. Although accuracy ratios overall followed the predicted order, that is, Group Eþ44Group E44Group U, the within-group variance canceled out most of the predicted between-group effects at the two posttests’ (J. White 1998: 101). ‘No positive effect was found for either enhancement or content familiarity on either the production or recognition tasks, but a negative effect was found on the subjects’ comprehension of the texts’ (Overstreet 1998: 228). ‘[D]rawing attention to the input can benefit processing, and that salience/ meaningfulness of the input is a major component in determining the extent of that benefit’ (Shook 1999: 39). ‘In general terms, these improvements would seem to demonstrate the effectiveness of devices that draw the subjects’ attention to specific aspects of the input through highlighting’ (Williams 1999: 29). ‘Results indicated no significant benefits of written input enhancement over unenhanced written input for (1) the amount of reported noticing of Spanish formal imperatives, (2) readers’ comprehension, or (3) readers’ intake’ (Leow 2001: 496). ‘Input enhancement failed to show measurable gains in learning despite the documented positive impact of enhancement on the noticing of the target form items in the input’ (Izumi 2002: 542). ‘The results indicated no significant benefit of textual enhancement over unenhanced input for (1) the amount of reported noticing of Spanish present perfect or present subjunctive forms, (2) learners’ intake of the forms, or (3) learners’ comprehension of the reading passage. The study did indicate a significant benefit of more salient forms (present perfect) over less salient forms (present subjunctive) for (1) the amount of reported noticing of the targeted verb forms, but not for (2) learners’ intake or (3) learners’ comprehension’ (Leow et al. 2003: 12). ‘[Textual enhancement] is not effective as a form of input enhancement on the acquisition of the French past participle agreement in relative clauses . . . Drawing learners’ attention to form via [textual enhancement] does not interfere with comprehension when whole clauses that contain the target form are underlined and when the form is a form of no communicative value’ (Wong 2003: 35). ‘[T]he main findings of the present study on discrete-item vocabulary learning were (a) no effect for enhancing 9 out of 24 words on learning rates for the enhanced words, (b) no effect for enhancing 9 out of 24 words on learning rates


17

18

19

20

21

615

for the unenhanced words, (c) a positive effect for enhancing 3 out of 24 words on learning rates for the enhanced words based on some but not all dependent measures, and (d) a negative effect for enhancing 3 out of 24 words on learning rates for the unenhanced words based on some but not all dependent measures’ (Barcroft 2003: 70). ‘No statistically significant benefit was found for enhanced input as compared to unenhanced input for (1) amount of reported noticing of targeted forms, (2) readers’ comprehension, or (3) readers’ intake’ (Bowles 2003: 406). ‘All of the 12 learners (6 dyads) in the [input enhancement] group reported that they had noticed some kind of visual enhancement in the reading given between Task 1 and Task 2’ (Park 2004: 14). ‘[Textual enhancement] alone did not aid form and meaning recognition of vocabulary, [and] [lexical elaboration] and [typographic enhancement] combined did not aid form recognition of vocabulary’ (Kim 2006: 341). ‘98% of the students in the Input [Enhancement] Group and 78% of the students in the Dictogloss Group experienced a positive change between the First and Last Essay. In contrast, only 38% of the students in the Conventional Group changed in a positive direction’ (Berent et al. 2007: 19). ‘Textual enhancement aided the learning of the target forms while having unfavorable effects on meaning comprehension’ (Lee 2007: 87)

NOTES 1 The 21 studies all had TE either as an independent variable or as a level thereof. 2 Of the 21 studies, Barcroft (2003) and Kim (2006) are the only two studies that focused on discrete-item vocabulary learning. There is arguably a distinction between vocabulary learning and grammar learning, which we will not deal with in this paper. 3 The cognitive processes may or may not be accessible to consciousness (Sharwood Smith, personal communication). 4 Barcroft (2003) provides evidence of a different kind of trade-off, ‘bidirectional effects’, namely that TE can draw attention to the enhanced items but away from the unenhanced items. 5 In order to account for the possible detracting effect of enhancement, J. White (1998) enhanced a different form than the target one in the reading texts for the unenhanced group.

6 The researcher explicitly told the participants to focus on the enhanced form. 7 What we have in mind here are the mainstream single-capacity models (see Robinson, 1997 for discussion). 8 The theoretical framework that underlies much of VanPatten’s research is known as ‘input processing’, which examines the form– meaning connection process in relation to intake. 9 See Greenslade et al.’s (1999) replication using aural and written input. 10 Wong (2005) sheds additional light on the lack of difference between the enhanced and the unenhanced group in this study, noting that both groups received explicit instructions to focus on verbal morphology, and that the effectiveness of the instructions may have overridden the effectiveness of the TE. 11 There is evidence from Barcroft (2003) that the fewer the enhanced

616


items, the more effective they are, suggesting that ‘distinctiveness’ can be a moderator variable to TE. 12 A related issue that is worth looking into concerns the nature of learning as a function of input enhancement. In discussing the research on input flood,

Wong (2005) insightfully notes that ‘input flood does not appear to be effective in showing learners what is not possible in the target language’ (p. 42) hence suggesting that the learning incurred by this particular strategy might be inadequate.

