Bilingualism and Social Relations: Turkish Speakers in North West Europe

Bilingualism and Social Relations Turkish Speakers in North Western Europe Edited by

J. Normann Jørgensen

MULTILINGUAL MATTERS LTD Clevedon • Buffalo • Toronto • Sydney

Library of Congress Cataloging in Publication Data Bilingualism and Social Relations: Turkish Speakers in North Western Europe Edited by J. Normann Jørgensen. Includes bibliographical references. 1. Bilingualism–Europe, Western. 2. Turks–Europe, Western–Languages. 3. Youth–Europe, Western–Language. 4. Social interaction in youth. I. Jørgensen, J. Normann. P115.5.E85 B55 2003 404'.2'094–dc21 2002015686 British Library Cataloguing in Publication Data A catalogue entry for this book is available from the British Library. ISBN 1-85359-650-7 (hbk) Multilingual Matters Ltd UK: Frankfurt Lodge, Clevedon Hall, Victoria Road, Clevedon BS21 7HH. USA: UTP, 2250 Military Road, Tonawanda, NY 14150, USA. Canada: UTP, 5201 Dufferin Street, North York, Ontario M3H 5T8, Canada. Australia: Footprint Books, PO Box 418, Church Point, NSW 2103, Australia. Copyright © 2003 J. Normann Jørgensen and the authors of individual chapters. This book is also available as Vol. 24, Nos 1&2, 2003 of the Journal of Multilingual and Multicultural Development. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. Printed and bound in Great Britain by the Cromwell Press Ltd.

Contents J. Normann Jørgensen: Linguistic Construction and Negotiation of Social Relations Among Bilingual Turkish-speaking Adolescents in North-western Europe

1

Volker Hinnenkamp: Mixed Language Varieties of Migrant Adolescents and the Discourse of Hybridity

12

nci Dirim and Andreas Hieronymus: Cultural Orientation and Language use among Multilingual Youth Groups: ‘For me it is like we all speak one language’

42

Jakob Cromdal: The Creation and Administration of Social Relations in Bilingual Group Work

56

Trine Esdahl: Language Choice as a Power Resource in Bilingual Adolescents’ Conversations in the Danish Folkeskole

76

Lian Malai Madsen: Power Relationships, Interactional Dominance and Manipulation Strategies in Group Conversations of Turkish-Danish Children

90

Erica Huls, Ad Backus, Saskia Klomps and Jens Normann Jørgensen: Adolescents Involved in the Construction of Equality in Urban Multicultural Settings

102

J. Normann Jørgensen: Languaging Among Fifth Graders: Code-switching in Conversation 501 of the Køge Project

126

iii

Linguistic Construction and Negotiation of Social Relations Among Bilingual Turkish-speaking Adolescents in North-western Europe Introduction J. Normann Jørgensen Department of Nordic, University of Copenhagen, Denmark

In recent years sociolinguistics has been increasingly interested in the language users’ work with creation and maintenance of social relations by means of language. Classical sociolinguistics studied linguistic variation as an effect of social structures, but the perspective has now been reversed, and social structuring is viewed as an effect of variation in language use. This entails different methods and different types of data in modern sociolinguistics. Typically, classical sociolinguistics collected large numbers of tokens of well-defined and unique linguistic variables from a sample of a total population. Prominent examples are Labov (1966) and Trudgill (1974), who both study a limited number of pronunciation variables with comparatively large numbers of informants. Based on the linguistic variation within their sample they draw conclusions about the overall pattern of variation in the population as a whole, including the relationship between the speaker’s social status (in socioeconomic terms) and her or his pronunciation of specific phonetic variables. Milroy (1980) represents another perspective on data collection, and another view of the social structures, but the linguistic variation observed is still considered an effect of social structures, now understood as different network patterns. A vast body of sociolinguistic literature draws on the same basic understanding of the relationship between social structures and linguistic variation. Classical feminist studies of gender and language is one example (for instance, Tannen, 1991). Studies of bilingualism, and particularly bilingual behaviour, is another example. Herman (1961) ascribes the patterns of language choice to situational variation, particularly the social relations among interlocutors (but also the speaker’s personal involvement). Fishman’s (1965) adaptation of Ferguson’s (1959) diglossia concept takes as its basis the observation that everyday use of two languages is arbitrary only in its most superficial form. In most real bilingual groups, or ‘speech communities’, there is a quite stable relationship between on the one hand the circumstances under which language is used, i.e. situational factors, and on the other hand language choice. ‘Proper’ usage, or common usage, or both, dictate that only one of the theoretically co-available languages will be chosen by particular classes of interlocutors on particular occasions. (Fishman, 1965: 67, italics from the original) Gumperz’s (1982) distinction between situational and metaphorical code-switching was an important development of the sociolinguistic understanding of bilingual behaviour. The situational circumstances of a particular 1

2

Bilingualism and Social Relations

conversation may determine the linguistic behaviour of bilingual interlocutors who share more than one language: the setting, participants, time, subject of discussion, etc. If the situational circumstances change, another language may become the appropriate one to use, and a switch will take place. Such situational code-switching follows Fishman’s description of bilingual behaviour. Bilinguals, however, often share languages, one of which is a societal minority language (we-code), the other one a societal majority language (they-code). In their mutual conversations the bilingual language users may choose to invoke the values attached to either of the languages. By switching from minority to majority language, the bilingual speaker invokes its value as a they-code. The message of the utterance will be influenced by the switch – the meaning will include the force and power attached to the majority language. On the other hand, the bilingual may switch from the majority to minority language, and the message will contain a reference to the solidarity and closeness of the minority language. Such metaphorical code-switches are determined by the intention of the speaker to relate a specific understanding of the content. In the 1990s the classical sociolinguistic view gave way to the social constructivist concept of language use as identity work on behalf of the language user. For instance, Johnson and Meinhof (1997) describe how linguistic differences are used by the gender to create identities rather than as a result of their gender, ie. gender is performed: There is no such thing as a ‘men’s language’. This does not mean that the notion of ‘difference’ has no part to play in the study of language and gender. But it would undoubtedly be more appropriate […] to shift the emphasis from ‘gender difference’ to ‘the difference gender makes’ (Johnson & Meinhof, 1997: 25) Bilingual behaviour, particular code-switching, creates social meaning in the specific contexts in which it is used. Bilingual speakers not only invoke values ascribed to their languages by society at large, they also constantly negotiate and rearrange the values of their languages when they interact. Code choice, and code-switching become acts of identity (LePage & Tabouret-Keller, 1985), and the act of selecting a linguistic item, be it a word, a sound, or a phrase, from one or the other language or variety is in itself a statement about who the speaker is, and what relationship he or she is involved in or getting involved in, with the interlocutors. The linguistic items involved may even stretch far beyond the two languages usually thought to be involved in bilingual interaction. In multilingual settings, particularly among adolescents, there need not be limits on what languages the individuals draw on for intergroup communication. Even monolingual language users may use items from languages around them which they do not command, and which are not ‘theirs’. But by being surrounded by speakers of these (typically minority) languages or varieties every day, as among youth groups in multilingual settings, even the monolinguals will know the existence of these languages, and they will know about the values ascribed to these languages in the larger society. By introducing items from a range of languages and varieties, the speakers, mono-, bi- or multilingual, may join in creating a linguistic atmosphere of crossing (Rampton, 1995) in which the borderline

Linguistic Construction and Negotiation of Social Relations

3

between ‘my’ language and ‘your’ language becomes blurred, and in which the values ascribed to the languages are challenged and renegotiated: Crossing … focuses on code alternation by people who are not accepted members of the group associated with the second language they employ. It is concerned with switching into languages that are not generally thought to belong to you. This kind of switching, in which there is a distinct sense of movement across social or ethnic boundaries, raises issues of social legitimacy that participants need to negotiate, and that analysts could usefully devote more attention to. (Rampton, 1995: 280) The view of linguistic variation as an instrument in social negotiation can be found with Sebba (1993). In a critical discussion of traditional variationist sociolinguistics, accommodation theory, and the act-of-identity approach to linguistic variation, Sebba concludes that: I have suggested that conversational interactions, not just for speakers who obviously ‘straddle two cultures’ but even for those who have traditionally been called ‘monolingual’ and ‘monocultural’, involve the speaker in animating a series of personas which are realised linguistically and derive their symbolic value from their association with stereotypes which have reality and symbolic value for the interactants. Such personations may operate at several levels, so that the speaker may create a persona for another individual who is talked about, while at the same time animating several personas which represent him or her self. Linguistic representations of different ‘selves’ and ‘others’ are thus woven by a speaker into a complex skein of conversation (or narrative), with feedback from other interactants at all times potentially affecting the blend. (Sebba, 1993: 137) Rampton (1999) suggests several examples of analyses that demonstrate how speakers represent and negotiate social structures through specific choices of language forms. In that particular context, the speakers are seen to describe indirectly and evaluate others while placing them in particular roles, for example, styling them in narratives. Several of the contributions in Auer (1998) discuss the relationship between specific code-switches and social relations outside the given conversation. Li Wei argues that conversational analysis may indeed help us understand code-switches, but it is not enough. The knowledge of values, role expectations, norms, and other societal phenomena is necessary to fully understand the meaning of code-switches. Although he specifically states that his main purpose is ‘to make a case for the conversational-analytic approach to code-switching’ he also concludes that: Thus, the fact that a bilingual speaker has chosen to code-switch invites a more detailed, perhaps multi-layered analysis which can demonstrate that in addition to its capacity of highlighting the status of the ongoing talk, code-switching as a contextualisation cue has the capacity to ‘bring about’ higher-level social meanings such as the speakers’ language attitudes, preferences, and community norms and values. While the need to avoid the wider contexts overshadowing the participants’ procedures is apparent, it

4

Bilingualism and Social Relations

is equally important to prevent entanglement in over-detailed description of conversation structures without making any sensible inference. (Wei, 1998: 173) The view of linguistic variation as an instrument for social negotiation is also prevalent in studies of youth language. Eckert (2000) describes the type of linguistic variation typically studied by classical sociolinguistics, and she sees it as a result of the identity work of the involved (monolingual) teenagers who manipulate their pronunciation as part of their social practice. For example, she argues that: The urban variables are the only ones that boys, and particularly burnout boys, can use with impunity. They are the only ones that clearly indicate an urban stance, and hence are associated with toughness, autonomy, and urban know-how – a kind of ownership of the local area. Variables that are more of a generally expressive resource, on the other hand, detract from a cool image, and conflict with the more inexpressive demeanour associated with both physical and technical masculinity. (Eckert, 2000: 225) The changed view of the relationship between social structures and linguistic variation has implications for sociolinguistic data collection. Classical sociolinguistics’ large-scale studies of one or a few variables with a large number of informants are scarce in newer sociolinguistics. There are of course several advantages with classical large-scale data collection. Execution of the data collection and primary data treatment is usually simple and fast. Vast masses of data are easily quantifiable, and generalisability is usually good. On the negative side, however, we find that classical studies are often restricted to description of variation, and they do not present very much in terms of the meaning of variation. Since exactly the meaning of linguistic variation has come very much in focus in newer sociolinguistic, other data collection methods must be used. Therefore we typically see studies of few speakers, but with the inclusion of – at least in principle – their entire language use, or at least their language use in particular settings. Such data may not give very much in terms of generalisability, but on the other hand we can get much deeper into the meaning of particular instance of language use by the few speakers. This leads us to a crucial difference between classical and newer sociolinguistics. The perspective of the unit of social grouping to be studied in scoiolinguistics has changed. Classical sociolinguistics works with a notion of speech community, and linguistic variation is seen as a reflection of social variation in a speech community. This concept has been crucial at least since Bloomfield: A speech-community is a group of people who interact by means of speech …. All the so-called higher activities of man – our specifically human activities – spring from the close adjustment among individuals which we call society, and this adjustment, in turn, is based upon language; the speech-community, therefore, is the most important kind of social group. (Bloomfield, 1935: 42) Rampton (1999a) says that classical sociolinguistics’ concentration on large societal units has been substituted by modern sociolinguistics’ focusing on much

Linguistic Construction and Negotiation of Social Relations

5

smaller units or groups. He also describes how the concept of speech community has been at least partly substituted by a concept of community of practice (cf. Eckert, 2000) through which the sociolinguist attempts to explain the linguistic variation in relation to norms and behaviour, for example in youth groups. To satisfy such an ambition, the sociolinguist must involve new methods of analysis. Anthropologically and sociologically based methods of analysis have found their way into sociolinguistics. A good example is the conversational analysis approach to linguistic data, an approach which was originally developed in sociology and later found its way into a range of linguistic disciplines (see e.g. Steensig, 2001 who claims to build the foundations of a grammar on the basis of the CA approach; or Wei, 1998 who points out how the CA approach can contribute to our understanding of code-switches). The study of bilingual behaviour as a branch of sociolinguistics has therefore undergone a change in emphasis, methods and data. With the introduction of social psychological perspectives, the relationship between social and linguistic variation has been turned around. To a large extent, studies from the latest decade emphasise how linguistic varitation creates social structures and social differences. This is certainly true of bilingual behaviour among minorities in the major cities of post-industrial Western Europe. In focus we have the simultaneous use of linguistic material from separate sets (‘languages’ or ‘dialects’ or ‘varieties’), or perhaps better: from sets which by some (typically monolingual) language users are considered to be separate languages or varieties. Auer (1999) attempts to cover the whole range of such simultaneous use from loans to so-called fused lects, and to unite structural descriptions with a conversational-analytical approach. The latter approach has been particularly fruitful in a number of the studies referred to above. There are also several ongoing projects which apply the same perspectives to their studies of linguistic minority youth members’ bilingual behaviour. Some of these studies are presented in this volume, which draw on particularly on two types of youth language projects. Both involve bilingual young people who speak Turkish as their mother tongue, and their majority language speaking peers. First, we have the Turkish-German studies. In Hamburg, Dirim and Auer’s study of the use of Turkish among young speakers of non-Turkish descent has provided the data for Dirim and Hieronymus’ paper in this volume. Altogether 25 non-native speakers of Turkish have provided tape-recordings of their conversations which involved Turkish in varying degrees. The data have been carefully transcribed according to CA conventions (see e.g. Auer & Dirim, 2000). Hinnenkamp relies on his extensive material of recorded conversations among Turkish-German bilinguals further south in Germany. These recordings were also made by the speakers themselves. Hinnenkamp has both audio-recorded and video-recorded data at his disposal, and he also works with careful and detailed CA-transcriptions of his data. A further study of the same kind is Kallmeyer & Keim (forthcoming) which involves Turkish-German bilingual girls and their multicultural peer groups. All these collections of conversational material have acquired their data through time-consuming processes that focused on a very few informants. Hinnenkamp’s data is a particularly good example, as it has grown steadily by way of the snowball method. This has the unintended effect that the majority of his speakers are male (see Hinnenkamp, this volume), but in this connection it

6

Bilingualism and Social Relations

does not matter. What matters is that the young males form a relevant community of practice, and Hinnenkamp can analyse the relationship between their language use variation and their social relations. The Turkish-German studies namely draw conclusions with regard to the specific youth groups in question (that is, why it matters whether one speaks about a certain quarter in Hamburg, or Køge in Denmark) and not about entire speech communities. This does not mean that such studies do not have implications for our macro-level understanding of the relationship between social structures and language variation. As Hinnenkamp’s results indicate, there are obvious consequences for the educational systems in North Western Europe. The range of languages and varieties which could be included in the regular schooling of these societies must be much wider than what we observe today. Furthermore, there is an obvious need to prepare grade school students of all ages to be able to understand their ‘own’ language spoken with different accents. This is probably most necessary for the mother-tongue speakers of the majority languages. The remainder of the contributions in this volume relate to Køge in Denmark. The Køge Project is a 10-year longitudinal study of the bilingual development of Turkish-Danish children and adolescents attending grade school in Køge, Denmark. The project has collected a wide range of data, linguistic, sociolinguistic, and pedagogical, from the bilingual children who began in Grade 1 in 1989. These children belong to the second generation of the labour migration wave which most of Northwestern Europe experienced around 1970. The students who take part in the study were born and raised in Denmark by immigrant Turkish-speaking parents (at the time almost the only minority represented in Køge), and they are thus second generation Turkish-Danes. The study has followed two classes of children from their school start through to Grade 9 and collected a wide range of data. The material includes transcribed tape-recordings of the bilingual students’ group conversations with peers, both in groups with three or four bilingual students and in groups with monolingual and bilingual students. These conversations were recorded every year for nine years. There are also face-to-face conversations between the bilingual students and adults, monolingual Danish speakers, and Turkish speakers respectively. For further details, see Turan (1999), Holmen and Jørgensen (2000), Jørgensen (2001). Cromdal’s, Esdahl’s, and Madsen’s papers deal with power relations among adolescent speakers involved in group conversations. Esdahl presents a quantitative analysis of the language choice patterns from Grade 1 through to Grade 9 of the bilingual students in the Køge Project, and this leads her to concentrate on the seventh grade. She finds that the change in language choice behaviour, especially among the girls, calls for special attention. Therefore she studies two conversations in the seventh grade qualitatively. One conversation takes place among girls, and the other one among boys. She finds certain similarities in the language choice patterns of the two genders. Both boys and girls are beyond the stage where one language is a we-code and the other one a they-code (although such a stage is possible at an earlier age), and both use code-switching as a tool in their negotiations of social relations. However, there are strong differences in the power relations within the groups. Contrary to classical feminist sociolinguistics, she finds that the girls compete, interrupt and attempt to

Linguistic Construction and Negotiation of Social Relations

7

suppress each other by means of language, much more than the boys do. This finding is supported by Madsen who analyses different variables (patterns of initiatives and responses, conflict outcomes) at different grade levels. She finds that girls do indeed compete more than the boys. She also finds that the girls adapt to the ways that the boys use language (see also Jørgensen’s paper) in mixed-gender conversations more than the boys change their ways. Her most important finding, however, is that the outcomes of linguistic power struggles seem to a large extent to be determined by social structures. There is very little change in who wins over whom from the early grades to the older grades in the students’ school years. Those who are linguistically strong at an early age seem to maintain their advantage through the years. They simply continue developing their skills, including their pragmatic skills such as code-switching, earlier and more efficiently than the others. This means that they are also better at using linguistic variation in the negotiations of social relations. Esdahl’s analysis of Conversation 702 documents this. The results imply that the hierarchies may be already set before the children begin school, or they are established very early in their school career. If this is the case, then the usual suspects cannot be rounded up for indictment. The bilingual students in the Køge Project were remarkably similar with regards to their parents’ socioeconomic background, education, employment, etc. We may have to look for social structures among the children when they are very young, which may possibly be established very early. Since most societies in Western Europe have some institutional socialisation of children before school, we may need to study the way specific social hierarchies are built by the children. Cromdal takes a step in this direction through a conversational analysis of a conversation from the eighth grade of the Køge Project. He finds that there is a certain division of labour between Danish and Turkish, although in a simple way. The ‘story-line’ developed by the participants is almost only in Danish while their negotiations around it are bilingual and involves code-switching. Cromdal shows how participation in the group activity initiated by the adults is used by the stronger member of the group to exercise power over the others. Thus the structures and values brought into the conversation (e.g. Danish for the official school activities) do play a role, but on the other hand, the social relations among the participants are certainly up for negotiation. Cromdal’s findings are supported by other studies from the Køge Project, notably Steensig, 2000. Cromdal finds, just like Esdahl, Madsen and Jørgensen, that the girl Esen maintains her strong hold over the other participants. The reciprocal relationship between societal structures and linguistic variation is a theme that runs through all the contributions to this volume, not only those from Esdahl and Madsen. Dirim and Hieronymus carefully explain that the spread of Turkish in unexpected groups of speakers does not amount to a revitalisation of Turkish in Hamburg. They find that it says more about social creativity in the communities studied. The young speakers use the resources they bring with them, ‘in their interactions and particularly as a means to negotiate identity’ as it says in the concluding remarks of Dirim and Hieronymus. This concept of resources brought into the linguistic interaction and used in social negotiations is a parallel to Madsen’s finding.