REFERENCES Alanen, R. 1995. ‘Input enhancement and rule presentation in second language acquisition’ in R. Schmidt (ed.): Attention and Awareness in Second Language Acquisition. Honolulu: University of Hawai’i Press, pp. 259–99. Allwright, D. 1984. ‘Why don’t learners learn what teachers teach? The Interaction Hypothesis’ in D. M. Singleton and D. Little (eds): Language Learning in Formal and Informal Contexts. IRAAL, pp. 3–18. Barcroft, J. 2003. ‘Distinctiveness and bidirectional effects in input enhancement for vocabulary learning,’ Applied Language Learning 13/2: 133–59. Bardovi-Harlig, K. and D. Reynolds. 1995. ‘The role of lexical aspect in the acquisition of tense and aspect,’ TESOL Quarterly 29/1: 109–31. Barry, S. and A. Lazarte. 1995. ‘Embedded clause effects on recall: Does high prior knowledge of content domain overcome syntactic complexity in students of Spanish?’ The Modern Language Journal 79/4: 491–504. Berent, G., R. Kelly, S. Aldersley, K. Schmitz, B. Khalsa, and J. Panara. 2007. ‘Focus-on-form instructional methods promote deaf college students’ improvement in English grammar,’ Journal of Deaf Studies and Deaf Education 12/1: 8–24. Bowles, M. 2003. ‘The effects of textual input enhancement on language learning: An online/offline study of fourth-semester Spanish students’ in P. Kempchinski and P. Pineros (eds): Theory, Practice, and Acquisition: Papers from the 6th Hispanic Linguistic Symposium and the 5th Conference on the Acquisition of Spanish & Portuguese. Somerville, MA: Cascadilla Press, pp. 359–411. DeKeyser, R. 1994. ‘How implicit can adult second language learning be?’ AILA Review 11: 83–96. Doughty, C. 1991. ‘Second language instruction does make a difference,’ Studies in Second Language Acquisition 13: 431–69.

Doughty, C. 1998. ‘Acquiring competence in a second language: Form and function’ in H. Byrnes (ed.): Learning Foreign and Second Languages. New York: Modern Language Association, pp. 128–56. Doughty, C. 2004. ‘Effects of instruction on learning a second language: A critique of instructed SLA research’ in B. VanPatten, J. Williams, S. Rott, and M. Overstreet (eds): Form-meaning Connections in Second Language Acquisition. Mahwah, NJ: Lawrence Erlbaum, pp. 181–202. Færch, C. and G. Kasper. 1986. ‘The role of comprehensible input in second-language acquisition,’ Applied Linguistics 7/3: 257–73. Gass, S. 1988. ‘Integrating research areas: A framework for second language studies,’ Applied Linguistics 9: 198–217. Greenslade, T., L. Bouden, and C. Sanz. 1999. ‘Attention to form and content in processing L2 reading texts: A conceptual replication study of VanPatten 1991,’ Spanish Applied Linguistics 3: 65–90. Han, Z.-H. 2007. ‘On the role of meaning in focus on form’ in Z.-H. Han (ed.): Understanding Second Language Process. Clevedon: Multilingual Matters, pp. 45–79. Han, Z.-H. and S. Peverly. 2007.’Input processing: A study of ab initio learners with multilingual backgrounds,’ The International Journal of Multilingualism 4/1: 17–37. Hulstijn, J. 1997. ‘Second language acquisition in the laboratory,’ Studies in Second Language Acquisition 19/2: 131–43. Izumi, S. 2002. ‘Output, input enhancement, and the noticing hypothesis: An experimental study on ESL relativization,’ Studies in Second Language Acquisition 24/4: 541–77. Jourdenais, R., M. Ota, S. Stauffer, B. Boyson, and C. Doughty. 1995. ‘Does textual enhancement promote noticing? A think-aloud protocol analysis’ in R. Schmidt (ed.): Attention and


Awareness in Foreign Language Learning. Honolulu: University of Hawai’i Press, pp. 182–209. Kim, Y. 2006. ‘Effects of input elaboration on vocabulary acquisition through reading by Korean learners of English as a foreign language,’ TESOL Quarterly 40/2: 341–73. Lee, S. 2007. ‘Effects of textual enhancement and topic familiarity on Korean EFL students’ reading comprehension and learning of passive form,’ Language Learning 57/1: 87–118. Leeman, J., I. Arteagoitia, B. Fridman, and C. Doughty. 1995. ‘Integrating attention to form with meaning: Focus on from in contentbased Spanish instruction’ in R. Schmidt (ed.): Attention and Awareness in Foreign Language Learning (Technical Report #9). Honolulu: University of Hawai’i Press, pp. 217–58. Leeser, M. 2004. ‘The effects of topic familiarity, mode, and pausing on second language learners’ comprehension and focus on form,’ Studies in Second Language Acquisition 26/4: 587–616. Leow, R. 1997. ‘The effects of input enhancement and text length on adult L2 readers’ comprehension and intake in second language acquisition,’ Applied Language Learning 82: 151–82. Leow, R. 2001. ‘Do learners notice enhanced forms while interacting with the L2?: An online and offline study of the role of written input enhancement in L2 reading,’ Hispania 84: 496–509. Leow, R., T. Egi, A. Nuevo, and Y. Tsai. 2003. ‘The roles of textual enhancement and type of linguistic item in adult L2 learners’ comprehension and intake,’ Applied Language Learning 13/2: 1–16. Lightbown, P. 1991. ‘What have we here? Some observations of the influence of instruction on L2 learning’ in R. Phillopson, E. Kellerman, L. Selinker, M. Sharwood Smith, and M. Swain (eds): Foreign Language Pedagogy Research: A Commemorative Volume for Claus Faerch. Clevedon: Multilingual Matters, pp. 197–212. Long, M. 1991. ‘Focus on form: A design feature in language teaching methodology’ in K. de Bot, D. Coste, C. Kramsch, and R. Ginsberg (eds): Foreign Language Research in a Cross-cultural Perspective. Amsterdam: John Benjamins, pp. 39–52. Long, M. 2007. Problems in SLA. Mahwah, NJ: Lawrence Erlbaum, pp. 75–118. Long, M. and P. Robinson. 1998. ‘Focus on form: Theory, research, and practice’ in C. Doughty and J. Williams (eds): Focus on Form in Classroom