8

Bilingualism and Social Relations

Hinnenkamp takes us a step further than realising that there is a reciprocal relationship between social structure and linguistic variation. He analyses in close detail excerpts from conversations with respect to code-switching patterns and the meaning of specific code-switches. He demonstrates that in some cases code-switches can only be understood in exclusively local terms. The given situation, the conversational context, and the involved languages may have led to a point in the conversation where a code-switch may carry important meaning that could not have been carried by a similar code-switch in other circumstances. Often we may not be able to understand the switch at all post hoc, because we do not have access to the situation and cannot repeat the circumstances. In other cases code-switches, and the practice of code-mixing, relates clearly to the general German majority discourse about die Ausländer. The constitution of a mixed language practice is also a construction of social space where the hybrid language practice becomes one we-code available to the minority. An indication of this is the minority speakers’ use of stylised immigrant German as mimicry (Bhabha, 1994) of the majority discourse. Hinnenkamp takes us through a range of code-switches with different functions. In total they show how attempts to explain and categorise code-switches according to their linguistic structure or their surface value must fail. Only when we see the code-switches as choices made by speakers in given circumstances can we describe them in their full range. It is obvious that social structures and societal evaluations are important factors that contribute to determine how language is used. But it is equally obvious, according to Hinnenkamp, that the speakers actively involve outside social structures and to a certain extent criticise and re-evaluate them through their language practice, in casual code-switching. The fact that Hinnenkamp’s results are drawn from data with young language users is probably no coincidence. In my paper I find that the code-switching practised by young Turkish-Danish speakers can be described in the same terms as youth language in general. I go into a fifth grade group conversation with two girls and two boys from the Køge Project. The analysis leads to the conclusion that it is meaningless to consider the speakers ‘bi-lingual’ as opposed to ‘mono-lingual’, as if one could count the number of languages or varieties available to them. ‘Bilingualism’ is probably anyway better understood as a term coined by the usual western narrow-mindedness when it comes to linguistic variation (see also Meeuwis & Blommaert, 1998). On their own terms, they are languagers like the rest of us, and it is of secondary or even lower importance that they may seem to share a few more varieties than others do. For the Western European nations who have built their self-understanding on 200 years of nationalist romanticism including a monolingual ideal, this is serious news. Even the most liberal language policy in the European Union takes linguistic Reinheit for granted. All the activities and initiatives emphasise the need to ‘preserve’ languages, which in the end means keeping them from changing. We have here seen linguistic production involving elements from varieties which are by some people considered ‘different’ languages. And we have seen how advanced and sophisticated such practices can be. In the political thinking and language planning of the European Union, such facts are not taken into account.

Linguistic Construction and Negotiation of Social Relations

9

A comparative study of linguistic politeness among young speakers in Europe (Huls, 1991) has inspired Huls et al.‘s paper in this volume. They have compared material from Køge with material from Rotterdam in the Netherlands, including Turkish-speaking bilingual adolescents and majority-language adolescents in both places. Huls (1991) devised a questionnaire to study variation in linguistic politeness by young speakers of different languages. In the questionnaire, the informants are presented with a task and they are given a range of possible linguistic reactions. They then have to choose among the specified linguistic reactions, according to whom they are talking. According to classical politeness theory, linguistic politeness is determined by social norms that predict how differences in status determine the linguistic choices made by speakers. Contrary to this, Huls et al. find that only one of the four groups (the bilinguals in Rotterdam) could be said to fit into the politeness theory. At the other extreme were the majority speakers in Køge, who showed very little inclination to calibrate their linguistic choices according to interlocutor. Huls et al. interpret this as this group’s linguistic construction of social equality. This study lends tremendous support to the claims made by Hinnenkamp and myself. If the detailed study of group conversations shows us that social negotiations are indubitably going on in peer groups, and extrapolation tells us that values brought into the groups from outside and from society at large (i.e. the adults), then Huls’s study gives us confirmation that the young speakers – at least in the Køge Project – intend to do differently than classical politeness theory and classical sociolinguistics predict. The papers that are presented here come from a remarkably small selection of studies, and at first glance a narrow one, too: small groups of adolescent speakers of Turkish in Rotterdam, Køge and a cople of German cities. This provides for a community of practice-studies, if anything. We are dealing with the relationship between on the one hand the children or grandchildren of labour immigrants from Turkey, and on the other hand urban, western, Germanic societies. We are indeed taken down to very small details, linguistically, culturally, and in the explanation of everyday phenomena in these societies. Nevertheless, the material is so rich and varied that the depth of our understanding of the speakers’ language use is profound. We find that they are actively involved in the linguistic negotiations of the social structures between them and around them. We have seen that linguistic variation is indeed a crucial phenomenon in these negotiations. Linguistic variation understood as the oscillation between different languages, or different varieties, or different styles, etc. does not function in any way differently from other aspects of language use. We have also found that the relationship between linguistic variation and social structures is reciprocal. Societal evaluations, differences in status, public and political discourse all set their mark on what goes on between the adolescents. However, the speakers’ active manipulation of identities sometimes handles these societal phenomena without very much respect. The values are re-negotiated, turned around, made fun of, and in other ways challenged. But they are also used with their surface value, when it suits particularly the stronger of the young speakers. The implications for the generally social democratic societies in Northwestern Europe is that variation must be embraced. The concept of one nation–one people–one language should be discarded as useless, absurd,

10

Bilingualism and Social Relations

even destructive. The young adolescents with a wider range of linguistic resources are as language users not inherently different from the majority of adolescents. They are certainly not poorer language users, either. But as Hinnenkamp demonstrates, their specific skills do not receive all the appreciation they deserve. Although the effect of general societal evaluations on the language use of our languagers is considerable, there seems to be little acknowledgement of their skills. And even less will to let the skills we have witnessed here have any bearing on the organisation or content of the educational systems. Correspondence Any correspondence should be directed to J. Normann Jørgensen, Department of Nordic, University of Copenhagen, Njalsgade 80, DK-2300 Copenhagen S, Denmark ([email protected]). References Auer, P. (ed.) (1998) Code-Switching in Conversation. Language, Interaction, and Identity. London: Routledge. Auer, P. (1999) From codeswitching via language mixing to fused lects: Toward a dynamic typology of bilingual speech. Intenational Journal of Bilingualism 3, 309–332. Auer, P. and Dirim, I. (2000) On the use of Turkish routines by adolescents of non-Turkish descent in Hamburg. In A. Holmen and J.N. Jørgensen (2000), pp. 157–194. Bhabha, H.K. (1994) The Location of Culture. London: Routledge. Bloomfield, L. (1935) Language (revised edn). London: George Allen and Unwin. Eckert, P. (2000) Linguistic Variation and Social Practice. Oxford: Basil Blackwell. Ferguson, C.A. (1959) Diglossia. Word 15, 325–340. Fishman, J. (1965) Who speaks what language to whom and when? La Linguistique 2, 67–88. Gumperz, J.J. (1982) Discourse Strategies. Cambridge: Cambrige University Press. Herman, S.N. (1961) Explorations in the social psychology of language choice. Human Relations 14 (2), 149–164. Holmen, A. and Jørgensen, J.N. (eds) (2000) Det er Conversation 501, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe. Copenhagen Studies in Bilingualism, the Køge Series, vol. K8. Copenhagen: The Danish University of Education. Huls, E. (1991) Een landenvergelijkend onderzoek naar de formulering van een verzoek. In R. van Hout and E. Huls (eds) Artikelen van de Eerste Sociolinguistische Conferentie (pp. 241–258). Delft: Eburon. Johnson, S.A. (1998) Theorizing language and masculinity. In S.A. Johnson and U.A. Meinhof (eds) Language and Masculinity (pp. 8–26). Oxford: Blackwell. Johnson, S.A. and Meinhof, U.A. (eds) (1997) Language and Masculinity. Oxford: Blackwell. Jørgensen, J.N. (2001) Bilingualism as an instrument for mastering sociolinguistic reality. In S. Björklund (ed.) Language as a Tool. Immersion Research and Practices (pp. 42–69). Vaasa: University of Vaasa. Kallmeyer, W. and Keim, I. (forthcoming) Linguistic variation and the construction of social identity in a German-Turkish setting. A case study of an immigrant youth group in Mannheim, Germany. In J. Androutsopoulos and A. Georgapoulou (eds) Discourse Construction of Youth Identities. Labov, W. (1966) The Social Stratification of English in New York City. Washington, DC: Center for Applied Linguistics. LePage, R.B. and Tabouret-Keller, A. (1985) Acts of Identity. Creole-based Approaches to Language and Ethnicity. Cambridge: Cambridge University Press. Meeuwis, M. and Blommaert, J. (1998) A monolectal view of code-switching: Layered code-switching among Zairians in Belgium. In P. Auer Code-switching in Conversation. Language, Interaction, and Identity (pp. 76–98). London: Routledge.

Linguistic Construction and Negotiation of Social Relations

11

Milroy, L. (1980) Language and Social Networks. Oxford: Blackwell. Rampton, B. (1995) Crossing. Language and Ethnicity Among Adolescents. London: Longman. Rampton, B. (ed.) (1999) Styling the other. Theme issue of the Journal of Sociolinguistics 3 (4). Rampton, B. (1999a) Speech community. In J. Verschueren, J. Östman, J. Blommaert and C. Bulcaen (eds) Handbook of Pragmatics. Amsterdam: John Benjamins. Sebba, M. (1993) London Jamaican: Language Systems in Interaction. London: Longman. Steensig, J. (2000) Notes on some uses of code-switches and other interactional devices in conversation 801. In A. Holmen and J.N. Jørgensen (2000), pp. 9–56. Steensig, J. (2001) Sprog i virkeligheden. Bidrag til en interaktionel lingvistik. Århusd: Aarhus Universitetsforlag. Tannen, D. (1991) You Just Don’t Understand: Women and Men in Conversation. London: Virago. Trudgill, P. (1974) The Social Differentiation of English in Norwich. Cambridge: Cambridge University Press. Turan, F. (ed.) (1999) A Text Collection of Turkish-Danish Bilingual Grade School Students’ Conversations. Copenhagen Studies in Bilingualism, the Køge Series, Vol. K6. Copenhagen: Royal Danish School of Educational Studies. Wei, L. (1998) The ‘why’ and ‘how’ questions in the analysis of conversational code-switching. In P. Auer (ed.) Code-switching in Conversation. Language, Interaction and Identity (pp. 156–176). London: Routledge.

Mixed Language Varieties of Migrant Adolescents and the Discourse of Hybridity1 Volker Hinnenkamp FH Fulda University of Applied Sciences, Intercultural Communication and European Studies, Fulda, Germany This paper is a study of mixed language use within the frame of interactional sociolinguistics. Adolescents of Turkish background in Germany have developed their own patterns of bilingualism, in particular mixed varieties of German and Turkish. By the analysis of some transcribed examples of Turkish–German mixed speech (code-switching, code-mixing, code-oscillation, stylised forms of ‘Immigrant German’) it is shown that the switching and mixing oscillates between local sequential functions and a more global ‘We-group’-function. The particular role assigned to ‘speaking mixed’ can only be appreciated against the specific background of the German migratory discourse. The adolescents’ use of hybrid language is an attempt at appropriating semantic space where their language is no longer the object of the migratory discourse as defined by majority society but constitutes an autonomous and exclusive form of (counter-) discourse in its own right.

Across a whole range of cultural forms there is a ‘syncretic’ dynamism which critically appropriates elements from the mastercodes of the dominant culture and ‘creolises’ them, disarticulating given signs and re-articulating their symbolic meaning otherwise. The subversive force of this hybridizing tendency is most apparent at the level of language itself where creoles, patois and Black English decentre, destabilise and carnivalise the linguistic domination of ‘English’. (Mercer, 1988: 57)

Introduction At around lunchtime at the school complex Alter Postweg in Augsburg hundreds of pupils pour into the tram. A multilingual jumble of voices arises. What reaches my ears is not only German, Turkish, Greek, Russian and other languages, but I can also hear mixed conversations in German and Turkish, German and Greek, or German and Russian, in which languages are switched at breathtaking speed. I listen, amazed by the pupils’ virtuosity, until my research interest takes me back to academic soberness, knowing that these adolescents’ and kids’ linguistic productions are hardly valued, and not at all respected, in the school classes they have just left. Monolingualism and German alone is what counts there. However, the conversations outside the official lessons’ discourse literally do speak another language; one that is varied and diverse, mixed, polyphonic and multilingual. This ‘language’ is ignored and despised by majority society and by the official guardians of the language norm. It is not seen as the logical, and likewise antithetical, outcome of a development that has its roots in the history of recent migration (since the 1960s) and in the consequential emergence of polylingual, polycultural and multiethnic areas in many urban centres in Germany (and elsewhere). 12

Mixed Language Varieties of Migrant Adolescents

13

The linguistic result of this situation is what is nowadays called ‘hybrid language’. In the following sections I present a few instances of this language, and show that although many of the examples of mixing may be explained by the local functions they serve within the sequential proceeding of a conversation, its emergence can finally only be understood within the context of migration history and the particular form of hybridity discourse it has fostered. Despite being called a ‘mixed’ or ‘hybrid’ language, the speech samples described are in no way homogeneous. However, they have many features in common. The most prominent is, of course, the switching between the languages we call ‘German’ and ‘Turkish’. The degree of switching and mixing is highly differentiated, as are the various functions within their interactional logic. What we find is a whole spectrum of bilingual patternings up to genuinely new and autonomous – some would say ‘creolised’ – forms, which do not belong to either of the languages involved. But that is not the only source of variation. Jargonised and dialectal features, and ethnolectal stylisations, also form part of the resources that are exploited for the switching and mixing of the language varieties involved. All of these I call ‘mixed language varieties’. The basis for my analysis is conversations between bilingual adolescents of ethnic Turkish background whose parents or grandparents had originally immigrated as ‘Gastarbeiter’ (guest workers) to West Germany (cf. Hinnenkamp, 1980; Sassen, 1996; Terkessidis, 2000). The conversations were audiotaped and partly videotaped in informal gatherings. It was mostly one of the participating parties that did the recordings. In most situations the adolescents did not know what kind of data the researcher was looking for.2 The majority of my informants were between 15 and 18 years old; some recordings were also made from the conversations of university students between the ages of 20 and 25. With some of the informants I conducted in-depth interviews about their usage of ‘mixed language’. Most speakers in my data are male. Just one of my recordings contain female ‘protagonists’. But the gender bias is solely due to the chain of my informants (male adolescents ask other male adolescents to make the recordings, for example). In female groups there is a similar range of switching and mixing, as can be seen from the data of the Mannheim project (cf. Kallmeyer et al., 2000). The excerpts I present here are solely those which include ‘heavy’ mixing and switching; not all the data did. Many stretches of talk were conducted more or less monolingually, mostly in Turkish. I ignored data with single item insertions and those where the switches were due to addressing monolingual participants.

From the Local-rhetorical Logic of Interaction to ‘Fuzzy Mix’: No Language Mixing à la carte Thematic and contextual contrasts The following excerpt is a sequence from a discussion between two 16-year-old friends, Ercan (E) and Hakan (H), about an ‘Inititiativkreis’ (Interest Group), a kind of social club for ‘foreign’ adolescents.3 Both speakers grew up in Germany and go to the final class of secondary school (Hauptschule). The conversation takes place at Hakan’s home:

Bilingualism and Social Relations

14

(E-1) Transcript ‘Initiativkreis’ (Interest Group) 1

Nerde bu Initiativkreis? Where is this interest group (initiative circle)? 2 E: Richtung Stadt böyle, ordan dümdüz gittiðin zaman It’s about direction of town if you go straight from there 3 Königsplatz çÏkÏyor karÕÏna you get to Königsplatz, 4 H: Ja::::::, ich weiß [(…) Yeahhhh, I know (…) 5 E: [Kennst du schon? You know that already? 6 H: Ja Yes 7a E: Ja, Yes, 7b iÕte ordan tam böyle hani o Initiativkreis tam böyle just from there that is that interest group’s place comes right up 8 Mitteye geliyor. in the middle. 9 O Einbahnstraße[nin tam Mittesinde böyle From the one-way street just quite exactly in the middle 10 H: [Mhh 11 E: Or- [orda The- there 12 H: [Was is das fürn Ding, so kolpingmäßig, oder? What kind of place is that, just kolpinglike or what? 13 E: Nein, nicht kolpingmäßig >{böyle/ ehh}< Lernstudio, No, not kolpinglike >{so/ uhh}< learning studio, ((Com.: ‘Kolping’ refers to a Catholic welfare institution)) 14 saz kurslarÏ °so was halt° (+) °ondan sonra° alles mögliche saz courses, just like that (+) and then all kinds of things ((Com.: ‘saz‘ is a Turkish string instrument)) 15 H: Ja und was bringt des? Yeah and what is it good for? 16 E: °Ja, die verdienen Geld° Yeah, they make money 17 H: Ja und #((laughing)) orda para kaybediyor yani # Yeah and #((laughing)) that means there you lose your money # 18 E: Nnnnnn nich ganz Nnnnn not quite 19 ((0.6 sec.)) 20 H: Ne iÕe yarÏyor? What is it good for? 21 E: °Eh e Geld verdienen, Mann° Uh u making money, man 22 H: Mann, du verstehst nich was ich meine Man, you don’t understand what I mean H:

Mixed Language Varieties of Migrant Adolescents

23

E:

24

H:

25

E:

26

H:

27

E:

28

H:

29

E:

30 31

H:

32

E:

33

15

Eh wie (h) wie yani? Uh how (h) how you mean? Onlar niçin gidiyor oraya? Why do they go there? Kimler? Who? Ja, o die Jugendlichen Yeah, these the young people Die wollen was lernen They want to learn something Lernstudiomäßig {yani}(?) Learning studio like {you mean} Lernstudiomäßig (+) ja, alles mögliche, >ne ararsan var orda< Learning studio like (+) yeah, all kinds, >whatever you look for you find< alles mögliche all kinds of things Cool (+) und nur Türken oder [so oder nur Ausländer? Cool (+) and only Turks or so or just foreigners? [Ähhh Uhhh °Ja° + °ziemlich° yes + rather

The excerpt combines a number of mixed language phenomena. Furthermore, in German as well as in Turkish we find some typical spoken language elisions (such as in ‘nerde‘ line 1 or in ‘nich’ lines 18, 22) as well as dialectal elements (‘des’ [dæ:s], line 15); in German there are also some typical youth language expressions like ‘kolpingmäßig’ lines 12, 13; ‘lernstudiomäßig’, lines 28, 29; or ‘cool’ line 31. The numeric relation of German to Turkish is roughly 2 to 1. However, we have to keep in mind that the suffixing principle of an agglutinating language like Turkish may pack much more information into one word. Also some formal aspects of language alternation are worth looking at: 12 out of about 26 turns of speaking lines are monolingually German as opposed to 5 or 6 turns in Turkish (including the insertion ‘Initiativkreis’ in line 1). Longer sequences tend to alternate languages. Here we find German dominance as in lines 12f. or Turkish dominance as in line 17. We must bear in mind that looking at the mixed data like this remains purely formal and normative. With a couple of lexemes it is in no way clear to which language they should be attributed; proper and quasi-proper names like ‘Initiativkreis’ and ‘Königsplatz’ are not counted (lines 1, 3, 7). But what about insertions such as ‘Mitte’ and ‘Einbahnstraße’ (lines 8 and 9)? By their suffixation we can see how they are fully integrated into Turkish. Although the four word sequence part ‘O Einbahnstraßenin tam Mittesinde‘ consists of two German content words as opposed to two Turkish functional items is it nonetheless a genuine Turkish sentence. As we can see a formal approach is very limiting. Just taking my own way of transcribing, intended to make reading easier, that is German in roman, Turkish in italics, it remains quite problematic as it proceeds on purely technical grounds according to language assignment.

Bilingualism and Social Relations

16

What is much more promising is taking a look at the switches themselves, such as those between lines 3 and 4, or between lines 11 (7ff. respectively) and 12, 17 and 18; or answering the question why there is a switch at all in line 17 or in lines 28 or 29. Bilingual speakers have both languages available, as we know. Some instances might be explained by word searching difficulties and other competence related reasons. Also, we know of preference principles such as following the previous speaker’s choice (Auer, 1988). But there are no really satisfying answers to be found here. Is ‘Richtung Stadt böyle‘ the adequate coupling with ‘bu Initiativkreis’ (line 1f.)? The answer has to be ‘no’. In fact, most of the sequence where the ‘Initiativkreis’ is located is negotiated in Turkish (up to line 11). Then we have a language switch. In what follows there is shift in language dominance (lines 12 to 16). The non-Turkish sequence, lines 4 to 7a, is clearly an insertional sequence in which H tries to explain that the ‘Inititiativkreis’ is in fact already known to him. It constitutes a formal opposition to E’s description of where the ‘Initiativkreis’ is located: it is an autonomous sequence refuting E’s over-explicitness (line 4), followed by a checking question (line 5), H’s confirmation of that (line 6) and its reconfirmation (line 7a) before both return to the status quo ante in line 7b ff. We thus find language choice-wise, as well as contextually, a hierarchical structure.4 We come across similar methods of opposing formats in the sequence containing lines 15 to 18, through H’s contradiction by his questioning of what the ‘Initiativkreis’ has to offer (line 15), E’s answer (line 16) and H’s ridiculing conclusion about that (line 17). For E’s mildly formulated protest against this he remains with his prior choice, i.e. German. That is, only in H’s concluding statement ‘orda para kaybediyor yani‘ (line 17) does he make use of this section’s contrastive language, Turkish; furthermore, marking it by laughter might also emphasise it as modally contrastive. Authenticity and narrative refinement As a matter of fact, formal, contextual and narrative oppositions serve as an ideal background for code-switching. As an example, let’s take a look at just a small excerpt from a rather long and extended account where Orhan (O), a student, tells two friends how he happened to meet his old schoolmate Matthias at Munich airport. Ayhan (A), also a student, gives him support at one point.

(E-2) Transcript ‘Matthias taucht auf’ (Matthias shows up) 1 2 3

4

O:

ndim, Selda’yÏ ar2yom bakÏyom. Got out of the car to look for Selda Bi baktÏm Matthias’Ï diyor hey kannsch du mi: mitnehmen? Then all of a sudden I see Matthias; says he hey can you give me a lift? Is- isn Freund von mir, mit dem ich früher inner Sch- eh Klasse war. He- he’s a friend of mine with whom I used to go to sch- uh into the same class. He:, kannschte mi: mitnehmen diyo, eh i hab niemand diyo sonst muss Hey, can you give me a lift, he says, uh nobody I know here, he says, otherwise

Mixed Language Varieties of Migrant Adolescents

17

5

ich mit- mitm Dings (+) [mitm Bus oder mit der U-Bahn I have to go by uhm (+) by bus or by tube 6 A: [Bus fahren go by bus 7 O: Augsburg’a gelmem lazÏm diyo + yi dedim, gel + baktÏm to Augsburg I have to go he says + Right I said, come on + I looked around 8 Selda da geldi, Selda’yÏ da ald2m (+) Und dann hab I: gsehn and Selda showed up as well, I picked Selda up as well (+) And then I saw 9 Veli kommt auch (+) Veli came as well (+) Again we find a couple of switches, such as in line 2, in line 4 and in line 8. Their local functions are quite obvious: the narrating language here is Turkish. At the point where Matthias is quoted O changes into German. Matthias is German, he is speaking in German. His direct question is introduced by the Turkish verbum dicendi ‘diyor‘ (he says; line 2). The switch not only serves narrative authenticity but also creates a nice contrast within the episodic progression. The same in line 4 and 5: Matthias is quoted, twice bracketed by the colloquial form ‘diyo‘ (he says). Ayhan immediately complies with the switch. His assistance in line 6 is hence in German, the language that is locally relevant here. Also, further above, the non-return into Turkish in line 3 makes sense, because it is not part of the main storyline, but constitutes – as already seen in the prior example – a sub-sequence which explains the history of Orhan’s relationship to Matthias. However, the switch back into Turkish in line 7 from ‘Augsburg’a‘ on onwards, i.e. after the directional suffix or postposition, does not comply with the interactional logic, because it is still Matthias who is talking here. But, as we know, transition points to another language may be fuzzy. The narrator may be anticipating the switch for the part where he continues his account how he met the others, etc. But at the same time, of course, he constrains his grammatical selection. In that O ‘suppresses’ the German preposition ‘nach Augsburg’ a grammatically ‘correct’ continuation is only postpositionally possible, namely in Turkish. Furthermore, O seems not to owe his bilingual audience further evidence of authenticity of what Matthias has said to him. Evidence that O is already back to his main narrative progression can further be seen by the change of present tense (diyo – he says) into the narrative past (dedim – I said), a tense marking difference which we can already find in line 1 and 2. About the last switch, in line 8, from Turkish into German, it could be said that it corresponds with the showing up of another person, Veli, thus stylistically marking another side event. In summing up this brief section we find quite a dramaturgic oscillation between personae and events in which the alternation of languages clearly serves the progress of the narration. Inverted opposition formats and other surprises: Playing with contrasts One gets a false sense of security, however, if one regards intrasequential opposition or other opposing patterns as a central key to explaining language alternation. The following example is a transcript from a conversational exchange that a student of mine spontaneously recorded at a bus stop. Here we come across the two 15-year-old adolescents Ferhat (F) and Ahmet (A) who are

Bilingualism and Social Relations

18

waiting for the bus and make nasty comments about the bus driver and the bus service in general.

(E-3) Transcript ‘Bushaltestelle’ (Bus stop) 1

Otobüse binecekmiyiz? Will we get on that bus? 2 A: #((laughing)) Ich weiss nicht# I don’t know 3 F: °{Lan}° + bugün zaten öðretmen kýzmϺtý bize Son + today the teacher told us off 4 A: #((laughing and sucking in air)) Echt oder?# Really? 5 F: Bugün geç kalmýºtým, otobüsü kaçýrmýºtýk Today I was a bit late, we missed the bus 6 A: Ben de saat acht’ta geldim camiye, lan hehehehehehehehe And I got to the mosque at eight, man 7 F: He:: der Busfahrer ist (h)ein Sack hey Hey, the busdriver is a bugger 8 A: Hehehe valla:::h hehe really 9 F: der kommt (h)der kommt immer zu spät he he comes he comes always too late ha 10 A: Otobüsün dolu olmasÏna çok gicik olyom hey Mann ge + voll It really gets on my nerves that the bus is so packed, hey man, ha + full 11 F: Ja weisch (+) girdik (h) {giriº/giriyoz = ºimdi} içeriye You know we got in {we got onto the bus/get on now} inside 12 A: [((laughs)) 13 F: [bi- bize (…) (+) seid mal leise diyor ehh das regt mich auf hey to us (…) (+) be quiet he says, hu, that really gets on my nerves 14 A: #((laughs for about 3 sec. swallowing words))(…)hohohohohehehehe 15 Ýyi mi? kötü mü?# ((sucks in air)) Is that fair or is that bad? 16 F: ((enervated)) Eh komm jetzt Eh come on now 17 A: ((self-controlled, with deep voice)) Ya tamam burdayýz=lan Yes alright, here we are=man 18 F: Wo bleibt der Bus hey Where’s the bus, hey 19 ((1 second)) 20 A: Ya abi çekiyo »hasch immer noch was zu sagen, oder?« Our brother’s recording. Still got something to say? ((Com.: meaning the interviewer)). 21 F: (…) 22 ((2.5 seconds)) 23 A: »Fenerbahçe’nin en son durumu kaçtý, lan« What’s Fehnerbahçe’s last position, son ((Com.: that’s a football club)) F:

Mixed Language Varieties of Migrant Adolescents

24

19

Ich weiß net (+) ich glaub die ham verloren [(…) I don’ know (+) I think they’ve lost (…) 25 A: [Zwei zu bir miydi? Was it two to one? 26 F: Na (+) zwei zu- (+) zwei, glaub=ich Nope (+) two (+) two, I think 27 A: Unentschieden Draw 28 F: Jaja Oh yeah The first impression is that the two languages are used more or less evenly (50 Turkish as against 58 German words). Again we find monolingual as well as bilingual sequences. A type of alternation we could spell out as ‘Speaker 1 speaks language A, speaker 2 language B’ is found e.g. in lines 1 to 5 or lines 6 to 9. Line 5 and 6 are in Turkish only and the last lines of the excerpt are almost only in German. Also, within a turn we find those typical one-word insertions like ‘acht’ in line 6. In lines 10, 11, 13, 20 and 25 we find switches within the respective turns which, in my opinion, are syntactically as well as functionally quite comprehensible. If we take line 13 as an example, we can immediately see that we have (a) the Turkish bracketing of a German quote and (b) the ensuing commenting of the quoted occurrence, a rhetorical and thematic differentiation by means of the two languages, German and Turkish. Or line 20, for example, is an obvious case of addressee specification: in the Turkish part, A refers to the Turkish student (in 3rd person) who makes the recording; in the German part – spoken in an accelerated manner – the immediate adressee is his mate F (in the 2nd person). Even if not every single switch is explicable, it is quite obvious that most of them nicely comply to phrase boundaries [//] such as in line 10: ‘Otobüsün dolu olmasýna çok gicik olyom // hey Mann ge + voll’ or in line 11 ‘Ja weisch (+) // girdik (h) {giriº/giriyoz=ºimdi} içeriye‘. Not all alternations, however, stick to these boundary rules. In line 25 ‘Zwei zu bir miydi?’ (Was it two to one?) the switch ‘respects’ no other boundary than just that between words. Its rhetorical function remains obscure. And if we once more take a look at the language alternation distribution according to speaking turn and speaker, more puzzles come up: the first five speaking turns seem to be characterised by the different language dominance of the interlocutors: F speaks Turkish, A replies in German. This might correspond to individual language preference or to the degree of (un)certainty in the respective languages. In the next turn, however, (line 6) A switches to Turkish, thus complying with F’s language choice. Surprisingly, F continues in German (line 7), likewise in the turns to follow (lines 8 to 10) we find the above pattern inversed before both speakers use both languages within their turns (lines 10, 11 und 13). To conclude: thus far we have not really found a reliable pattern of interturn language alternation unless we declare the maxim ‘Don’t use prior speaker’s language!’ as at least valid for parts of the conversation. We would thus have discovered a further pattern based on opposition – on formal speaking turn opposition. There is not much narrative or dramaturgical logic behind that; more the potential for playing with oppositional resources. What, we have to ask, is it that stimulates F:

Bilingualism and Social Relations

20

the adolescents to conduct their conversational interchanges in such manifold code alternating ways – comprehensible as well as surprising ways? From extensive alternations to oscillations: About the successive increase of mixing density We find extensive patterns of alternations side by side with intensive patterns. The following two excerpts stem from a rather long account about a dramatic accident with nearly fatal consequences. The excerpt ‘Accident I’ is from the introductory part of the account, ‘Accident II’ is about 12 minutes later when it comes to a dramatic climax within the narration. Nineteen-year old Remzi (R) is the narrator. He came to Germany when he was eight years old and is still in vocational training. His listener is 15-year-old Yasemin (Y). She was born in Germany and goes to grammar school. The two are at R’s home. Y, who has not yet been informed about how the accident happened, is an eager listener to R’s detailed descriptions.

(E-4) Transcript ‘Accident I’ 1

R:

2

Y:

3

R:

4

Y:

5

R:

6 7 8 9 10 11 12 13 14 15

Y: R:

A- arkadaºýma gittim {bilyomusun} kaza yaptýlar I had gone to my friend {you know} they had the accident Wo? Where Bilmiyom. Hab i dir des nicht verzählt oder? I don’t know. Haven’t I told you that? °Hayýr° No Ach so. Ich wa:r + halt mit (h) meim (h) Freund bissl unterwegs + I see. I was just away with my friend »{und dann/ondan} dedik ki arkadaºýmýza gidelim dedik {then/because of that} we thought let’s go to see our friend kaza yaptýk, [ya (?) geçen cuma« we had this accident last Friday [Mh ݺte bir hafta oldu ya + camideydik (+) sonra iºte + bir iki About a week ago + we were in the mosque (+) then + we were one two dört kiºiydik dört arkadaº Kenan, Tahir (+) ich und Ahmet + four of us were there, four friends Kenan, Tahir (+) me and Ahmet + und dann sind dort zwei dazu kommen Taner’le Baki geldiler + there two others came as well Taner and Baki came Katzdorf’da oturuyorlar, ya(?). Namazý kýldýktan sonra camide they live in Katzdorf , you know(?). After praying in the mosque + kahve içmeye gidelim dediler bize + + they said to us come on let’s have a cup of coffee + pe- peki tamam. Neriye gidelim? (+) Lauenberg’e gidelim ri- right, okey. Where shold we go? (+) Let’s go to Lauenberg »hani Taner orda« niºanlandÏ ya(?) that’s where Taner has his fiancée, right(?)