617

Second Language Acquisition. Cambridge, UK: Cambridge University Press, pp. 16–41. Lorch, R. 1989. ‘Text-signaling devices and their effects on reading and memory processes,’ Educational Psychology Review 1: 209–34. Mackey, A. and J. Goo. 2007. ‘Interaction research in SLA: A meta-analysis and research synthesis’ in A. Mackey (ed.): Conversational Interaction in Second Language Acquisition: A Series of Empirical Studies. Oxford: Oxford University Press, pp. 407–52. Ortega, L. and G. Iberri-Shea. 2005. ‘Longitudinal research in second language acquisition: Recent trends and future directions,’ Annual Review of Applied Linguistics 25: 26–45. Overstreet, M. 1998. ‘Text enhancement and content familiarity: The focus of learner attention,’ Spanish Applied Linguistics 2: 229–58. Park, E. S. 2004. ‘Constraints of implicit focus on form: Insights from a study of input enhancement,’ Teachers College, Columbia University Working papers in TESOL & Applied Linguistics 42: Retrieved 7 September 2005, from http:// www.tc.edu/tesolalwebjournal/Park.pdf Robinson, P. 1996. ‘Learning simple and complex second language rules under implicit, incidental, rule-search, and instructed conditions,’ Studies in Second Language Acquisition 18: 27–67. Robinson, P. 1997. ‘Generalizability and automaticity of second language learning under implicit, incidental, enhanced, and instructed conditions,’ Studies in Second Language Acquisition 19: 223–47. Robinson, P. 2003. ‘Attention and memory during SLA’ in C. Doughty and M. Long (eds): The Handbook of Second Language Acquisition. Oxford: Blackwell Publishers, pp. 630–78. Schmidt, R. 1990. ‘The role of consciousness in second language learning,’ Applied Linguistics 11: 129–58. Schmidt, R. 2001. ‘Attention’ in P. Robinson (ed.): Cognition and Second Language Instruction. Cambridge: Cambridge University Press, pp. 3–32. Selinker, L. 1972. ‘Interlanguage,’ International Review of Applied Linguistics 10/2: 209–31. Sharwood Smith, M. 1981. ‘Consciousnessraising and second language acquisition theory,’ Applied Linguistics 7/3: 239–56. Sharwood Smith, M. 1986. ‘Comprehension versus acquisition: Two ways of processing input,’ Applied Linguistics 7/3: 239–56.

618


Sharwood Smith, M. 1991. ‘Speaking to many minds: On the relevance of different types of language information for the L2 learner,’ Second Language Research 72: 118–32. Sharwood Smith, M. 1993. ‘Input enhancement in instructed SLA,’ Studies in Second Language Acquisition 15: 165–79. Shook, D. 1994. ‘FL/L2 reading, grammatical information, and the input-to-intake phenomenon,’ Applied Language Learning 52: 57–93. Shook, D. 1999. ‘What foreign language reading recalls reveal about the input-to-intake phenomenon,’ Applied Language Learning 10/1&2: 39–76. Skehan, P. 1996. ‘A framework for the implementation of task-based instruction,’ Applied Linguistics 17: 38–62. Truscott, J. 1998. ‘Noticing in second language acquisition: A critical review,’ Second Language Research 14/2: 103–35. VanPatten, B. 1990. ‘Attending to form and content in the input,’ Studies in Second Language Acquisition 12: 287–310. VanPatten, B. 1991. ‘Grammar instruction and input processing’. Paper presented at the The Special Colloquium on the Role of Grammar Instruction in Communicative Language Teaching, Concordia University and McGill University, Montreal. VanPatten, B. 1996. Input Processing and Grammar Instruction: Theory and Research. Norwood, NJ: Ablex.

VanPatten, B. 2004. Processing Instruction. Mahwah, NJ: Lawrence Erlbaum. White, J. 1998. ‘Getting the learner’s attention’ in C. Doughty and J. Williams (eds): Focus on Form in Classroom Second Language Acquisition. Cambridge: Cambridge University Press, pp. 85–113. White, L. 1991. ‘Adverb placement in second language acquisition: Some effects of positive and negative evidence in the classroom,’ Second Language Research 72: 133–61. Williams, J. 1999. ‘Memory, attention, and inductive learning,’ Studies in Second Language Acquisition 211: 1–48. Williams, J. and J. Evans. 1998. ‘What kind of focus and on which forms?’ in C. Doughty and J. Williams (eds): Focus on Form in Classroom Second Language Acquisition. Cambridge: Cambridge University Press, pp. 139–55. Wong, W. 2001. ‘Modality and attention to meaning and form in the input,’ Studies in Second Language Acquisition 23/3: 345–68. Wong, W. 2003. ‘The effects of textual enhancement and simplified input on L2 comprehension and acquisition of non-meaningful grammatical form,’ Applied Language Learning 14: 109–32. Wong, W. 2005. Input Enhancement: From Theory and Research to the Classroom. San Francisco, CA: McGraw-Hill (Second Language Professional Series).

Applied Linguistics 29/4: 619–644 ß Oxford University Press 2008 doi:10.1093/applin/amn011 Advance Access published on 29 April 2008

The Cultural Productions of the ESL Student at Tradewinds High: Contingency, Multidirectionality, and Identity in L2 Socialization STEVEN TALMY University of British Columbia Although the originators of the language socialization (LS) paradigm were careful to cast socialization as a contingent, contested, ‘bidirectional’ process, the focus in much first language LS research on ‘successful’ socialization among children and caregivers may have obscured these themes. Despite this, I suggest the call for a more ‘dynamic model’ of LS (Bayley and Schecter 2003), while compelling, is unnecessary: contingency and multidirectionality are inherent in LS given its orientation to socialization as an interactionally-mediated process. This paper foregrounds the ‘dynamism’ of LS by examining processes comprising ‘unsuccessful’ or ‘unexpected’ socialization. Specifically, it analyses interactions involving ‘oldtimer’ ‘Local ESL’ students and their first-year teachers at a multilingual public high school in Hawai’i. Contingency and multidirectionality are explicated through analysis of two competing ‘cultural productions of the ESL student.’ The first, manifest in ESL program structures and instruction, was school-sanctioned or ‘official.’ Socialization of Local ESL students into this schooled identity was anything but predictable, however, as they consistently subverted the actions, stances, and activities that constituted it. In doing so, these students produced another, oppositional ESL student identity, which came to affect ‘official’ classroom processes in significant ways.