Mixed Language Varieties of Migrant Adolescents

16

Y:

17

R:

18 19 20 21 22 23 24 25 26 27 28 29 30

Y:

31

R:

32 33 34

Y:

21

Ja ja eh eh in (h) zwei Autos oder in drei Autos? Yes yes uh uh in two cars or in three cars Ýki arabayla, ja zwei Autos. Aber es war ja so: + wir ham bloß ein With two cars, yes two cars. But it happened like this + we just had one Auto da gehabt. Bende yayan gitmiºtim Camiye=Kenan’lar car with us. I had walked to the mosque=Kenan and Ahmet de sonradan geldi (+) Baki’ler Taner de bir arabayla Ahmet came later as well (+) Baki and Taner seem to have come by car gelmiºler (+)karar verdik kahve içmeye (+) we decided to go for a cup of coffee gidicez Lauenberg’e. Çýktýk camiden (+) sonra tabii bir araba var to Lauenberg. We came out of the mosque (+) then of course we had a car altý kiºi, sýðmýca:z bir arabaya (+) we were six but didn’t fit into one car (+) ne yapalÏm edelim Taner bana sordu so what should we do, Taner asked me araban burdamý diye yok evde dedim. Ahmet‘e de sordu if my car was there, no, it’s at home, I said. He asked Ahmet as well Ahmet’inki de Tiefgarage deydi + Ahmet’s car was in the underground car park, he said + °ondan sonra° ben dedim beni eve atsÏn Baki arabayÏ alÏp geliyim and then I said Baki should bring me home so I could go and get my car iki arabayla gidelim dedi{m} (+) then we can go with two cars {he/I} said (+) Ahmet te ayný Vorschla/kÏ (+) söyledi (+) und die ham and Ahmet made the same proposal (+) and they said no abglehnt °{halt}°, weisch, die ham alle beide so Sportautosyou know, both of them have these sport cars Echt so [Zweisitzer? Really this kind of two-seat car? [{Ja} {Yes} Ne ne scho Viersitzer aber so: + hohe PS-Zahl und so, weisch do No no normal four seat car but with really high HP etc, you know und so richtig Farbe and real colour ((i.e. power)) Ama Taner’inki Zweisitzer deðil mi? (+) But is the one of Taner not a two-seat car?

So much for the first excerpt from the beginning of the accident account. Most of what R says is in Turkish, Y responds in German as well as in Turkish. Her short contributions correspond to her role as listener. In addition, there are some, quite short and comprehensible, sequences of code alternation that inform us about unequally distributed dominances in either language. In line 16 Y’s checking question is in German, not in the hitherto dominating language of the account. R responds first in Turkish but he then repeats the prepositional phrase in German. He eventually answers the question in German; the story, however, continues in Turkish: ‘Ýki arabayla, // ja zwei Autos. Aber es war ja so:

Bilingualism and Social Relations

22

+ wir ham bloß ein Auto da gehabt.// Bende yayan gitmiºtim Camiye=Kenan’lar Ahmet de sonradan geldi …’. (line 18f). Also, another sequence adds to this suspicion of unequally distributed competences: in line 28f. R alters into German ‘Ahmet te ayný Vorschla/ký (+) söyledi (+)// und die ham abglehnt {halt}, weisch, die ham alle beide so Sportautos-’ and it is just here that R is interrupted by another checking question by Y – in German (line 30): ‘Echt so Zweisitzer?’. And R answers in German (line 32f.): ‘Ne ne scho Viersitzer aber so: + hohe PS-Zahl und so, weisch do und so richtig Farbe’. One gets the impression that R changes into German to play a speaking turn on to Y. However, in the next line this pattern is reversed when Y asks in Turkish: ‘Ama Taner’inki Zweisitzer deðil mi?‘ (line 34). Thus, we find at least some insights into some of the patterns of the participants’ code-switching behaviour. Other examples contradict these patterns. The main narrative line in the excerpt discussed is in Turkish; there is no doubt about which is the matrix language. More astonishing, however, is the development of the narrative to quite a dense oscillation of languages at the point where the dramaturgy of the accident seems to dominate the account, as we can see in the following excerpt.

(E-5) Transcript ‘Accident II’ 1

R:

2 3 4 5 6

Y:

7

R:

8 9 10

Y:

11

R:

12 13 14

Ondan sonra o lafý bitmiº (+) tam o viraj gelince (+) hatta gmeint Then that subject seem to be finished (+) just the very moment when the curve showed up (+) he thought durch den Nebel (+) dass da ne grade Strecke kommt, doch kei because of the fog (+) that the next bit was straight, no Kurve, grade Strecke. Dann ist er in der Kurve dimdirek curve, just straight on. Then in the curve he straightly came off yoldan çýktý (+) aðaca çarptýlar the road (+) into a tree, aðaca çarpmýºlar and obviously got right into a tree Es- is da jemand verletzt worden? Did- did anybody get hurt? (h)arkadaºýn ikinci Wirbeli kýrýldý eðer üçüncü Wirbel one friend broke his second vertebra if it had been the third vertebra °wenns gebrochen wär,wär er normal tot° (+) °°da wär er nicht if it had been broken he would be dead by normal (+) he would not mehr jetzt (h)unter uns°° be with us anymore Ehm, Beifahrerin yerindemi oturuyodu oder (h)was? Uh did he sit on the passenger seat or what? Yo:, Ahmet, Ahmet arkada oturuyodu aber er hat sich nicht No, Ahmet, Ahmet sat in the back but he hadn’t had his belt on angeschnallt ghabt, er hat sein Gurt irgendwie nicht gefunden, he somehow hadn’t found his belt, weisch, wo wo so drauf gsteckt war irgendwie you know, where where it’s tagged on was somehow unterm Sitz versteckt (+) i weiß au net, hat es irgendwie net

Mixed Language Varieties of Migrant Adolescents

15

16

17 18 19 20

Y:

21

R:

22 23 24

Y:

25

R:

26 27

28 29 30

Y:

31

R:

32

33 34

23

hidden away under the seat (+) I don’t know either, somehow didn’t gefunden und dann nach dem Roller wo wir yol giderken (+) find it and then after the scooter ((they had nearly knocked over before)) where we continued plötzlich warn sie weg + wir sind schnell hinterher, weischt du they had somehow disappeared + we went faster to catch up, you know (+) aber wir sind dann schneller geworden hýzlý sürmeðe baºladý (+) but we speeded up he began to speed up as well yakalyalým onlarý diye (+) ama yoklardý to catch up with them I think (+) but they’ve been away {ortadan} kaybolmuºlardý (+) nasýl olduysa somehow they had disappeared (+) however it might have happened Siz görmedinizmi onlarýn (h)[çýktýºýný Didn’t you see how they got off [Yo: direkmen çýktýðýný biz görmedik No, the very moment they got off the road we didn’t see ondan sonra virajý biz yavaº döndük Taner biliyormuº dass da (+) then we just took the curve softly, Taner seem to know (+) hat there was scharfe [Kurve ist (+) a sharp curve [Kurve curve und dann plötzlich ham wir alle sað tarafa baktýk (+) Kenan, and then all of a sudden we all looked to the right (+) Kenan Ahmet- Kenan, Taner und ich. Ondan sonra, das war wie im Ahmet- Kenan, Taner and I. And then, that was like on Fernsehn, weisch, wenn du so von Rally-Autos Unfälle siehsch television, you know, when you see these kind of accidents from rally-cars ayný televizyondaki kazalar gibiydi (+) was just like those accidents on television (+) bi baktýk {karp-} alle blicken nach rechts we looked {…} all looked to the right Ÿey olmuºmuydu überschlagen? Was it uhm turned upside down? Yok, über- hats nicht überschlagen (+) yapmamýºlar=iyiki ondan No, upside- no didn’t turn upside down (+) they didn’t=how lucky they were önce iki aºacýn arasýndan geçmiºler iki aðactan birine çarpsalar before they drew right through between two trees, had they hit one of the two trees, zaten takla atalardý bilyomusun (+) ondan sonra hepimiz ham wir they would have toppled, you know (+) and after that all of us we looked auf einmal nach rechts geschaut (+) wir sind alle voll blass to the right all of a sudden (+) our faces just turned pale

24

Bilingualism and Social Relations

35

geworden °im Gesicht° ondan sonra Taner hat zurückgeschaut and then Taner looked back Kenan baktý Taner’e ben de ikisine bakýyom (+) Taner hat Kenan looked to Taner and I looked to both of them (+) Taner gesagt °kaza yapmýlar°. Und dann hat er Vollbremsung eingelegt said they seem to have made an accident. And then he really slammed on the brakes hemen arabayý park ettik kenara (+) wir sind aus dem Auto raus we parked the car at the side of the road (+) we got out of the car gesprungen arabadan nasýl çýktýmýzý bilmiyoz (+) yolu nasýl geçtim but how we got out of it we don’t know (+) how I crossed the road karêý tarafa kaza tarafýn onu bilmiyon={eðer} bi araba gelse how I got to where the accident had happened I dont know= {if} a car had come hätt es mich grad überfahrn. it would have just knocked me over.

36 37

38 39 40 41

In this excerpt there are 26 instances of language alternation within an utterance, not including the integrated one-word switches like ‘Wirbel’ (line 7) or ‘Beifahrer’ (line 10). As we deal here with the roles of narrator and listener, there are only four switches between turns. Of course, some of the switches are easily comprehensible again. Take for example the emphasis in line 37f. ‘Taner hat gesagt °kaza yapmýºlar°‘ where Taner is probably authentically quoted; this is, furthermore, doubly marked by lowered voice as well as by language choice. At other places we find German–Turkish doublings such as: ‘aber wir sind dann schneller geworden hýzlý sürmeðe baºladý‘ (line 17f.); or ‘das war wie im Fernsehn, weisch, wenn du so von Rally-Autos Unfälle siehsch aynÏ televizyondaki kazalar gibiydi‘ (lines 26–28). Redoublings of this kind can either function as stylistic-rhetorical emphases or as concessions vis-à-vis a bilingually less competent partner. Irrespective of not being able to find plausible explanations for each single switch, such local, stylistic and competence-oriented assignments to a switch’s function do not inform us about the increase of switching and mixing density. It seems that the degree of mixing is made a criterion for the narration’s quality, so to speak. We might well conclude that the increase of narrative density, of dramaturgy and of emotional involvement corresponds to the increase and density of language alternations. The document at hand nicely displays the richness of switching variations. This includes, of course, language internal variation, in particular dialect variation, with which the accident account abounds. Variation, unless arbitrary, after all, is always an expression of variable competence. Thus, stylistic-rhetorical, recipient-designed and metaphorical alternations of language are hardly to be gauged as deficits in bilingual competence, rather the opposite: they are an expression of skilfully handling different languages, varieties and registers simultaneously. Formally we deal with alternating forms which move to and fro between insertions, code-switching and code-mixing. There are phases of clearly negotiated switches, and of contextually relevant and plausible alternations; but there are also phases of increasingly dense and accelerated language alternations.

Mixed Language Varieties of Migrant Adolescents

25

‘Code-oscillation’ is a good term for this constant shift in density and functionality. This is, of course, a purely formal description. It does not explain why even among a constant constellation of participants (and even a constant participation framework, we might add) the mixing and switching takes place in such a highly differentiated and variable way. Uncertainties in language competence may always be cited as one reason for code-switching: if one speaker does not know how to continue in one language he or she may revert to the other available language. This is often indicated by signs of uncertainty, by abortions, self corrections and resumptions in the other language. This, of course, is not the whole story as we know from those fast grammatically correct continuations in another language in the middle of a phrase or a constituent, as in our examples (E-2) line 5/7 and (E-3) line 25 or in (E-5) lines 15 and 25. Those without practice in switching would hardly be able to create these mixed language forms. Compositions of this kind require the apt combination of different grammars, demanding, for example, topological and pre- and postpositional anticipations.5 Mixtures of this kind are not infrequent and we should remember that the denser the alternations, the more complex is the grammatical performance in compatibilisation. Reducing them purely to deficits in either of the languages contradicts my empirical findings. Furthermore, many of the things adolescents talk about in a mixed way may also be expressed monolinguallly in either of the languages alternated. Lack of language competence certainly is not a sufficient explanation for the phenomenon of code-mixing. Interim result: The limits of local explanations of language alternations In all the excerpts cited so far, we come across different forms of alternations and switches which we might subsume under the following typology: (1)

(a)

(b)

(c)

Languages alternate within, as well as between, utterances for which we find either local-sequential reasons in terms of an immediate interactional logic and functionality (negotiated or indexical code-switching), or stylistic-rhetorical reasons (not ruling out both at the same time, of course). Its form is not grammatically or compositionally fixed, and in general it corresponds to phrase structure rules; the switches are likewise extensive. Also cognitive reasons and reasons based on competence (as might be documented, for example, by self-corrections, difficulties in finding the proper word or the proper connection) can be decisive for the switch. In (E-1), line 13f. we find a good example of this kind of motivation. We also find discursive and recipient-designed routines, mostly ‘external’ to the utterance. Their function is mostly in the stylistic-rhetorical realm and often with a kind of ritual character, as we see in ‘Lernstudiomäßig yani‘ (E-1, line 28) or the excerpt below from a conversation documented elsewhere (Hinnenkamp, 2000a) where the comment is underlined:

Bilingualism and Social Relations

26

(E-6) 1 2

E:

Geçen tramvayda gidiyom, biliyonmu + bi tane oðlan Recently I took the tram you know + one guy das gibt’s gar nicht you wouldn’t believe it

(2) Some alternations do not have a comprehensible logical function at the particular place where they occur. They are not seen as a result of negotiation, are neither local or indexical, are obviously not stylistically or rhetorically motivated and do not occur for reasons of incompetence. On the formal level we come across more extensive alternations according to speakers (such as in the transcript (E-3) in which each speaker seemed to respond according to the principle of not using the predecessor’s language), but also alternations of the dense kind as we found in (E-1), (E-4) and (E-5). These examples give evidence of switches which cannot be accounted for in terms of their local functions. Furthermore, these kinds of switches do not respect any phrase structure constraints and also create new and autonomous language forms. It should be noted here that there is no static border between (1) and (2). And, of course, their subsumptions are not totally independent from the observer’s interpretative and analytical assumptions. However, this dual categorisation will persist. If the locally negotiated, and hence meaningful, language switch is made an essential criteria for code-switching, then of course (2) would not count anymore as a proper instance of code-switching. In this sense code-switching is constituted by the meaningful alternation of codes (languages, varieties, types of texts, styles, expressive forms). ‘Meaningful’ means that the distinction of code X and code Y has a local function and will serve as a resource for the ongoing interaction or for the interpretability of a text. It’s the here and now of the alternation that offers the potential for negotiating the next step of the interaction. It leads to such implicit questions as ‘how to interpret the switch?’, ‘what does it mean for the continuation of the interaction?’ and ‘what of our momentary relationship?’. The alternation may, in an interaction as well as in a text, also serve as an intertextual or metapragmatic commentary: ‘why a language switch at this particular point?’, ‘what does it imply for further interpretation?’, ‘how is it connected to prior parts of the text?’. If any local meaningfulness is unretrievable then we have to search for other, more global explanations for language alternation (cf. Auer, 1998a, 1998b; Swigart, 1992). It is here that we might find a semantic correlation between formally different types of alternations: a differentiation between code-switching and code-mixing according to formal criteria looks quite difficult. Extensive and intensive alternations are not essential criteria, although it may be generally stated that more density goes hand-in-hand with more fuzziness. As such, I think that the differentiation holds between code-switching and code-mixing, where code-switching has local relevance and code-mixing does not.

Mixed Language Varieties of Migrant Adolescents

27

On the other hand, oppositions of this kind are always problematic as they serve as contrastive placats. The ‘higher hierachies’ of a global analysis as opposed to a local one are also understood as taking the societal level of an act into account. However, as was convincingly shown by analyses in Interactional Sociolinguistics, society always percolates through into every move and speech act of a conversation (cf. Hinnenkamp, 1989, 1990). It’s just the focus that is shifted: whereas local analysis focuses on hic-et-nunc-motives for alternations, mostly seen as a rationale of local relevance, global analysis focuses on motives for alternations which can be found in categories such as identity, group cohesion, co-membership, that is, mostly in a rationale of global relevance to the speakers involved. On the other hand, the very fact of talking, be it switching or mixing, in this particular way, can only be understood against the background of Turkish migration and the established discourse of migration. This also situates their way of speaking into the global context of migration and hybridity as an integral part of a one-world globalisation. We do not get precision and sharpness in the internal differentiation of language alternation, a language mixing à la carte as I have called it above. We thus have to make do with formally fuzzy boundaries. Skilful mixtures: Playing with languages and playing with normativity

Polylingualism as a resource for poetic language games Let us return to the world of language oscillations and their fuzziness. In the next excerpt this aspect of mixing will become clear, as we will be able to see how bilingual competence becomes the resource for language plays and even for a kind of extempore poetry. Excerpt 7 nicely displays the tension between normative consciousness and an awareness of language, and its simultaneous undermining by hybrid language use. In the following scene, three adolescents Mehmet, Uður and Kamil, aged 15 and 16, hang around in a self-service shop in their neighbourhood. They buy doughnuts and fool about. At one point Mehmet swallows a piece of his doughnut the wrong way and starts coughing, which Kamil responds to by slapping his back and ironically telling him to enjoy the meal. This prelude continues down to line 5. (E-7) Transcript ‘Gang-ster’ 1 2 3 4 5 6 7 8

Afiyet [olsun Enjoy your meal M: [((coughing)) M: Afiyetle beraber olsun All of you enjoy it U: Geber Die hard! K: Afiyet ºeker olsun Enjoy it sweet as sugar ((2 sec.)) U: Stirb langsam Die hard M: hahaha + bizde (+) kaseti açtý=’stirb langsam’ yazýyor K:

28

9 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

Bilingualism and Social Relations

hahaha + in our place he put in a cassette with the title ‘Die hard’ on it #((gradually starts to laugh))U-Uður ‘sýtýrb langsam’ okuyor hahaha# U- Uður reads ‘sýtýrb langsam’ #((Laughing continues for about 6 sec., K. and U. join in)) M: ha ha ha ha ha ha ha ha ha ha ha [ha ha ha ha ha ha ha ha ha ha ha# K: #((emerging from laughter)) [Stirb langsam (…) Die hard {sýtýrb/stirb} langsam,{Alter}# die hard old man U: {°…°} M: ((emerging from laughter)) Bak orda ne yazÏyor, Ei-gang hahaha Look, what’s written there, Ei-gang ((Eingang with a ‘n’ missing; resulting in lit. egg-walk or egg-corridor)) (+) {Ei/Ay}Gang6 (+) egg/moon walk/corridor U: Ei{n}gang K: Nerde bunun {ayý /Ei-ý} Where is its {moon /egg} M: He? Hu? ((What do you mean?)) K: Nerde bunun {ayý /Ei-ý} Where is its {moon /egg} M: ((coughing, gradually merging into laughter)) eh [eh ha ha ha K: [{Ay /Ei}-Gäng Moon/ egg-gang7 M: Doðru lan Right man U: Nerde bunun {ayý / Ei-ý} oðlum Where is its {moon / egg}, son} M: Yoa: + #((strong draw)) *{ay/Ei}-yýn- gang*# (+) {ay/Ei}{ý}gang haha ya No: + moon /egg walk (+) moon /egg walk hahah ya #((imitating American accent)) ein geyn zwei geyn#8 U: Ayýnýn Gangý (+) hýhýhý the walk of a bear K: Eingang (+) Zweigang Entrance (+) two gear9 ((0.5 sec.)) M: [ha! U: [{Weitergang /zweiter Gang} Continuation /second gear /course K: {°…°} M: Dün ne filmleri vardý? What kind of films were there yesterday? U: Dün mü? Yesterday? K: Saate baksana Just take a look at the time

Mixed Language Varieties of Migrant Adolescents

29

The episode that is of primary interest here starts after the pause of two seconds, which Uður introduces with ‘Stirb langsam’ (line 7). Uður thus comments on Mehmet’s ongoing coughing. ‘Stirb langsam’ reminds Mehmet of an episode when Uður pronounced the same title of the video film as ‘Sιtιrb langsam’ for which he gets a loud laughter in response (line 8–13). The strong reaction is probably due to the pronunciation of ‘sιtιrb’ for ‘stirb’. This epenthesis is regarded as a typical Turkish accent and is highly stigmatised.10 While still laughing about Uður’s mispronunciation Mehmet directs his friends’ attention to a sign in the shop in which they hang around. This sign originally reads ‘EINGANG’ (entrance) but the first ‘N’ had dropped leaving ‘EI GANG’ (line 15). This leads the three adolescents to a brief, fast and effective word play, which actually cannot be adequately represented by the transcript above (or by any other transcript). The sequence from line 15 to line 31 or 32 respectively is fully dedicated to the polyfunctionality and to the associations of the truncated presyllable ‘Ei’ which in German of course means ‘egg’ and which is homophonous with the Turkish noun ‘ay‘ (moon, month) or – if extended by the Turkish vowel [I] – with ‘bear’ (ayý). If the adolescents alternate languages they do it in a way which seems – as we have seen in the above examples – to have no restrictions in terms of adding Turkish suffixes to German words (not vice versa, however). This means that for the word play above all kinds of German–Turkish combinations have to be taken into account. Thus a German ‘Eigang’ (egg walk) may also be thought of as a German-Turkish ‘ay Gang’, a combination of ‘moon’ or ‘month’ with ‘Gang’ (walk, corridor, course (of a meal)). It is probably because this phrase can be understood in so many ways that Kamil asks ‘Nerede bunun ayý/Ei-ý’ (line 18 and line 20), to which Mehmet reacts with a laughter and which leads Kamil to the variant: ‘Ay/Ei Gäng’ (line 22). Kamil thus transforms ‘Gang’ into ‘Gäng’, but the German orthography does not show that Kamils pronunciation is indeed [aI gæŋ] bringing a third language, i.e. English or American, into play. At this point Uður, who has been teased before, enters the play as well (line 24), though it’s not clear whether his contribution is one of participation or one of checking. Also Mehmet, who had started the play, offers another variant (lines 25 and 26): Mehmet pronounces the complete German EINGANG now with a strong draw as if it there were three instead of two syllables. This is also an interesting parallel to Uður’s alleged epenthetic pronunciation (line 9,) in that Mehmet inserts an additional vowel between the semivowelised [y] and the reinserted [n] thus pronouncing it in a very Turkish way as *{ay/Ei}-yýn-gang*. Is this an allusion to Uður’s use of the stigmatised epenthetic form? Mehmet continues by returning to Kamil’s Anglo-American variant, caricaturing the ‘heavy accent’ of a German-speaking American: [aIn geIn svaI geIn] (line 26). The intonation pattern is roughly like that: ¯° 2°. At the same time Mehmet’s voice goes one pitch up. This variant derivates the verb ‘gehen’ from ‘Gang’. On the paradigmatic level ‘ein’ is substituted by ‘zwei’. ‘Ein gehen’ or ‘eingehen’, on which ‘ein geyn’ is based, is a proper German verb (construction) with different meanings. ‘Zwei gehen’, on the other hand, makes only sense as ‘two (persons etc.) go’. It is, however, the parallelism that counts which Kamil reconverts into the nominal forms ‘Eingang (+) Zweigang’ (line 28). Uður enters the game now with a new version in that he brings the ‘bear’ (ayý)

Bilingualism and Social Relations

30

into play (line 27) using a full fledged Turkish genitive construction ‘the bear its walk’ (ayý- GEN Gang-POSS. ÷ ayý-n-ýn Gang-Ï). Uður’s bear-version, however, is not elaborated upon. Rather, Uður adapts to Kamil’s ‘Eingang (+) Zweigang’ to which he adds ‘Weitergang’ or ‘zweiter Gang’ respectively (line 31). And as if ‘Weitergang’ should be taken literally, Mehmet opens up a new subject (line 33). Obviously, the language play has reached its end here. And indeed the last two contributions were fully in German, presenting real existing words, far enough away from the initial word. In Table 1, the word play, which lasted only a few seconds is represented in a kind of overview just by enumerating its various stages.

(E-8) Table 1 ‘Gang-ster’ (line 15)

Ei gang

(line 16)

{Ei/Ay}Gang

(line 17)

Ei{n}gang

(line 18)

{ayý / Ei-ý}

(line 20)

{ayý / Ei-ý}

(line 22)

{Ay/Ei-Gäng / Ei-Gäng} ((engl.? [aw gæõ]))

(line 24)

{ayý / Ei-ý}

(line 25)

*ay yýn gang*

(line 26)

ay{ý}gang

(line 24)

#((amer. accent)) ein geyn zwei geyn# (([awn gewn svaw gewn]; ¯° 2°))

(line 27)

ayýnýn Gang-ý

(line 28)

Eingang (+) Zweigang

(line 31)

{Weitergang /zweiter Gang}

It is quite normal for children and adolescents to play with language, testing it and turning words upside down. That Mehmet, Kamil and Uður do this in two languages, that they extract and exploit the language material and the ambiguities to play with from two languages is certainly the privilege of bilinguals. Mehmet, Kamil and Uður go to secondary school. Their educational record is not brilliant. Pupils like them are very often regarded as semilingual or as defective bilinguals – these are the terms used within the school’s institutional discourse. However, we also find a high degree of language awareness expressed, for example, through the episode of stigmatising Uður’s epenthetic pronunciation or the caricaturing of the American accent, not forgetting the missing ‘n’ of ‘Eingang’ that served as the immediate cause of the play. All this reflects their normative awareness of language. Part of the play is on the word derivations, conversions, paradigmatic substitutions, parallelisms and continuous ambiguities that are borne out of the two languages’ in-betweenness, thus fully exploiting the potentials of bilingualism.

Mixed Language Varieties of Migrant Adolescents

31

Playing in-between the two (and sometimes more) languages (or codes) is not infrequent. Mehmet and Kamil do it quite often, as do other adolescent bilinguals. At one point there is a kind of extempore composition triggered by talking about a football coach named ‘Wolfgang’, which is presented in Table 2 (without the context in which it is embedded).

(E-9) Table 2 ‘Wolfgang’ Wolfgang adý Wolfgang [Wolfgang his name is Wolfgang] Wolfgang Wolf’un oðlu Molf [Wolfgang wolf his son molf] Wolfgang Wolf’un oðlu Molfgang [Wolfgang wolf his son Molfgang] 11

Wolfgang Wolf’un oðlu in Wolfsburg [Wolfgang wolf his son in Wolfsburg]

Adam dreimal Wolf oldu Doppelwolf [The man was three times wolf doublewolf] Ama Wolfsburg’da oynuyor [But he plays in Wolfsburg] Wolfgang oynuyor ama wo wo [Wolfgang plays but where where]

Besides the alliterative play with the letter ‘o’ it is also the bilabial initial sound ‘m’ (Wolf’un oðlu molf) which is conspicuous here. This ‘m’- alternation is a typical Turkish etcetera-form: cf. Hasan Masan meaning Hasan and his friends – a pattern used here in an expressive-poetic function. To give another, less extensive, example: at one point, whilst talking about fights between football fans, the adolescents use an imaginary gun and rhythmically shoot around, which is supported by a Turkish counting-out rhyme which is further accompanied by clapping hands: ‘Bir sana bi hava / bir sana bi hava‘ (‘One for you, one into the air / one for you, one into the air’), which then was altered to ‘Hava Ana‘, to ‘Mother Eve’, which then profanely ends as a German ‘Havanna Zigarre’ (Havana cigar). This kind of language performance in two languages often displays extempore poetry. In doing so, the performers not only make use of their bilingualism but also exploit its possibilities for boundary crossing by fusing and blending words, as well as expressive mechanisms (such as the bilabial etcetera-marker). The bilingual language players thus display quite a high normative awareness of language and its potentials, even of word order processes. In a situation where standards of bilingual language use and linguistic creativity were assessed, speakers such as these would be thought to have a high level of linguistic reflection and consciousness. Besides vernacular and dialectal elements, stylised elements of Gastarbeiterdeutsch (guest worker German, immigrants’ Pidgin-German) are integrated into the performances. Stylised speech has been well researched by Rampton as one of the ‘crossing’ phenomena in language (‘stylized Asian English’, cf. Rampton, 1995). Stylised German is the – mostly exaggerated – imitation of the first generation migrants’ accent when speaking German. Uður’s alleged pronunciation of ‘stirb’ as ‘sýtýrb’ in (E-7) is a typical instance of this accent.

32

Bilingualism and Social Relations

Stylisations as mimicry The use of Gastarbeiterdeutsch, or marked elements of it (emblems), as part of the mixed language repertoire plays an important role in itself. As a matter of fact, it functions as a kind of intertextual quotation. On the one hand, this stylised variety is a copy or quotation of a particular language variety as it was, and still is, spoken by the parent generation of migrants; on the other hand, it is also an imitation of majority society’s ascriptions vis-à-vis migrants in general (as is used in foreigner talk, in caricature, comedy shows and elsewhere). During the 40 years of postwar immigration to Germany (at least since its visibility has called for political measures) proper command of German has become the tertium comparationis for integration as defined by majority society, and non-compliance to this demand can at any time be made the rationale for distinction and discrimination (cf. Hinnenkamp, 1980, 1990). When using a mixed language, however, it is no more than a quotation and a stylistic ingredient to play with. At the same time, and it is important to emphasise this, it is not their authentic language as a category of speakers but represents re-appropriation by mimicry. This multiple role is nicely exemplified in Excerpt 10. Mehmet (Me) is speaking again, this time sitting together with a friend in his room listening to techno music. Mehmet’s little niece and nephew are playing on the floor. In the background one can sometimes hear the voice of Mehmet’s mother. When she enters the room the following minor dialogue develops between mother (Mo) and son: (E-10) Transcript ‘Wie geht’s (How are you?) 1 2 3

4 5 6

Mo: ((calling her grandchild)): NEREDESN GI:::Z? Where are you, girl? ((1 sec.)) Mo: ((enters the room; directed towards her son)): WIE GE::::ST? How are you? Me: NIX GU:AT No good Mo: NIX GU:AT? No good? Mo: ((breathes deeply and picks up the child)) °hopala°

As we can see, Mehmet’s mother does not approach her son in Turkish but in a loud and extremely exaggerated Gastarbeiterdeutsch way of speaking. Asking ‘WIE GE::::ST?’ is certainly not meant as a kind of welcoming remark towards her son or his friend. Both have been there all afternoon and have been in frequent contact with her. The question is not only marked by the metathesis of ‘TS’ to ‘ST’ (the Standard German form is ‘Wie geht’s’) but in particular on the prosodic level by its loudness, the vowel lengthening and the high pitched voice. Mehmet responds in the same extreme and exaggerated way. ‘NIX GU:AT’ is an apt and adequate answer in tone and voice, and furthermore diphthongises the German [u:] in ‘gut’ (good) into ‘guat’, which is an exaggerated form of Bavarian dialect. Furthermore it contains the highly stigmatised negation particle ‘nix’. ‘Nix’ is the passepartout negator in Gastarbeiterdeutsch and is also Bavarian. Phrases like ‘nix

Mixed Language Varieties of Migrant Adolescents

33

verstehen’ (no understand) and ‘nix deutsch sprechen’ (no speak German) are caricature classics when ridiculing migrant speech. His mother’s checking question terminates the sequence. She picks up the child and leaves the room. The isolation, the displaced topic and the conditional irrelevance of this threeturn-sequence strongly suggest a metaphorical intertextual language play between mother and son, in which the stylised form of Gastarbeiterdeutsch comes to fruition. This variety is omnipresent. Its figurative use even releases ritual clichés from rituality and topics from their thematic boundedness. Its function is purely phatic: a We that reassures itself of its own identity via an exaggerated and caricatured use of voices that are not their own (anymore) but which become re-appropriated in play, this time, however, stripped of any threatening connotations.

Discussion Based on the self-characterisations of the migrant adolescent users of mixed language, I want to describe in particular its place in migratory history and immigrant society, and to locate it within the wider societal context of migratory discourse. ‘Speaking mixed’ – an active discourse in its own right In the interviews which I conducted with the adolescents who switch and mix codes in the described ways, they described their way of talking as gemischt sprechen or karýºýk konuºmak (both meaning ‘speaking mixed’); some limit their description to ‘speaking half-German half-Turkish’. My informants in Augsburg refer to it quite metaphorically as yarým yamalak konuºmak which could be glossed as ‘patchwork speaking’. But whichever way they label their way of speaking, there are two things that are crucial: (1) They have given this specific way of speaking an autonomous name. They thus distinguish between this and other varieties of language and even other languages. An internal differentiation according to the degree of switching and mixing does not exist. (2) The characterisation of this variety by its users always expresses an activity. It is not named by a noun such as ‘mixed language’ or ‘patchwork language’ but is always combined with verba dicendi- formulations such as German sprechen and reden or Turkish konuêmak (all meaning ‘to speak’). That is, when they are ‘speaking mixed’ they are doing something very active.12 The relevance of this kind of activistic self-reference becomes clear in the background of how the language of their parents (and sometimes grandparents) was labelled. They spoke Gastarbeiterdeutsch. This characterisation of the language variety spoken by immigrants has even entered Hadumod Bußmann’s German ‘Lexikon der Sprachwissenschaft’ (Linguistic Dictionary), where we find the following entry: ‘Gastarbeiterdeutsch is a pidgin variant which developed in Germany since the 1960s and 1970s and is characterized by paratactic sentence patterns, limited lexicon, little redundancy, deletion of article, preposition, conjunction and verbal inflection. All these features are generally occurring irrespective of the speaker’s native language’ (Bußmann, 1983: 157; 1990: 262f., my

34

Bilingualism and Social Relations

translation). The name Gastarbeiterdeutsch is rooted not in the users themselves but in other-characterisation by, for example, wider society, the media and linguists. As we learn from the above entry it was not conspicuous because of its genesis in emergency multilingual language situations or in its supportive function in untutored processes of making oneself understood. Only its deficits are focused on here. The migrants’ native languages have to be remembered as well: they did not play a prominent role in the linguistic discussion. Even if they did, they were of interest only insofar as they were responsible for learning inhibitions (negative transfers etc.). Also, the attempts of the next generation, called Gastarbeiter- or Ausländerkinder (migrant children, foreigner children), to get along in two languages (their parents’ language and the German varieties) led to them being only too often denounced as ‘semilingual’ (cf. Hinnenkamp, 1990, 2000b). None of these labels originated from the immigrant community themselves, but they were names given to them by majority society. In contrast to this, the generation of ‘speaking mixed’ has given this name to their language (variety) themselves, without being labelled so by others. Majority society, furthermore, is no longer the direct addressee of this language; all it provides is overhearers. When speaking to majority language monolinguals the speakers who mix have German at their disposal; with Turkish-speaking people they have Turkish, and among themselves the mixed variety. These are the general options. The patterns are ‘monolinguality’, ‘bilinguality’ or ‘mixed language’. Of course, reality does not fit such a clear-cut picture. It would be naïve to pretend that the options or choices selected in actual discourse follow a pattern. We find imbalances in language dominance, have to deal with incompetence and with emergency solutions in order to reach a communicative goal. The ways in which bilingual speakers deal with imbalances can be seen when within a group the more versatile speaker adjusts vis-à-vis a speaker who is less fluent in switching languages, by converging towards ones partner’s preferred or stronger language, for example. Another recipient design in this respect is the doubling principle, repeating or paraphrasing the utterance (or part of it) in the other language. All of this can be found in the accident-account (E-4) and (E-5), where R repeatedly adjusts his mixing strategies towards his linguistically less versatile partner Y.