INTRODUCTION The socialization of children or novices by adults or experts into particular roles, identities, and world views has been the topic of scholarship for decades across the social sciences. Concerning as it does ‘the activity that confronts and lends structure to the entry of nonmembers into an already existing world’ (Wentworth 1980: 85), the nexus of socialization research engages such longstanding problematics as ‘agent vs. structure’, ‘voluntarism vs. determinism’, and ‘macro vs. micro’. Depending on disciplinary origin and theoretical orientation, emphases have varied in the diverse socialization literature on the influence that, for example, society has on the individual, or genetics has over the environment. Notwithstanding differences in emphasis, however, early theories of socialization—from the psychoanalytic tradition of Freud (1939), to the sociological functionalisms of Durkheim (1997) and Parsons (1937),

620

THE CULTURAL PRODUCTIONS OF THE ESL STUDENT AT TRADEWINDS HIGH

to the cultural anthropology of Mead (1961)—shared a general concern with the outcomes of socialization, characteristic of what Wentworth (1980) calls a ‘socialization-as-internalization model.’ In contrast, theories within a ‘socialization-as-interaction model’ have focused on the processes of socialization, including symbolic interactionism (Blumer 1969), phenomenology (Schutz 1967), ethnomethodology (Garfinkel 1967), and structuration theory (Giddens 1979). Despite substantive differences, these latter theories all consider ‘the medium through which the ability to produce society is transmitted [sic] from member to novice’ (Wentworth 1980: 79), that is, ‘the interaction that constitutes socialization’ (1980: 83). In doing so, they work to transcend binarisms such as ‘individual vs. society’, the inevitability and unidirectionality of a totalizing socialization, and a conception of the individual so ‘malleab[le] and passiv[e] . . . in the face of all powerful social influences’ (White 1977: 5) that s/he could be considered ‘oversocialized’ (Wrong 1961). It is within this ‘process’ orientation that language socialization (LS), the theory of socialization most often adopted in applied linguistics, can be situated. LS originated in the 1980s (e.g. Schieffelin and Ochs 1986a, 1986b) as a response both to earlier anthropological studies of child socialization, which ‘proceeded as if language was irrelevant’, and to the ‘invisibility’, in psycholinguistic studies of first language (L1) acquisition, of culture ‘as a principle that organized speech practices and their acquisition’ (Kulick and Schieffelin 2004: 349; also see Ochs and Schieffelin 2008). Consistent with its orientation toward socialization as a process that is accommodated, contested, and transformed by agents in interaction (see, e.g. Schieffelin and Ochs 1986a, 1986b; Garrett and Baquedano-Lo´pez 2002), LS provides a rigorous analytic framework for examining socializing processes in situ. This allows LS analyses to be grounded in ways unavailable to other models of socialization. It also offers the means to demonstrate the fundamental contingency and multidirectionality of socialization as it is—or is not—collaboratively achieved. Despite these analytic means, there has been some recent debate, particularly among second language (L2) LS scholars, concerning the extent to which unpredictability and mutuality in socialization have been accounted for in LS research. Bayley and Schecter (2003), for example, have strongly argued for ‘a more dynamic model’ of LS (2003: 6), one that ‘emphasize[s LS] as an interactive process, in which those being socialized also act as agents rather than as mere passive initiates’ (2003: 3). Such a call is an important corrective, they argue, to the view in ‘traditional’ LS studies of socialization as ‘relatively static, bounded, and relatively unidirectional’ (Schecter and Bayley 2004: 605). Yet such assessments are at odds with early claims made by Schieffelin and Ochs (1986a: 165) that ‘socialization is an interactive process’ and that ‘the child or novice . . . is not a passive recipient of sociocultural knowledge’ who ‘automatically internalize[s] others’ views’ but is a ‘selective and active participant . . . in the process of constructing social worlds’ (cf. Duff and Hornberger 2008). I argue that the call for a more ‘dynamic model’ of LS, while compelling and important, is in fact unwarranted given the basic premise in LS of socialization

STEVEN TALMY

621

as an interactionally-mediated process. That notwithstanding, however, it certainly is incumbent on LS researchers to highlight the ‘dynamism’ of LS in their analyses, lest the image emerge that socialization, and cultural and social reproduction, are ‘inevitable’ (Kulick and Schieffelin 2004; cf. Willis 1977). In the study below, I join a growing number of LS researchers (e.g. Duff 1995, 2002; Siegal 1996; Cole and Zuengler 2003; He 2003; Schecter and Bayley 2004) whose work foregrounds contingency and multidirectionality in LS by examining processes comprising what might be considered ‘unsuccessful’ or ‘unexpected’ LS. Specifically, I undertake an analysis of socializing interactions involving ‘oldtimer’ (Lave and Wenger 1991) ‘Local English-asa-second-language (ESL)’ students and their ‘newcomer’ teachers at Tradewinds High,1 a multilingual public high school in Hawai’i. My argument concerning contingency and multidirectionality takes shape around two competing ‘cultural productions of the ESL student’ (cf. Levinson et al. 1996) at Tradewinds. The first, manifest in ESL program structures and instructional practice, I gloss as school-sanctioned or ‘official’. Socialization of oldtimer ‘Local ESL’ students into this particular schooled identity was anything but certain, however, as they consistently withdrew participation in and otherwise subverted the acts, stances, and activities that constituted it. In doing so, these students produced another, oppositional ESL student identity which came to affect ‘official’ classroom processes in significant ways, contributing, perhaps paradoxically, to the social reproduction (Willis 1977, 1981; Giddens 1979) of the marginality and ‘stigma’ of ESL (also see Talmy forthcoming-a, forthcoming-b). In the following section, I consider the debate concerning contingency and multidirectionality in LS. Afterwards, I introduce the 2.5-year critical ethnography the data below is drawn from. I then analyse four classroom interactions involving Local ESL students and their newcomer teachers using a critical pragmatic (Talmy 2007) microanalytic framework that combines LS, interactional sociolinguistics (Gumperz 1982), membership categorization analysis, and ‘applied’ (ten Have 2001) conversation analysis (Sacks 1992). I use the first two interactions to illustrate contingency in LS by focusing on the competing ‘cultural productions of the ESL student’ mentioned above. My analysis of the latter two interactions focuses more directly on multidirectionality in LS, specifically in terms of how oldtimer students’ actions were increasingly accommodated by newcomer teachers over the course of a school year, suggesting their socialization into an ESL teacher identity that was at once adversarial and infantilizing. I conclude by considering implications of the analysis for LS, agency, and cultural reproduction.