Gemischt sprechen and identity ‘Speaking mixed’ is not simply one language option among many – it is also an expression of a particular identity within the migratory process. The adolescents who grow up under such polycultural and multilingual conditions are confronted with contradictory ideas of which kind of linguistic and cultural conduct are apt. If we fall back upon another of those often cited essentialisms within the migratory discourse, we might say that these adolescents’ use of a ‘split language’ is analogous to the way in which they possess a ‘split identity’. Hence, their mixed language can be seen as just another expression of their confusion between two languages, two cultures and two socialisations. This, of course, is extremely simplified if not wrong. It portrays language, culture and identity in a one-to-one relationship and operates on a basis of rigid essentialist concepts. Identity is regarded as a fixed and ready-made entity, like a suit which either fits or does not fit. The constitution of identity (or identities, I should say) is

Mixed Language Varieties of Migrant Adolescents

35

a permanent process and communication plays an important part in it. One does not have a single identity but one operates, interacts and struggles with different identities. Its formation is a continuous debate with other people, with different social and societal demands (cf. Antaki & Widdicombe, 1998). These identities are also borne out of boundary marking against majority society or against one’s parent generation. We find them transformed into ‘acts of identity’ (Le Page & Tabouret- Keller, 1985), in communicative acts, in which an inventory of categories is used: how, for example, to deal with one’s own and with others’ typifications and ascriptions; how to claim, to confirm or mark off membership and affiliations; how to include oneself in and exclude oneself from groups and communities. Acts of identity make these and other categories of localising oneself individually or socially relevant. One’s localisation is neither free of contradictions nor permanent and stable. It implies a permanent struggle between a chosen self and ascribed identities. Gemischt sprechen is not only an expression of a transitional social identity. It does not simply juxtapose elements of different languages, but blends them, creates new compositions, hybrid forms and fills up a semantic room that was hitherto unoccupied and undefined. It mirrors an autonomous approach by way of language alternation, language mixing and appropriations from both linguistic communities and both ‘cultures’ (if we allow this simplification for a moment). In this sense gemischt sprechen represents an autonomous hybridolectal We-code. That is, code alternations do not correspond to alternations of metaphorical We versus They-affiliations along the lines of Ingroup–Outgroup/WeCode–They-Code (cf. Gumperz, 1982) but represent a We-code in its own right (see also Hinnenkamp, 2000a; Sebba & Wootton, 1998). Hybridity, hybridolect and gemischt sprechen Hybridity, hybridisation and hybridolect are terms to characterise the process of language mixing and code-oscillation as described and analysed above. The terminology itself is biologistic (cf. e.g. Whinnom’s, 1971, approach to contact languages). The new hybridity discourse within the humanities, however, has transmuted this concept into a critique of essentialist, heteronomous and dependential positions in theories of literature, culture, sociology and sociolinguistics. Popular theorems of colonial dependency and modernisation, as well as the installation of a concept of multiculturalism that is based on the fixation of oppositions in language, culture, identity and ethnicity, have been challenged by the deconstructive work of the hybridity discourse.13 In this respect it is only too obvious that the hybridity discourse also presents an apt frame for locating the discussion of the global motives for migrant adolescents’ language mixing. To use the notion of hybridity within the text sciences has a long tradition (cf. Bakhtin, 1981). In recent times it has particularly been established in the so-called postcolonial debate on literature and culture theory (cf. among many others García Canclini, 1992; Pratt, 1992; Young, 1995). Hybridity emphasises the – sometimes unexpected – blending of linguistic and cultural systems. Its main focus is oriented towards the reactive development of new linguistic, cultural and identitary forms in conflict with majority society, hegemony and (ex) colonial

Bilingualism and Social Relations

36

society. One breeding ground for such reactive forms is the urban multicultural society. The forms of language, culture and identity that have been developing there fit neither into the image of ethnic minorities in terms of ethnocultural folklore nor in terms of an integrative acculturation.

Conclusion The adolescent actors who grow up in the polycultural and polylinguistic space of urban migration centres develop specific transitory forms and creations out of the pool of codes at their disposal. This bricolage-argument is well known. The bilingual and bilingually mixed conversations of migrant adolescents are not just code-switching in the sense of juxtaposing rules out of two languages and their local indexical functions. What we get is an autonomous hybrid code, oscillating between two languages, representing both languages, and, at the same time constituting something third, which they call gemischt sprechen, karýºýk konuºmak, yarÏm yamalak konuºmak, kauderzanca and the like – a linguistic code in its own right (also cf. Meeuwis & Blommaert, 1998; Swigart, 1992). This mixed language functions like a mirror of the historical, social, cultural and linguistic conditions under which these adolescents grow up. Historically, the code is the critical response to the majority society’s demands for integration. The demand is to be in good command of German and at the same time to be allowed to preserve one’s Turkishness as an ‘ethnic identity’, which of course will then remain the last resort of discrimination and segregation. Sociolinguistically the adolescents react with an ingroup-language, a We-Code, which implies both deficit and competence, but first and foremost, however, difference and autonomy. The latter lead to exclusion, of the parent generation on the one hand, and of majority society on the other hand. Both, however, become re-integrated into an autonomous code which is made up by the ‘donating languages’ – to adapt a word from the beginnings of creole studies – but which are also distorted, caricatured and reinterpreted. At this point, the quote from Kobena Mercer (1988) that begins this paper will speak for itself. Across a whole range of cultural forms there is a ‘syncretic’ dynamic which critically appropriates elements from the mastercodes of the dominant culture and ‘creolises’ them, disarticulating given signs and re- articulating their symbolic meaning otherwise. The subversive force of this hybridising tendency is most apparent at the level of language itself where creoles, patois and Black English decentre, destabilise and carnivalise the linguistic domination of ‘English’. (Mercer, 1988: 57) Mercer wrote this with respect to ‘black film practice’. In a more general sense this pertains to all ‘mastercodes’, so that English could be substituted by other dominant languages or hegemonial codes. The hybridolect as used by the Turkish adolescents thus represents a kind of feedback effect to the ‘mastercode’, it constitutes a re-appropriation and re-contextualisation of a discourse that so far had been defined solely by others. In this respect the adolescents’ mixed code with all its implications is part and parcel of the hybridity discourse.

Mixed Language Varieties of Migrant Adolescents

37

Acknowledgements I am grateful to Katherina Meng, Carol Pfaff and Normann Jørgensen for their critical and valuable comments. As usual, all responsibility is solely mine. The data that form the basis of my contribution were collected during a research project that was financially supported by the University of Augsburg. Correspondence Any correspondence should be directed to Dr Volker Hinnenkamp, Fulda University of Applied Sciences, Intercultural Communication and European Studies (ICEUS), D-36039 Fulda, Germany ([email protected] .de) Notes 1. A word has to be said about ‘migrant adolescents’ in the German context (‘Migrantenjugendliche’). The adolescents whose language will be focused on in this essay are certainly no migrants themselves. The level of the German discussion, however, is that in the public discourse adolescents with an ethnic background (Turkish, Greek, Croatian, etc.) are still regarded (also by law) as ‘Ausländer’ (‘foreigners’) and as ‘Ausländerjugendliche’ (‘young/adolescent foreigners’ or ‘second’ or ‘third generation foreigners’). Regarding them as ‘immigrants’ or migrants is a progress that acknowledges at least their status being borne out of the migratory context (of their parents or grandparents). Furthermore, these adolescents have so far been mainly the object of this discourse, hardly the subject of it. This relationship is also reflected in the notion of ‘migrant’. Change in perception and acknowledgement comes slowly (cf. Terkessidis, 2000). It’s mainly the articulations of ‘migrant adolescents’ themselves that promote this change. This essay is dedicated to the further de-objectification of the migratory discourse and to the respect of autonomous forms of expression. 2. Special thanks to Tuna Döger and Ahmet Atasever for their support in getting the data. Most of the data were gathered in informal situations by one of the adolescents themselves, mostly during spare time activities. Not all participants of a taped conversation were thus informed beforehand. They were then asked afterwards if they agreed in using their data. They all agreed. Furthermore, they were very enthusiastic that interest was shown vis-à- vis their language. The participants were also asked for additional information about some personal data. With some of the adolescents, interviews were made about their usage of mixed language. 3. See my comment in Note 1 above. 4. Here we have to ask if H and E are still talking about the same subject: H is asking for a place which at the same time he rejects as already known. In this respect we have to deal with a misframing of contexts. 5. Particularly salient are periphrastic German-Turkish verb constructions with ‘yapmak‘ or ‘etmek‘ (to make, to do), which have been researched by other authors as well (e.g. Backus, 1992, 1996; Hayasi, 1999; Pfaff, 2000). Cf. instances such as (i) Ötekini nach dem ‘welches’ einsetzen yaparsan. [If you inserted the other after the ‘welches’ (which)] (ii) Langzeitig denken yapyoruz. [We think in long terms] As we can see from these examples this construction requires the harmonisation of two grammars: in both cases the pronominal drop of ‘Du’ (you, tu) in (i) and ‘wir’ (we) in (ii) requires the continuation in the PRO DROP-language Turkish. 6. This is a genuine Turkish genitive construction: bu-GENITIVE SUFFIX ay/Ei-POSSESSIVE SUFFIX. 7. K’s pronunciation here is different: the German Umlaut ‘ä’ signifies that he says it with an English pronunciation, so the equivalent translation would be ‘gang’.

38

Bilingualism and Social Relations

8. This construction is actually not translatable into English: The German verb ‘eingehen’ means primarily ‘to die, to decay’ and in another sense also ‘to enter’ (in relation to ‘Eingang’ – ‘entrance’). Of course, the German prefix ‘ein-’ is identical to the number/indefinite article ‘ein’ (one, a). Thus, ‘zwei gehen’ with ‘zwei’ meaning ‘two’ is a parallel construction to ‘ein gehen’. The meaning of this two verb construction is manifold: e.g. in guestworker pidgin ‘one walk two walk’. 9. This is the parallel noun construction to line 26. However, the nouns do not correspond to the verbs. ‘Weitergang’ translated as ‘continuation’ could also be written ‘weiter Gang’ meaning ‘wide corridor’. 10. Uður’s alleged realisation of ‘stirb’ does not only allude to the highly stigmatised pronunciation of Turk’s ‘Gastarbeiterdeutsch’ (‘guest workers’/immigrants’ German) of inserting vowels between consonant clusters as a transfer from Turkish which has many more restrictions on consonant clusters than German. Secondly, the German [i] is furthermore relaxed into a Turkish centralised [I]. Thirdly, German dialectical features are lost. It also remains unclear, if alone Ugur’s defective pronunciation is responsible for the laughter or if this is also due to a particular role constellation within the group. 11. ‘Wolfsburg’ is the German town of the Volkswagen motor works. 12. In the critical and antiethnicist movement of migrants other labels are used as well which partly have become popularised by the books of Feridun Zaimoglus (Zaimoglus, 1995, 1997, 1998). ‘Kanak Sprak’ based on the xenophobe invective ‘Kanake’ is one such label which formally and semantically has a wider extension than gemischt sprechen (cf. Zaimoglu’s preface in ‘Kanak Sprak’; also cf. Pfaff, 2003). ‘Kanak Sprak’ reflects and absorbs the negative ascriptions as much as it is an expression of new self-confidence and identity. But as it consists mainly of stylised and jargonised forms (and is not bilingual) it becomes easily majorised by non-ethnic jargon and comedy shows (cf. Keim & Androutsopoulos, 2000; Füglein, 2000). 13. On some earlier and highly enlightening treatises on this subject matter, cf. for example Hewitt, 1992; Le Page & Tabouret-Keller, 1982; see also Rampton’s ethnographic work on language crossing (1995). Another aspect is the critique of institutionalised folklorisation of cultural and ethnic difference. Young (1995: 5) comments on this: ‘the doctrine of multiculturalism encourages different groups to reify their individual and different identities at their most different, thus (…) encouraging extremist groups, who have become ‘representative’ because they have the most clearly discernibly different identity’.

References Androutsopoulos, J. and Hinnenkamp, V. (2001) Code-Switching in der bilingualen Chat-Kommunikation: ein explorativer Blick auf #hellas und #turks. In M. Beißwenger (ed.) Chatkommunikation. Sprache, Interaktion, Sozialität & Identität in synchroner computervermittelter Kommunikation (pp. 367–401). Stuttgart: Ibidem. Antaki, C. and Widdicombe, S. (1998) Identity as an achievement and as a tool. In C. Antaki and S. Widdicombe (eds) Identity in Talk (pp. 1–14). London: Sage. Auer, P. (1988) A conversation analytic approach to codeswitching and transfer. In M. Heller (ed.) Codeswitching. Anthropological and Sociolinguistic Perspectives (pp. 187–213). Berlin/New York/Amsterdam: Mouton. Auer, P. (ed.) (1998a) Code-Switching in Conversation. Language, Interaction and Identity. London/New York: Routledge. Auer, P. (1998b) Introduction: Bilingual Conversation revisited. In Code-Switching in Conversation. Language, Interaction and Identity (pp. 1–24). London/New York: Routledge. Auer, P. (1998c) From code-switching via language mixing to fused lects: Toward a dynamic typology of bilingual speech. InLiSt [Interaction and Linguistic Structures] 6. Freiburg im Breisgau. Backus, A. (1992) Patterns of Language Mixing. A Study in Turkish-Dutch Bilingualism. Wiesbaden: Otto Harrasowitz.

Mixed Language Varieties of Migrant Adolescents

39

Backus, A. (1996) Two in One. Bilingual Speech of Turkish Immigrants in the Netherlands. Tilburg: Tilburg University Press. Bakhtin, M.M. (1981) (orig. 1935) The Dialogic Imagination. Four essays. Austin: Texas University Press. Bußmann, H. (1983/1990) Lexikon der Sprachwissenschaft. Stuttgart: Kröner. Clarke, J. (1976) Style. In S. Hall and A. Jefferson (eds) Resistance through Rituals. Youth Subcultures in Post-war Britain. London: Hutchinson. Eastman, C. (ed.) (1992) Codeswitching. Journal of Multilingual and Multicultural Development 13 (1/2). Füglein, R. (2001) Kanak Sprak. Eine ethnolinguistische Untersuchung eines Sprachphänomens im Deutschen. Unpublished paper, University of Bamberg. García Canclini, N. (1992) Culturas Híbridas. Estrategias para entrar y salir de la modernidad. Buenos Aires: Editorial Sudamericana. Gumperz, J.J. (1982) Discourse Strategies. Cambridge: Oxford University Press. Hayasi, T. (1999) Insertion of German verbal elements into Turkish discourse by Turkish-German bilingual speakers in Berlin. Tokyo University Linguistic Papers (TULIP) 18, 215–230 (in Japanese, summary in English). Hewitt, R. (1992) Language, youth and the destabilisation of ethnicity. In C. Palmgren, K. Lövgren and B. Göran (eds) Ethnicity in Youth Culture (pp. 27–42). Stockholm: Youth Culture at Stockholm University. Hinnenkamp, V. (1980) The refusal of second language learning in interethnic context. In H. Giles, P. Robinson and P.H. Smith (eds) Language: Social Psychological Perspectives (pp. 179–184. Oxford & New York: Pergamon. Hinnenkamp, V. (1989) Interaktionale Soziolinguistik und Interkulturelle Kommunikation. Gesprächsmanagement zwischen Deutschen und Türken. Tübingen: Niemeyer. Hinnenkamp, V. (1990) ‘Gastarbeiterlinguistik’ und die Ethnisierung der Gastarbeiter. In E. Dittrich and F.-O. Radtke (eds) Ethnizität. Wissenschaft und Minderheiten (pp. 277–298). Opladen: Westdeutscher Verlag. Hinnenkamp, V. (2000a) ‘Gemischt sprechen’ von Migrantenjugendlichen als Ausdruck ihrer Identität. Der Deutschunterricht 5, 96–107. Hinnenkamp, V. (2000b) La comunicazione interculturale. Aspetti dei suoi sviluppi in Germania. In F. Ricci-Garotti and M. Rosanelli (eds) Programmi di scambio con i paesi di lingua tedesca e dimensione interculturale (pp. 128–143). Milano: Franco Angeli. Hinnenkamp, V. and Meng, K. (eds) (2003) Sprachgrenzen überspringen. Sprachliche Hybridität und polykulturelles Selbstverständnis. Tübingen: Narr. Kallmeyer, W., TandoÈan-Weidenhammer, D. and Keim, I. (2000) Deutsch-Türkisches. Sprache und kommunikativer Stil von Migranten. Sprachreport 3, 2–8. Keim, I. and Androutsopoulos, J. (2000) Hey Lan, isch geb dir konkret Handy. Deutsch-türkische Mischsprache und Deutsch mit ausländischem Akzent: Wie Sprechweisen der Straße durch die Medien populär werden. Frankfurter Allgemeine Zeitung (FAZ), 26 January. Le Page, R.B. and Tabouret-Keller, A. (1982) Models and stereotypes of ethnicity and of language. In Journal of Multilingual and Multicultural Development 3 (3), 161–192. Le Page, R.B. and Tabouret-Keller, A. (1985) Acts of Identity. Cambridge: Cambridge University Press. Meeuwis, M. and Blommaert, J. (1998) A monolectal view of code-switching: Layered code-switching among Zairians in Belgium. In P. Auer (ed.) Code-Switching in Conversation. Language, Interaction and Identity (pp. 76–100). London/New York: Routledge. Mercer, K. (1988) Diaspora cultures and the dialogic imagination: The aesthetics of Black independent film in Britain. In M.B. Cham and C. Andrade-Watkins (eds) Blackframes: Critical Perspectives on Black Independent Cinema (pp. 50–61). Cambridge, MA: MIT Press. Pfaff, C.W. (2000) Development and use of et- and yap- by Turkish/German bilingual children. In A. Göksel and C. Kerslake (eds) Studies on Turkish and Turkic Languages (pp. 365–373). Wiesbaden: Otto Harrasowitz.

40

Bilingualism and Social Relations

Pfaff, C.W. (2003) ‘Kanaken in Alemannistan’: Feridun Zaimoglu’s representation of migrant language. In Sprachgrenzen überspringen. Sprachliche Hybridität und polykulturelles Selbstverständnis. Tübingen: Narr. Pratt, M.L. (1992) Imperial Eyes. Travel Writing and Transculturation. London and New York: Routledge. Rampton, B. (1995) Crossing. Language and Ethnicity among Adolescents. London and New York: Longman. Sassen, S. (1996) Migranten, Siedler, Flüchtlinge. Von der Massenauswanderung zur Festung Europa. Frankfurt am Main: Fischer. Sebba, M. and Wootton, A. (1998) We, they and identity: Sequential versus identity-related explanation in code-switching. In Code-Switching in Conversation. Language, Interaction and Identity (pp. 262–289). London/New York: Routledge. Swigart, L. (1992) Two codes or one? The Insider’s view and the description of codeswitching in Dakar. Journal of Multilingual and Multicultural Development 13 (1/2), 83–102. Terkessidis, M. (2000) Migranten. Hamburg: Europäische Verlagsanstalt/Rotbuch. Whinnom, K. (1971) Linguistic hybridization and the ‘special case’ of pidgins and creoles. In D. Hymes (ed.) The Pidginization and Creolization of Languages (pp. 91–115). London and New York: Cambridge University Press. Young, R.J.C. (1995) Hybridity and Diaspora. In R.J.C. Young Colonial Desire. Hybridity in Theory, Culture and Race (pp. 1–54). London and New York: Routledge. Zaimoglu, F. (1995) Kanak Sprak. 24 Misstöne vom Rande der Gesellschaft. Hamburg: Rotbuch. Zaimoglu, F. (1997) Abschaum. Die wahre Geschichte von Ertan Ongun. Hamburg: Rotbuch. Zaimoglu, F. (1998) Koppstoff. Kanaka Sprak vom Rande der Gesellschaft. Hamburg: Rotbuch.

Mixed Language Varieties of Migrant Adolescents

41

Appendix: Legend to Transcriptions {?kommt}

doubtable reconstruction

{fährt /Pferd}

potential alternatives of hearing and interpreting

(....)

incomprehensible (with variable extension according to length)

(( ))

commentary, e.g. ((1.5 sec.)), ((laughter))

#((Com.)) dadada#

scope of commentary

wie-

abortion of utterance

sa:gt, sa:::gt

vowel lengthening, degree of lengthening

lanngsam, dasssss

holding of consonant, according to intensity

ein

assimilation of unstressed endings such as ‘ein’ instead of ‘einen’

damit

stressed, emphasised

DAS

high volume

EdaE

low volume

BBdaBB

very low volume

*ach was*

slowly

>darüber
aklÏn durur (.) arkhadan böyle araba knick oldu abi knick totalschaden yani Brother, they crashed into the Mondeo, can you believe it? Brother, it bent. It was written off.

Fikri:

(

) mi yapÏyorlar abi ne yapÏyorlar

If they ……………., brother, or what are they doing?

Hans: Fikri:

totalschaden oldu arkadaê It was written off, my friend. bunlar simdi hep böyle yapÏyi belki böyle para alÏyolar They do it always like that, perhaps they get money like that.

Hans:

para (.) êimdik o parayÏ çÏkartÏyo abi direk veriyo belki êey leasing fabrike Money, well the money they get, perhaps they give it (the car) straight back to the leasing company.

Fikri:

tabii Sure.

Hans:

firmaya [ya: To the company, oh yeah.

Fikri: Hans:

[bunlar simdi böyle yeni araba alÏyolar pahalÏ araba They buy a new car, an expensive car. sÏfÏrdan almÏê sÏfÏrdan yepyeni ford It had zero kilometres, a brand new Ford.

Cultural Orientation and Language Use

Fikri:

51

tamam iête (.) vurduruyolar para almak için sigortadan dolandÏrÏyolar Yeah, they provoke an accident to get the money back from the insurance company. They are cheating.

Hans: Fikri:

((sighs)) bana öyle geliyo It looks like that.

Hans:

valla öyle That’s true.

((The conversation continues.)) The conversation of Hans who acquired Turkish like other young people in the neighbourhood, is monolingual Turkish. For variation he uses the Turkish language. He expresses in a variety of words that the car in question is a brand new one: ‘yeni’ (new), ‘yepyeni’ (brand new) and ‘sÏfÏrdan’ (having zero km). He integrates some German expressions into the Turkish flow of speech: ,Kennzeichen’ (registration), ‘Knick’ (bent) and ‘Totalschaden’ (total damage).With this integration he seems to follow a strategy of stressing key expressions in his speech. Example 2: Kral Musa Andreas: Greek, 15 y, m.; Musa: Turkish, 15 y, m.; Kevin: German, 15 y, m.; DoÈan: Turkish, 15 y, m.; Sami: Turkish, 15 y, m. ((Andreas, Musa, Kevin, DoÈan and Sami are talking in the park. Before the excerpt transcribed somebody was whistling a melody loudly.)) ?: eine one/a

Andreas:

eine legende von kral musa a legend of Musa the boss.

Andreas:

so’n mensch (-) kral musa Such a man (-) Musa the boss.

?:

kal musa meinst du You mean ‘stay Musa’,

?:

kal mÏ yoksa kral mÏ ‘stay’ or ‘boss’?

Andreas, Kevin, Sami and Musa: kral musa others: (hundeschänder) Dog fucker

others: ?:

kral musa gef¬ngnis

(others): ?: ?:

kral musa ( ) (geschwängert)

Prison

Pregnant

others:

kral musa

Bilingualism and Social Relations

52

?: Andreas:

( ) bekloppt (–) kral musa (–) (ist gleich) kral musa (–) gefährlich (–) kral musa (–) sein lieblingsjob (–) abstechen (–) wer ist das (–) kral musa stupid – kral musa – (is similar to) kral musa – dangerous – kral musa – his favourite job – stabbing – who is that – kral musa

((The word game is finished, after that some incomprehensible expressions follow.)) The part Turkish is playing in the realisation of this word game seems to be very limited. The necessary language competence of Turkish is minimal. But the cultural competence needed to perform this sequence is substantial. The talk of the boys follows a traditional pattern, that means that they show a well established kind of interaction. At the beginning of the transcipt the word ‘eine’ (one/a) initiates this sequence, which is continuously stressed by the word ‘kral’ (originally standard Turkish for a European king). This sequence reminds us of contemporary hip hop songs, like the ones by the band ‘Cartel’, which were bilingual in German and Turkish. The boys direct their interaction towards Musa. They ascribe to him negative traits and actions in a joking manner. At the same time they praise him in the ‘refrain’ with the word kral, which is part of the local male register. Linguistic and cultural knowledge of different backgrounds culminates here. The root of this oral tradition goes back to Turkey as well. It becomes clear that the young people initiate a ritual by which they reassure their group identity. It is obvious that in some situations language competence does not matter, in the sense of knowing a certain amount of words and structures. However getting the spirit of the interaction is more important. ‘Kral’ functions as an element of speech, constituting group identity. In the first part of our study we discovered that in some parts of Hamburg a German–Turkish variety is used among young people. Young people from Altona, Wilhelmsburg and Mümmelmannsberg reported that it is essential to have active knowledge of Turkish words and expressions, if you want to be ‘in’. They do not only use single words like ‘para’ (money) or greeting forms but also well extended question-answer-sequences, which become integrated into the German flow of speech (for example asking for time or answering the question for time). We want to introduce a further example of Turkish–German mixed talk. Example 3: ‘Meraba kÏzÏm’ Maike: German, 19 y, f.; Tanja: Libyian-German, 17 y, f.; Aischa: Afghan., 20 y, f. ((Maike visits Tanja. Aischa phones to say that she is coming as well. Shortly afterwards the bell rings.)) Maike: ( ) Tür geklingelt Tanja The door bell rang, Tanja.

Tanja:

warte mal (2.0) das ist Aischa Wait a second, that’s Aischa.

Maike:

ja (-) schon so früh? Korrekt Really, so early? Yes.

Tanja:

ha Aischa (-) meraba kÏzÏm

Cultural Orientation and Language Use

53

Hello, my dear.

Aischa:

n=aber kÏz What’s up, girl?

Tanja:

iyi misin How are you?

Aischa:

iyim I’m fine.

((Maike tells about her visit to the dentist.)) This example shows that Turkish is used even when no one of Turkish background is around. Turkish is a well-established practice in adolescent conversation in these parts of the city. It is not so important to speak ‘correctly’ (in the sense of monolingual norms) as to integrate Turkish into German in a particular way. This style of speech is not functional in terms of fulfilling all necessities of communication but as an identity-forming element in groups.

Language Strategies in a Diverse Society In our research we discovered a diversity of strategies for dealing with the multilingual world. The strategy of submission to the standard language (German) and the development of a consciousness about the appropriate choice of language in well-defined situations are two of them. According to Bourdieu’s theory of the linguistic market (Bourdieu, 1990), language acquisition is characterised not only by incorporation of linguistic elements starting in early childhood (for example, words or syntactical elements) but also by a social ability to use this competence adequately in well-defined situations. The results of part one of our research confirms this assumption. Another strategy is the learning of Turkish by young people who do not have a Turkish background at home. It goes along with creative language alternations, which young people classify as mixed language and which appears to them as a proper language. German is characterised as such a proper language. At the same time, this mixed language has to be exclusive, understood only by insiders of the group. Their speech practices are directed towards the inside and the outside of the group. We might see the two spheres as a representation of the two as purely imaginary forms (German and Turkish), which mix in a field of transition. They become ‘impure’, ‘dirty’. This leads us back to the field of the liminal. The transition can proceed in one of two directions. The ‘impure’ is resolved in ‘pure’ gestalts. The older they get, the more they develop a stable bilingual constellation, in which German and Turkish exist side by side or alternatively Turkish is driven out by German. This results in a more or less monolingual German situation. This can be observed in the high school, and Gymnasium. Another possibility is the social production of multilingual situations, in which different languages exist side by side and new forms of languages come into existence. With our work, we can show that the old identity between language and an ethnically defined group in Germany is crumbling. Research conducted in European cities over the past few years shows that it is not only in bilingual or migrant families that multilingual speech is practised. Even among children and young people of different linguistic backgrounds an intensive language crossing

54

Bilingualism and Social Relations

(Rampton, 1995) is practised in everyday life. Children and young people acquire and use elements of different languages in their peer-groups (For the developments in Sweden see Kotsinas (1998), United Kingdom Rampton (1995) and Hewitt (1986, 1994), in Germany Auer & Dirim (2000, 2001); Auer and Hieronymus (1997) and Hieronymus (2000)). This indicates that multilingual language use is not restricted to particular adolescents, or to a particular age group, or to a particular sex, or to particular areas of a city. During the first part of the research into the linguistic-cultural hybrid worlds, it looked like a phenomenon of the lower classes rather than one found in the upper classes. The second part, however, showed that Turkish also has a value of prestige (Bourdieu) among the pupils of the Gymnasium which encouraged the acquisition and the use of Turkish elements. The popular understanding of mother tongue is challenged by our research. It is not only Turkish as a national language, but as a local means of communication which is located in the worlds of Hamburg (most frequently in areas like St Pauli, Altona, and Wilhelmsburg where large groups of migrants with different backgrounds live). This local variety is no longer congruent with standard Turkish in Turkey. Therefore we cannot talk of a revitalisation of Turkish (Heitmeyer et al. 1997 for example, suggests this in his misleading study of fundamentalism). In our opinion it has more to do with social creativity of local speech communities. We can speak of the joint language acquisition of young people from different backgrounds outside of school and the home. In the multilingual language use which results from such language acqusition, Turkish plays a dominant role. The two parts of our study also show that languages find their way into the language practices of the young people; for example Romanes, ‘Yugoslavs’, and Persians. We can now say that children and adolescents may jointly integrate into their own particular linguistic background a wide range of linguistic resources of their world. They use these resources in their interactions and particularly as a means to negotiate identity. Because this language does not single out specific functions, we cannot really talk of domains, in the sense of Fishman (1965). It looks more likely that multilingual language use in unofficial space fulfils the function of creating relations and positioning in the social environment. Correspondence Any correspondence should be directed to Dr nci Dirim, Kroprinzenstraße 26, D-22587 Hamburg, Germany ([email protected]).

References Auer, P. (1983) Zweisprachige Konversation. Code-Switching und Transfer bei italienischen Migrantenkindern in Konstanz. Konstanz. Auer, P. (1995) Türkisch in gemischt-kulturellen Gruppen von Jugendlichen im schulischen und außerschulischen Bereich. Hamburg. Auer, P. and Dirim, . (2000) Vorläufige Überlegungen zur Verwendung des Türkischen in gemischtethnischen Jugendlichengruppen in Hamburg. In I. Gogolin and B. Nauck (eds) Pluralität, gesellschaftliche Differenzierung und Bildung (pp. 97–113). Opladen: Leske & Budrich. Auer, P. and Dirim, . (2001) On the use of Turkish routines by adolescents of non-Turkish descent in Hamburg. In A. Holmen and J.N. Jørgensen (eds) Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children in North Western Europe. Copenhagen Studies in Bilingualism (pp. 157–194). The Køge Series, Vol. 7. Copenhagen: Danish University of Education.

Cultural Orientation and Language Use

55

Auer, P. and Hieronymus, A. (1997) Das versteckte Prestige des Türkischen. In Kreisjugendring München Stadt, Dokumentation, Multikulturalität in den Metropolen (pp. 77–88). München. Bourdieu, P. (1990) Was heißt sprechen? Die Ökonomie des sprachlichen Tauschs. Wien: Drammüller. Fishman, J. (1965) Who speaks what language to whom and when? La Linguistique 2, 67–88. Geertz, C. (1994) Dichte Beschreibung. Beiträge zum Verstehen kultureller Systeme. Frankfurt a.M. Glaser, B.G. and Strauss, A. (1979)[1967] The Discovery of Grounded Theory. Strategies for Qualitative Research. 10. Auflage. New York. Goffman, E. (1996) Stigma. Über Techniken der Bewältigung beschädigter Identität. Frankfurt a.M. Hartig, M. and Kurz, U. (1971) Sprache als soziale Kontrolle. Neue Ansätze zur Soziolinguistik. Frankfurt a.M. Haust, D. (1993) Formen und Funktionen des Code-Switching. Linguistische Berichte 144, Opladen. Heitmeyer, W., Müller, J. and Schröder, H. (1997) Verlockender Fundamentalismus. Frankfurt a.M. Hewitt, R. (1986) White Talk Black Talk. Cambridge: Cambridge University Press. Hewitt, R. (1994) Sprache, Adoleszenz und die Destabilisierung von Ethnizität. Deutsch lernen 19 (4), 362–376. Hieronymus, A. (2000) Ibo lan, das ist der kral! Qualitativ-heuristische Explorationen in vielsprachige Lebenswelten. P u b l i s he d at : ht t p ://w w w .s ub .uni-hamb urg.de /disse/228/index.html Kleining, G. (1994) Qualitativ-heuristische Sozialforschung. Hamburg. Kotsinas, U.-B. (1998) Language contact in Rinkeby, an immigrant suburb. In J.K. Androutsopoulos and A. Scholz (eds) Jugendsprache (pp. 125–148). Frankfurt a.M.: Lang. Rampton, B. (1995) Crossing: Language and Ethnicity Among Adolescents. Harlow. Schütz, A. and Luckmann, T. (1994) Strukturen der Lebenswelt Band 1. Frankfurt a.M. Searle, J.R. (1983) Sprechakte. Ein sprachphilosophischer Essay. Frankfurt a.M. Selting, M. et al. (1998) Gesprächsanalytisches Transkriptionssystem (GAT). Linguistische Berichte 173, 91–122. Stoll, R. (1994) Ausländische Arbeitnehmer und Arbeitnehmerinnen. In Literaturdokumentation zur Arbeitsmarkt-und Berufsforschung, Nürnberg, Nr. S 3, 7 Turner, V. (1995) Vom Ritual zum Theater. Der Ernst des menschlichen Spiels. Frankfurt a.M. Wittgenstein, L. (1990) Tractatus logico-philosophicus. Philosopische Untersuchungen. Leipzig.

Appendix The interviews of part one have been transcribed verbatim. All names are fictitious. For rules of transcription, see Selting et al. (1998). Font: large: original expressions small: translation German and English bold: Turkish parts of conversation ((xxx)) non-verbal interaction <xxx> the conversation accompanying interaction = immediate follow-up, unclear talk (xxx) reconstructed meaning ( ) incomprehensible passage (.) micropause (-) (- - ) (- - -) short, middle, long pause (2.7) measured pause ? rising intonation acceleration

The Creation and Administration of Social Relations in Bilingual Group Work Jakob Cromdal Department of Child Studies, Linköping University, S-58183 Linköping, Sweden While recent studies of task-focused activities in bilingual groups of students focus on the social nature of knowledge construction as part of the (language) learning process, they sometimes leave out the role of such bilingual practices as language choice and alternation in the interactional accomplishment of social relations. The present study highlights the bilingual aspects of social interaction in a group of four Danish–Turkish students engaged in creating a cartoon strip. In-depth analysis of a 45-minute session reveals some organisational features of the group’s work. First, the construction of the cartoon is informed by a storyline, which is narratively produced throughout the best part of the session. Second, the narration of the storyline is in turn informed by a linguistic division of labour, specifying that narrative contributions are produced in Danish and leaving language choice open for other types of actions. Third, it is shown that participation in narrative activities may be exploited to regulate the group’s work, and the analysis highlights participants’ use of language choice and code-switching in forming alliances and opposing contestable actions in relation to story narration. Thus, to some extent, participation in the group’s work is asymmetrically organised, and the analysis is discussed in terms of ‘power’ as an interactionally accomplished feature of the students’ social conduct.