CONTINGENCY AND MULTIDIRECTIONALITY IN LS In earlier ‘first generation’ (Garrett and Baquedano-Lo´pez 2002) LS studies, the primary foci were on the ways that caregivers socialized children in and through the L1 to become ‘competent’ members of sociocultural groups.2

622


The focus on ‘successful’ cases of children’s L1 socialization, on recurrent language practices, universals, and monolingual communities, has, according to some, obscured the basic orientation in LS of socialization as contingent and multidirectional. In terms that evoke criticisms made of earlier functionalist theories of socialization, He (2003) maintains that LS research tends to emphasize the efforts made by the experts . . . to socialize the novices. . . . Less visible are the reactions and responses of the novices. Consequently, the process of socialization is often characterized as smooth and seamless, and novices are often presumed to be passive, ready, and uniform recipients of socialization. (128) Garrett (2004), for one, disputes He’s assertion, arguing that this ‘is simply untrue: Even the earliest [LS] studies stressed the agency (and the capacity for resistance and creativity) of the child or novice, and conceptualized socialization as a two-way process’ (778). This is indeed the case (see above); however, it should be noted that there is a difference between ‘stressing’ these themes, and demonstrating them empirically. In an apparent nod to the contention that LS has downplayed the ‘dynamism’ of LS by focusing on communicative processes involved in socialization that is realized rather than that which is not, Ochs (1999) herself has noted certain ‘undesirable effects’ deriving from the foci of earlier LS studies, including fixity, essentialism, and reductionism: We say . . . that Samoan caregivers communicate one way, EuroAmerican caregivers another way, Kaluli yet another, and so on. Our accounts . . . seem like fixed cameos, members and communities enslaved by convention and frozen in time rather than fluid and changing over the course of a generation, a life, and even a single social encounter. (231) Kulick and Schieffelin (2004) address the matter more directly: That the majority of [LS] studies have focused on [cultural and social] reproduction is a strength—they provide us with methodological and analytical tools for investigating and interpreting . . . continuity across generations. But the focus on expected and predictable outcomes is a weakness if there is not also an examination of cases in which socialization doesn’t occur, or where it occurs in ways that are not expected or desired. To the extent that [LS] studies only document the acquisition of normatively sanctioned practices, they open themselves up to the charge that they are merely behaviorism in new clothes. (355) The issue thus appears to come down to at least some combination of empirical focus in LS research (e.g. highlighting processes involved when L1 socialization is achieved) and historical moment (i.e. establishing and

STEVEN TALMY

623

elaborating the LS paradigm) (cf. Ochs and Schieffelin 2008), rather than some problem endemic to LS itself. For this reason, the focus of this paper can be considered part of an historically-situated effort to ‘push [LS’s] boundaries . . . further,’ as Garrett puts it, by examining ‘bilingual and multilingual settings . . . later stages of the lifespan . . . [and] contexts associated with those later stages, such as peer groups, schools, and workplaces’ (2004: 776; also see Duff and Hornberger 2008). Further discussion of this point is available online as Supplementary Material 1 for Subscribers at http:// applij.oxfordjournals.org

THE STUDY The data utilized here come from a larger study consisting of 625 hours of observation in 15 classrooms, including eight dedicated-ESL classes, over 2.5 years. Observational data were generated in field notes and supplemented by audio-recordings of 158 hours of classroom interaction. A total of 58 formal interviews were audio-recorded with 10 teachers and 37 students, and materials, classwork, and other site artifacts were collected for analysis.

First-year teachers in first-year ‘ESL-A’ classes This paper concerns the three first-year ‘ESL-A’ classes that I observed, each taught by an instructor in his/her first year at the high school.3 I observed Ms Cheney’s ESL-A (1X),4 Ms Ariel’s ESL-A (2W), and Mr Day’s ESL-A (2X) for 48, 68, and 64 hours respectively, with 26 hours of audio-recording in the ESL-A (2W) class, and 29 hours in ESL-A (2X).5 The ESL-A classes were the largest, most heterogeneous, and instructionally challenging classes in the Tradewinds ESL program. One reason for this was a policy that determined ESL placements based not on L2 proficiency but length of enrollment at the school. Thus, the ESL-A classes were for students in their first year at Tradewinds, regardless of age, grade-level, or L2 proficiency. The three classes I observed averaged over 30 students, aged 14–18, about one-third of whom were at early levels of L2 development and/or had interrupted formal educations; another third of whom had lived in the US for 3–10 years, many of whom I identified as Local ESL (see Table 1); with the remainder at levels of English expertise in between.

‘Local’ and ‘Local ESL’ ‘Local’ is an identity category in wide circulation in Hawai’i, signifying a racialized ‘sociopolitically constructed panethnic formation comprised of Asian and/or Pacific Islanders who were born and raised in the islands’ (Labrador 2004: 292).6 It is a relational category, defined as it is through a system of ‘categorical opposition[s] between groups considered [L]ocal and

624


Table 1: Local ESL students in ESL-A (2W) and ESL-A (2X) (age and length of residence at mid-point of school year; self-report data) Student pseudonym

Sex Age L1

Country of origin

US length of residence

Ms. Ariel’s ESL-A (2W) 618 F Ash M Barehand M China M Camp Kill Yourself (CKY) M Eddie M Maria F Benz M Nat M Raven M

14 14 16 15 14 15 14 15 14 15

Cantonese Cantonese Korean Cantonese Vietnamese Ilokano Tagalog Korean Marshallese Mandarin

China Hong Kong Korea Hong Kong Vietnam Philippines Philippines Korea Marshall Islands Taiwan

4.5 years 7.5 years 3.5 years 4.5 years 9.5 years 4 years 4.5 years 2.5 years 8 years 8.5 years

Mr. Day’s ESL-A (2X) CJ Computer Ioane Iwannafuckalot (IwannaFAL) Jennie

M M F M

17 17 16 14

Ilokano Korean Tongan, English Vietnamese

Philippines Korea US (Hawai’i) Vietnam

7 years 5.5 years n/a 9.5 years

F

14

Korean

US (California)

Laidplayer

M

15

Palauan

Palau

Mack Daddy Nina Joyleen

M F F

15 15 16

Chuukese Chuukese Chuukese

Chuuk, FSM Chuuk, FSM Chuuk, FSM

Shorty

F

16

Sky Blue

M

15

Samoan, Tongan, US (Hawai’i) English Cantonese Macau

2.5 years (returnee) 5 months (see below) 5.5 years 4.5 years 6 months (6 years in Guam) n/a 3 years

those considered non[L]ocal, including haole [i.e. ‘white people’], immigrants, the military, tourists, and foreign investors’ (Okamura 1994: 165). The production of ‘Local’ identity centrally involves participation in a range of social practices, including the use of Pidgin (Hawai’i Creole), the creole language of Hawai’i (see Sakoda and Siegel 2003).

STEVEN TALMY

625

‘Local ESL’, an analyst’s category, is conceptualized as a situationally contingent identity (Zimmerman 1998). It refers to students who were institutionally-identified as ESL by Tradewinds, yet who displayed the following: cultural knowledge of and affiliation with Local culture, cultural forms, and social practices; experience with US and Hawai’i school expectations and practices; difference from newcomer or lower-L2-proficient classmates, who many Local ESL students characterized as the ‘real’ ESL students; awareness of—and participation in—‘linguicist’ (Phillipson 1988) discourses about ESL in circulation in the wider social context; and, crucially, the L2 expertise and interactional competence to participate in these practices.7

School-sanctioned productions of ESL What I call the school-sanctioned productions of ESL can be explicated in terms of instructional practice and the ‘structural productions’ (Eisenhart 1996) of ESL. Drawing on Nespor (1990), Eisenhart (1996) uses the concept of ‘structural productions’ to analyse how the subject position ‘student’ is produced within and by particular institutional arrangements, that is to say, how program organization, intra-institutional relationships, and curriculum constitute certain preferred ways of being a student in those contexts. Such an idea is especially salient for programs such as ESL, which tightly control student enrollment and exit, promoting access to certain forms of learning and school experience, and denying it to others. Such structures have significant ‘implications for how students spen[d] their time (e.g. with whom, doing what) and what they learn’ (Eisenhart 1996: 172), and can result in the ‘produc[tion] of cohorts . . . with shared outlooks, ambitions, definitions of reality, and strategies for acquiring and using knowledge’ (Nespor 1990: 231–2). Analysing structures of the Tradewinds ESL program can therefore offer insight into how the category ‘ESL student’ was institutionally conceived. ESL at Tradewinds was structurally articulated as a hypernym, suggesting that students classified as ‘ESL’ were essentially the same, with negligible differences in L1 or L2 abilities, educational backgrounds, or needs. This was exemplified by the placement policy described above, whereby length of enrollment rather than L2 expertise determined what ESL classes students were to take. It was also signified by an undifferentiated curriculum: although the ESL-A classes were large and diverse, students were, with few exceptions, given the same assignments, activities, and tests, each with the same demands and deadlines. In addition to homogeneity, ESL at Tradewinds was structurally elaborated as an afterthought to pre-existing institutional arrangements. This was indicated by the status of ESL classes as ‘required electives’; by an absence of communication between ESL and subject-area departments; and a lack of administrative, curricular, and physical integration with the rest of the school. These arrangements were amplified by the assignment of ESL courses to new or junior faculty, most of whom had little background in ESL,

626


and who were provided minimal institutional support despite frequently overwhelming instructional circumstances. The ESL curriculum reflected, and helped to re/produce, the program’s institutional standing. This was most evident in the choice of L1-English juvenile fiction to form the basis of the ESL curriculum. Books such as James and the Giant Peach (Dahl 1961) and Dogsong (Paulsen 1985) were below grade-level and often had tangential relevance to high school academics or L2 learning. In addition to these books were assignments that evidently presumed students knew about and ‘automatically affiliated’ (Talmy 2007) with the cultures, customs, and languages of ‘their’ countries (Duff 2002; Talmy forthcoming-a).8 Assignments introducing initiates to cultures and customs of the US often appeared, as well, especially around holidays, as did other ESL mainstays such family tree activities and map exercises, which most Local ESL students maintained they had been assigned in earlier grades, often repeatedly. Thus, ‘ESL student’ as it was institutionally articulated at Tradewinds connoted a monolithic out-group of recently-arrived cultural and linguistic ‘Others’, that is, an iconic, stereotypical ‘ESL Student’, or ‘FOB’ (‘fresh off the boat’) (Talmy 2007, forthcoming-a, forthcoming-b). Local ESL students’ responses to these school-sanctioned productions of the ESL student, were, as might be expected, largely negative. Indeed, these responses became so pervasive in the ESL-A classes I observed that they ultimately affected what was taught, how it was taught, and who taught it, with the knowledge, orientations, and practices of an ‘institutionally disapproved interstitial community of practice’ (CoP) (Lave 1991: 78) coming to assume varying degrees of prominence in each of these classes.9 The ‘joint enterprise’ of this CoP, produced as Local ESL students ‘mutually engaged’ in a ‘shared repertoire’ of social practices (Wenger 1998), was the contestation, if not destabilization, of the school-sanctioned productions of ESL. Implicated in these efforts were Local ESL students’ prior experiences in ESL classes, as well as their connections to communities beyond the ESL program, where ESL was stigmatized, and English monolingualism, the ‘mainstream’, and ‘regular students’ were valorized as preferred ideals.

Competing productions of the ESL student The interactions I discuss next occurred between the third and sixth months of the school year. By this time, curriculum and classroom processes had begun to appreciably change in all three ESL-A classes as teachers contended with Local ESL students’ participation in practices, both subtle and overt, that worked to undermine ‘official’ curriculum and instruction. These included leaving assigned materials ‘at home’; not doing homework; completing assignments that required minimal effort (e.g. worksheets) but not others (e.g. writing activities); starting class late; and finishing early. The more overt practices included bargaining for reduced requirements on classwork and

STEVEN TALMY

627

extended time to complete it; refusal to participate in instructional activities; teasing students who did participate; and the often delicate negotiations with teachers that resulted (Talmy forthcoming-b). These in part constituted the shared repertoire of the Local ESL CoP described above, and the oppositional productions of the ESL student. The first extract I analyse involves Laidplayer, a 9th-grader from Palau. When I first met Laidplayer, I was surprised to learn he had only been in Hawai’i for a few weeks, due to his advanced expertise in L2 English and Pidgin, and his knowledge of Local culture. As it happened, Laidplayer had extended family in Hawai’i with whom he lived as he went to school, and who provided him with the resources to act and sound—to be—Local. Though he claimed to have had no idea what ‘ESL’ meant when he first came to Tradewinds, by the end of the year he was participating in a host of practices that were indexical of an ESL oldtimer. This interaction occurred five months into the school year, near the end of a ‘study hall’ class session, in which students were, in Mr Day’s words, to be ‘working behind or ahead’, that is, to be catching up on overdue work. Laidplayer had to this point spent 45 minutes on a short exercise from an audiolingual-era grammar book (Dixon 1956), then had talked with two Local ESL classmates, before finally falling asleep. Mr Day had woken him twice already to get him to finish an overdue summary from Shiloh (Naylor 1991). As the interaction begins, Laidplayer is fixing his watch and talking to IwannaFAL seated beside him (see Appendix for transcript conventions). Extract 1. So izi 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Mr. Day:

[ELA41XmdS10: 1185–1208]

((calls across room)) Laidplayer, you finish your summary. (0.4) IwannaFAL?: (no.) (0.5) ↑finish your summary it’s due today! do Mr. Day: you need help? (0.7) Mr. Day: I ↓kno:w, you want me to sit down and read the book with you. alri::ght. (0.7) ((Mr. Day approaches)) Mr. Day: >what’s wrong?< Laidplayer: I was: fixing my watch. (0.9) ((Mr. Day sits down)) Mr. Day: what’s wrong with your watch? Laidplayer: it’s too big. (2.9) IwannaFAL: oh ma:n, the bell’s gonna ring! (0.6) ?FS: ↑wha::t? (0.7) Mr. Day: u:h no you got about ten minu[tes before] we’re outta here.

628


24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

Laidplayer:

Mr. Day: Laidplayer:

Mr. Day: Mr. Day: Laidplayer: Mr. Day: Laidplayer:

Mr. Day:

Laidplayer: Mr. Day: Laidplayer: Mr. Day: Laidplayer: Laidplayer: IwannaFAL:

[(sh:it.) ] (0.7) ((clicking sound of metal watch band)) why don’t you try get it done that way you won’t have to do it [at home. [SO IZI ai jas– IT’S SO EASY I just– (0.5) >so do it now!< (0.5) °>so do it now.ai dono wai >go in the bathroom.< hhh huh? hey, I like to do that too. that’s the best time. just don’t get it dirty. HHHHH (0.4) hhhHHHHH. (1.0) ((Mr. Day gets up and leaves)) ho wipe my ass with (.) the book. ((clears throat)) oh yeah oh yeah.

In his first four turns, Mr Day enumerates or implies a series of acts and activities that constitute particular ESL student and teacher identities. Associated with the teacher identity are the rights, obligations, and competencies that warrant assigning ‘schoolwork’ to students, setting deadlines to complete it, monitoring classroom performance, making assessments about it, and if deemed problematic, pursuing some form of remedy, for example, in the form of ‘help’. The student is someone who has already completed the assigned reading of Shiloh so s/he can produce a summary of it, who is endeavoring to meet the due date of ‘today’, and is complying with Mr Day’s instructions in the larger activity of doing ESL coursework. This is, in other words, a student who is accommodating the production of a schoolsanctioned ESL student identity. It is clear that Laidplayer does not participate in the acts, stances, or activities associated with the production of this student identity: this has, in fact, occasioned Mr Day’s visit to him. Despite repeated ‘reminders’ to complete his overdue summary, during a ‘study hall’ class specifically intended for this purpose, Laidplayer has spent the period doing a 22-item grammar worksheet, talking with friends, sleeping, and is now fixing his watch. This is accountable

STEVEN TALMY

629

behavior and in his first turns, Mr Day displays an orientation to its problematicity. In lines 1–2, he calls Laidplayer by name, asking if he’s completed the summary as instructed. Although this may be a ‘real’ yes/no question, it is likely given the context that its polarity is reversed (Koshik 2002). Reversed polarity yes/no questions (RPQs), similar to rhetorical questions, are not information-seeking, but convey a ‘negative assertion’ (1852) about some form of social action and can thus ‘be heard as a complaint’ (1863). Whether this is an RPQ or not, the imperative, higher pitch, emphatic stress, and exclamatory intonation of Mr Day’s next utterance (lines 6–7) index a stance of unambiguous disapproval. Despite this, Laidplayer does not provide the sequentially-projected reply. In fact, it is not until more than 11 seconds after Mr Day calls his name that Laidplayer finally speaks. Even then, the account he provides, ‘I was: fixing my watch.’, is not what might be expected, since this provides denotative ‘proof’ that he has not been ‘finishing the summary’. Still, he offers no elaboration when one might, once more, be anticipated. In short, Laidplayer ‘does’ the same resistance interactionally that his utterance references denotationally; it is a mark of his L2 interactional competence that this utterance accomplishes both simultaneously. Interestingly, after a nearly 1-second delay, Mr Day accommodates rather than problematizes Laidplayer’s topicalization of his watch (‘what’s wrong with your watch?’). Although in line 16, Laidplayer (finally) provides an account (‘it’s too big.’), it is minimal. The silence that follows suggests that Mr Day rejects the account as inadequate (cf. Davidson 1984), and is awaiting further elaboration (cf. Pomerantz 1984). IwannaFAL’s sudden ‘oh man, the bell’s gonna ring!’ is thus particularly important, working as it does to reorient Mr Day from the pursuit of an elaboration to the matter of when class will end (lines 22–23). Coming as it does with more than 10 minutes left in the class (cf. the female student’s high-pitched ‘wha::t?’), this may be an instance of a Local ESL ‘oldtimer’ helping out a comparative apprentice during an interactional difficulty with his teacher. In lines 27–28, Mr Day returns to the matter of Laidplayer’s summary but in distinctly different terms. With the mitigated directive in question format (‘why don’t you try get it done’) and the added inducement (‘that way you won’t have to do it at home.’), Mr Day frames this utterance as a suggestion. When compared with Mr Day’s first two turns, it appears that he has oriented to Laidplayer’s disaffiliative actions and is accommodating them. With his line 29 utterance, ‘So IZI ai jas–’, Laidplayer at last provides the sought-after account about ‘what’s wrong’. However, the affective and epistemic stances indexed by the overlap, the amplified volume, the sentenceinitial placement of the intensifier ‘so’, the lack of hedging, and the sentence incompletion transform this utterance from an explanation, to a rejection that what is accountably ‘wrong’ is attributable to Laidplayer. It instead shifts the problem to the assigned material, that is, the award-winning children’s book Shiloh.

630


Significantly, Mr Day’s repeated ‘so do it now’ utterances in lines 31 and 33 do not challenge Laidplayer’s assessment, but implicitly concur with it. These can be seen as additional aligning moves, consistent with Mr Day’s earlier uptake regarding the watch, the mitigated directives in lines 27–28 and 40–42, and his final remark in lines 48–49. Indeed, the discourse marker ‘so’ that prefaces both line 31 and 33 turns marks a ‘fact-based result relation’ (Schiffrin 1987: 201–4), meaning the assessment is not only concurred with but used as an incentive for the student to ‘do it now!’ This transforms Laidplayer’s subsequent admissions in line 34 (bat ai neva rid da (h)as(hh)ai(h)nmen.) and lines 37–38 (WEL AI DON RID DIS BUK wen ai get hom, 4ai dono wai5.) into unaccountable actions. More significantly, it indexes broader pedagogical accommodations that have been implemented in this class to make it ‘easier’ for students to ‘finish’: the class time that has been allotted for students to complete overdue assignments, and the deadlines that have been extended. In other words, there is evidence here of this first-year teacher’s uptake to his students’ socializing practices. Laidplayer’s refusal to participate in the acts, stances, and activities that comprise the school-sanctioned ESL student identity, or in the interactional norms of a participation structure such as this (cf. Mehan 1979), are constitutive of the production of an alternative ESL student identity. As outlined in Table 2, this is a student whose disaffiliative actions index resistance to, difference from, and lack of investment (Norton 2000) in the ‘official’ productions of ESL.10 These points are further illustrated in Extract 2 from Ms Ariel’s ESL-A (2W). This interaction occurred early in the fourth quarter of the school year, by which time the Local ESL CoP in the class was also firmly established. It took place during a discussion activity for The Wanderer (Creech 2000), another award-winning book that by the publisher’s estimation was intended

Table 2: Acts and activities constituting productions of the ESL student at Tradewinds School-sanctioned ESL student

Local ESL student

Bringing required materials to class Reading assigned juvenile fiction Doing bookwork (summaries, questions, vocab) Meeting due dates Following instructions Working for full class session Not participating in proscribed activities (sleeping, talking with friends, playing cards, ‘off-task’ behavior)

Not bringing required materials to class Not reading assigned juvenile fiction Doing some bookwork (worksheets; little extended writing, e.g. summaries, questions)

Not meeting due dates Not following instructions Interactional resistance Participating in proscribed activities

STEVEN TALMY

631

for audiences aged 8–12. The reading group for this book was comprised almost entirely of Local ESL students: China, Ash, Raven, Barehand, Benz, 618, Sou Li (a low L2-proficient newcomer who had been paired with 618) and CKY, who was shuffling a deck of playing cards intermittently. For 7 minutes prior to the interaction, Ms Ariel had tried and failed to get a discussion started among these students about the following questions, which concerned Sophie, the 13-year-old protagonist of The Wanderer:

What is Sophie’s story? Why is she so attracted to the open ocean? Is it unusual for a 13-year-old to be the only female on the boat? What does this mean? (ELA41Wdoc: L). Extract 2. Cards 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

Ms. Ariel:

Ms. Ariel: China: Ms. Ariel: Ms. Ariel:

China: China: Ms. Ariel:

Ms. Ariel: Benz?: Ms. Ariel:

China: Ms. Ariel: Ms. Ariel:

China: Ms. Ariel: China: Ms. Ariel: Benz: China: Benz: Ms. Ariel: Benz: China:

[ELA42WmdS11: 565-593]

you have to think about Sophie’s personality. (2.6) ((Ss talking in other group)) ((China opens recorder case, examines it)) (1.3) ((Ss talking in other group)) you have to think about what you’ve read. ((blowing into mic?)) SSSS[S [China. (0.6) why is she so attracted to the ocean. that is going to have a lot to do with your own (.) thoughts.= =I THINK IT’S QUIET AND PEACEFUL, (2.4) ((Ss talking in other group)) an:d dolphin is real[ly nice. [how many people have started on the second set of questions. (0.6) (as is us[ual.) [(nah.) (1.8) okay. start on them today, (0.3) finish the first three (0.7) before you start the se[cond set. [((heavy sigh)) (2.0) ((Ss talking in other group)) okay, and they’re essay style questions. (1.0) ((to CKY)) and don’t bring your cards in future please. (0.7) YAE CK↓Y. YEAH CKY. I’ll throw them away. are you looking at me, Miss? okay I’m gonna go see what they’re doing,= =kay. kay.= =bye. up to you,= =bye. hh

632


42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

CKY: 618: China: Barehand: China:

Raven: China: Benz: China: Barehand: China: Benz: China: CKY:

hhh[h [(you’re cold.) ba’bye. >have a good day.

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

Recommend Documents