Introduction Research on multilingual talk in educational settings has traditionally focused on teacher discourse and, in particular, on its relation to students’ progress in mastering the new language, where progress is assessed through a variety of methods (for reviews see, e.g. Chaudron, 1988; van Lier, 1988). In contrast, recent studies within ethnographic, sociocultural and constructivist frameworks favour the view of multilingual classrooms as settings where social interaction is organised around various pedagogic activities, and where learning is both accomplished and displayed through participation in such activities (e.g. Durán & Szymanski, 1995; Tuyay et al., 1995; Willett, 1995). In short, these studies look at the mundane social practices of the participants, rather than solely at intraindividual processes of learning. One empirical outcome of this has been a shift in focus from ‘official’ talk in teacher-fronted activities to the talk constituting students’ collaborative activities in small groups. Such activities typically entail the production of written text, which may either be the sole purpose of the task, or merely serve to document the groups’ work progress, or its results. Analyses of bilingual text production have shown how students focus joint attention and participate in structuring, negotiating and commenting on the evolving text and its minute details (Tuyay et al., 1995), including the negotiation and correction of specific language use (Durán & Szymanski, 1995). These studies clearly make the point that, in focusing on joint text, students are not merely involved in solving a mental or linguistic task, which may be facilitated by collaboration. Rather the (constructivist) idea is that such shared activities inescapably involve students’ management of social relations. 56

Social Relations in Bilingual Group Work

57

But what is the role of bilingualism in the students’ social conduct? While the studies cited provide inspiring analyses of collaborative work in relation to the construction of knowledge in the peer group, we learn little about how the availability of two languages enters into these discourse processes. For instance, Tuyay et al. (1995: 89) briefly state that the students’ negotiation of language (choice) is ‘not solely a matter of academic content’, but their analysis leaves us uninformed about the relevance of language choice and switching practices for situated interaction, i.e. the very conduct through which knowledge is constructed. One particularly relevant collection of studies stems from the Køge project, which investigates longitudinally the bilingual behaviour of second generation Turkish immigrants to Denmark (see Jørgensen et al., 1991, for a presentation). For instance, examining conversations within task-oriented groups, Holmen and Jørgensen (1997) and Maegaard (1998) raise the issue of students’ development of bilingual conversational skills, highlighting the locally strategic use of language choice and alternation in the deployment and coordination of social actions. Also, combining these conversational data with student interviews, Møller (1998) discusses their practices of language choice and code-switching in terms of individual and social ethnic identities. Furthermore, Jørgensen (1998) investigates the students’ code-switching as a means of regulating and even dominating interaction within the group, a practice he labels ‘power wielding’. Finally, grounding his analysis in the minute fabric of social interaction, Steensig (2000a) provides a more detailed account of the local deployment of code-switching as one in a set of resources for the accomplishment of social relations. In the broadest sense, the study at hand attempts to shed some further light on social interaction in task-oriented groups of bilingual students. The specific purpose of the study is twofold: first, using a subset of data from the Køge corpus, the analysis highlights the students’ language choice and alternation as a means of organising a group’s task activity, which is to produce a cartoon strip. Second, it will be shown how participation in the task activity itself becomes a resource in the management, and indeed regulation of social relations. Thus, the analysis of participation in the task activity raises the question of interactional asymmetry and dominance, issues which may be properly glossed in terms of ‘power’ as an interactionally accomplished phenomenon. With these aims, in-depth analyses of talk-in-interaction will be presented, combining the task-related as well as interpersonal aspects of students’ actions, and treating the issue of language choice as an integral part of their social conduct.

Methodological Issues The present analysis draws on a collection of transcripts from the Køge project, including multiparty interactions in task-oriented activities among bilingual Turkish-Danish students at school (Turan, 1999). Specifically, the present study examines Conversation 801, which involves a group of eighth grade students engaged in the production of a cartoon strip. The group comprised three girls, Asiye, Esen and Selma, and one boy, Erol. Prior to the recording, the group was informed that their task was to produce a cartoon strip using the following materials provided by the researchers: a sheet of plain cardboard, scissors, glue sticks, marker pens, a collection of teenage

58

Bilingualism and Social Relations

magazines as well as advertising and artistic postcards. The 45-minute audio recording was accomplished by supplying each of the participants with a microphone connected to a mixer board (see Turan, 1999 for details concerning data collection procedures). The original transcripts used here were made in compliance with the CHAT conventions of CHILDES (MacWhinney, 1995). In addition, the present analysis draws on a conversation analytic transcript (e.g. Hutchby & Wooffitt, 1998; see also Appendix) of the initial 15 minutes (approximately) provided by Steensig (2000b). For reasons of uniformity, the CHAT transcriptions have been converted to CA notations in the excerpts presented in the sections to follow. Naturally, this only means that the actual symbols have been changed, not the level of transcription detail. An important resource in analysing these data was the sequential ordering of talk that played an important organisational part in the social interaction taking place within the group. This analytical standpoint is grounded in the theoretical conception of social interaction as an inherently dialogic endeavour, in which mutually recognisable interactional projects are accomplished through sequentially organised trajectories of action.

Creating a Monolingual Story in Bilingual Conversation – A Background Analysis Elsewhere, I have shown that the present group’s work with the assignment is organised as an extended, collaboratively produced narration (Cromdal, 2000). That is, all the activities involved in producing a cartoon-like strip (finding, cutting out, discussing and finally pasting newspaper and magazine clips) relate, in different ways, to the joint production of a verbal narrative. Along the way, this narration is written down under the pictures. Hence, each text strip serves as a document of a single visual image (either a single picture or a small collage of clips) rendering the image as interpretable in terms of the storyline. In Garfinkel’s (1967) words, the written storyline serves as a scheme of interpretation for each visual image. The previous analysis detailed a crucial aspect of the bilingual organisation of the group’s work, namely the division of labour between Turkish and Danish that informed the participants’ verbal conduct. In brief, whereas both languages were used for a variety of interactional purposes, the storyline itself was produced in Danish. Naturally, producing a strictly monolingual storyline in the midst of bilingual talk, and indeed, producing a continuous storyline in the midst of a variety of other interactional projects, calls for minute organisation of verbal actions. For instance, when a narrative sequence begins in a Turkish sequential environment, the use of Danish for the narrative turn sets off the storyline from other talk. In contrast, when a narrative sequence is initiated in a Danish environment, speakers tend to exploit a variety of methods, such as different temporal markers or preface units, to set off their turn from preceding talk. Moreover, in cases where several speakers produced extended narrative sequences, the preference for Danish was observed without exception. That is, none of the participants involved in co-narration would switch to Turkish within a narrative sequence.

Social Relations in Bilingual Group Work

59

Finally, the transition from storyline talk to non-narrative interaction was frequently marked by a ‘postnarrative’ switch into Turkish. In sum, the group’s inherently bilingual organisation of this assignment implies that the joint production of the storyline is normatively conducted in Danish, and Cromdal (2000) shows that a handful of deviations from this pattern can be accounted for in terms of participants’ orientation to this norm. With this task organisation as a backdrop, the remaining part of this article will highlight some important aspects of the participants’ relational work, which may also inform our general understanding of social interaction in bilingual assignment groups. Specifically, I aim to show how the jointly recognised task, to produce a coherent cartoon-like story, may be exploited for a variety of relational projects unfolding during the group’s work.

Managing Social Alignments in Task Negotiation An essential aspect of the children’s management of social relations involves their ways of establishing, sustaining and terminating local alignments in the unfolding interaction. Importantly, once the group members recognise and align with a common task, they need to work out jointly acceptable ways of contributing to the group’s work. Ideally, participants need to coordinate their actions in such a way that their contributions can be examined, evaluated and accepted by the other group members. Needless to say, such a practice invites various constellations of social alignments with respect to specific ideas and suggestions brought up by each participant during the course of the group’s work. Let us therefore consider the very first negotiation of what becomes the joint focus for the participants, to create a cartoon-like story about two young people spending the night out on the town. As it were, in this episode, the participants establish a sort of procedural scheme guiding their contributions to the story, namely that all contributions to the actual storyline be produced in a narrative format and that they be produced exclusively in Danish. The excerpt begins a few minutes into the recording, immediately after Esen has presented her idea as to how they should proceed with the task, winning at least the partial attention of the other participants (see Steensig, 2000a, for a detailed account of this sequence). Excerpt 1 [Esen has just announced that she has an idea about how to do the assignment. Simplified version of transcript from Steensig (2000b: 41–44). Original transcript in Turan (1999: 211–212).] 1

Erol:

2

Esen:

3

Erol:

4

Esen:

Coca cola var! Valla bunu birêeye kullanÏrÏz= (there’s coca cola! gosh we can use this) =dinle bir [dinle] (listen just listen) [r e h]klam (commercial) onu °çÏkart° (.)°bak° hm bak (.) hun ringer til (take it off and look hm look she calls the)

Bilingualism and Social Relations

60

5 6 7

Erol:

8 9 10

Selma:

11 12

Selma: Esen:

13

Selma:

14

Esen:

15

Selma:

16

Erol:

17

Selma:

Erol:

18 19

Erol:

20

Esen:

21

Selma:

22 23 24

Erol: Esen:

25

Selma:

26

Erol:

27

Esen:

28 29 30

Erol:

biografen for at bestille en billen så ankommer hun (theater to book a ticket then she arrives) (0.7) og så køber hun popcorn og °ser° Speed (and then she buys popcorn and watches Speed) vay yavrum Esen! og coca co[la reklame] (woaw baby Esen! and coca cola commercial) [­WU:hh!­ ] (1) [ha önde ] hh hahh hah hhh (in front) [YEA:H!] °og° hun køber (.) (and she buys) COCA COLA ay êunu bir= (COCA COLA oh this one) =(xx) ne:[j (no:) [hall[o du kan]= (hello you can) [ne:j ] (no:) =(jo også købe) (.) (v:e: øh) popcorn og (du køb) (also buy eh popcorn and you buy) coca ¯colaer= (coca colas) =JA: (EXACTLy)= (yes) =sammen med sin (.5) [kæreste] (together with her boyfriend) [kæreste] ((hands clapping)) (boyfriend) ­WU:U:! hu hihh ((clapping)) (.) hng hh [hhh [­JU U:! (nej s) den vil jeg godt (x x)= (no that one I’d like to x x) =og [bagefter skal h]un til museum (.9) ­I[kke? (and afterwards she’s going to the museum right?) [(tar) man det?] (does one take it?) [(nåja) (allright) og bagefter der var også en koncert med ham der (and afterwards there was a concert too with that) Tom Jones. (guy Tom Jones) kiz ne mahsus size. (girl what’s peculiar to you)

Social Relations in Bilingual Group Work

31

Selma:

61

okay men hvor fanden er han (okay but where the hell is he)

In the beginning of the excerpt, we can see that Erol’s enthusiastic presentation of the Coca-Cola advertisement is partly disrupted by Esen, who makes an effort to get the others’ attention (lines 2 and 4) and presents her previously announced idea, producing a short narration about a girl visiting a movie theatre (lines 4–6). We can see Erol vividly aligning with Esen’s story in line 7 (‘vay yavrum Esen!‘), his use of Turkish resulting, on the level of conversational structure, in a code-switch. Thus, Erol’s token of alignment with Esen’s idea stands as an isolated Turkish unit in the midst of talk in Danish, and one interpretation of such language choice may be that code-switching serves to enhance expression of affect (cf. Aronsson, 2000). We should also note that Erol uses this turn space to reintroduce the Coca-Cola commercial. Notably, to accomplish this he switches to Danish, appropriating Esen’s story on the level of language choice. As this move receives no immediate response, Erol elaborates his idea, ‘ha önde‘ (in front), that the commercial should go in front (in the illustration). It is difficult to pinpoint the local meaning of this switch into Turkish, but it may be that it is an effect of the lack of response from the others. Thus it would resemble Auer’s (1984b) analysis of ‘nonfirst firsts’, where turns receiving no response from their recipients are repeated in a different language. Also, but not necessarily alternatively, the switch may be used to mitigate his disruption of the story (note the laugh tokens following on line 10). However, Erol’s turn in line 10 is overlapped with Selma’s enthusiastic support for his suggestion (line 11), and again the contrastive language choice (this time of English) seems to do the sort of interactional work Aronsson (2000) describes as displaying intensified affect. In effect, what we have here is an ‘incipient alignment’ between Selma and Erol. This term was used by Aronsson to refer to a relational state where a first step has been taken towards a likely alliance between (minimally) two participants against a third party, but where the alliance has not yet been confirmed by both prospective allies (cf. Steensig, 2000a). In the present case, this begs the question of Selma’s delay in taking this first step. The issue is important for the participants’ coordination of actions, because at this moment (i.e. the beginning of turns 10 and 11), Erol seems to be acting upon the belief that he has no allies with regard to the Coca-Cola commercial. In my interpretation, the delay has to do with the sequential placement of Selma’s supportive token (to Esen’s story) in line 8, which overlaps with the very part of Erol’s utterance (line 7) to which Selma subscribes in line 11. She simply needs the extra time to consider (and respond to) Esen’s suggestion. In any case, at this point in the unfolding sequence of actions, Esen does not respond to either party. Instead, she picks up the storyline with the words ‘og hun køber’ (‘and she buys’) followed by a micropause. This provides a structural slot for further suggestions for objects that should be part of the narration, a slot that Selma exploits to insert the Coca-Cola suggestion (line 13); and again we may see Turkish being used to disrupt the projected continuation of Esen’s story. This time Esen explicitly rejects the suggestion (lines 15 and 16), and Selma

62

Bilingualism and Social Relations

upgrades the opposition initiating her turn with the token ‘hallo’ and arguing her case that if the storyline character can buy popcorn, she can buy Coca-Cola too. It might be argued here that implicit in her argument is the suggestion that popcorn and Coca-Cola go together in a movie theatre context. Such an argument may be akin to ethnomethodological work on commonsense categories that inform mundane reasoning practices (Sacks, 1992: Vol. 1). In brief, it seems likely that teenagers would recognise Coca-Cola and popcorn as a sort of ‘kit’ when it comes to snacks at the movies. Indeed, it would seem that Esen expresses precisely this recognition through his intense agreement (line 19) with Selma’s argument, thereby ratifying the alliance between the two of them. Again, the delivery of a strongly supportive action involves a code-switch, this time to English. Instead of pursuing this controversial topic further, Esen resorts to the storyline (line 20), introducing an entirely new element, namely the girl’s boyfriend (‘kæreste’). Again, this is met with strong appreciation by both Selma and Erol (lines 21 through 23), and the production of the narration continues as Selma volunteers another ‘brick’ for the storyline (a visit to the museum), which, due to no immediate response, she tries to qualify with the others through the tag-positioned agreement token ‘ikke?’(’no?’) in line 25. Upon this, Esen ratifies Selma’s contribution (‘nåja’/’alright’) and adds another story element concerning a concert with ‘ham der Tom Jones’. Selma approves to this, but points out that the papercut image of Tom Jones is temporarily lost, and a search for the picture ensues. With this action, the narration work is temporary suspended, and the children engage in various tasks such as finding the glue or inspecting other visual materials. In terms of the group’s task organisation, the result of this event is that they now have a sort of procedural agenda: any images, pictures or other visual materials they may bring to attention will be discussed, evaluated and indeed interpreted in terms of the storyline for which the grounds have been laid in the above transcript. To recap then, the analysis above has highlighted a patterned chain of local alliances with respect to several issues arising as part of the interaction. First, we have seen Esen bid for, and receive enough attention to present her idea for the story. Also, both Erol and Selma align with Esen on that issue, and with each other on the issue of Coca-Cola. However, the issue of Coca-Cola becomes a point of disagreement and mobilises Erol and Selma to collaborate in a negotiation with Esen, which ends in a ‘silent acceptance’ on Esen’s behalf. Once this potential threat to the group’s cooperation is out of the way, several other story elements are introduced, and accepted by the group without further argument. Moreover, with Selma’s narrative contribution in line 25, the storyline is now not merely collaboratively constructed in the sense that several participants present ideas to be incorporated in the story proper, but co-narrated as well. As we will see in the subsequent sections, issues of participation in narrating the story proper will lead us to consider the group’s organisation of the task in terms of interactional asymmetry and dominance.

Social Relations in Bilingual Group Work

63

Using the Storyline to Regulate Social Interaction In the above transcript, we have seen how a story about a young couple’s night on the town has been introduced to guide the group’s work with the cartoon strip, and also that the storyline is produced in Danish. In the remaining part of the analysis it will be shown how this task, to narrate a coherent storyline, serves as a resource in organising the group’s interaction. Specifically, we will consider how the creation of the storyline proper as well as activities directly related to its narration are exploited by one of the participants in particular, Esen, to impose upon the group her own version of a proper working order. Using the storyline to suppress alternative activities An important aspect of the organisational features of story narration for the interaction is that it may be deployed in such a way as to override other activities taking place within the group. The following excerpt shows how the narrative activity may be used to manage a potential conflict brewing between two of the participants. Excerpt 2 [All participants have been discussing illustrations for the movie theme. From Turan (1999: 215–216).] 1

Erol:

2 3

Selma:

4

Erol:

5

Asiye:

6

Selma:

7

Erol:

8

Selma:

9 10

(?): Erol:

11

Esen:

12 13

Erol:

14

Esen:

aha bunu bul drømmeseng para kazanmÏê lottodan (aha find this dreambed they have won money in a lottery) så bliver de rig til sidst (then they’re gonna be rich in the end) aj for fanden (no God damn it) valla oldu Esen doÈru (gosh that’s it Esen that’s right) (xxx) daha iyi olacak (xxx it’s going to be much better) hallo vi kan også bare bruge denne her (hello we could also just use this one here) hold kæft (shut up) der er en kæreste øje nej (it is a boyfriend eye no) (xxx) du skal altså holde kæft først tage det roligt (first you shut up then take it easy) vil I høre historien hun ringer til biografen for (do you wanna hear the story she calls the theater) at bestille nogen billetter (to book tickets) køber (buys) til en filmfestival så da hun ankommer der køber (for a film festival then when she arrives she buys)

Bilingualism and Social Relations

64

15

hun popcorn cola og chips (popcorn cola and potato chips)

In rough outline, Erol’s suggestion for the continuation of the storyline is blatantly rejected by Selma. Upon this rejection, he switches to Turkish while trying to solicit Esen’s alignment (line 4). It may be that the linguistic contrast arising with the code-switch serves to stress Erol’s disaffiliation with Selma’s actions altogether. In any case, whereas Esen does not ratify the alignment with Erol, Asiye seems to express her approval, and there are now at least two participants in support of his idea. Selma then makes a countermove, suggesting that they use another visual image instead, at which point Erol, now speaking Danish, plainly tells her to shut up. Disregarding that, Selma proposes that the image is the boyfriend’s eye, and now Erol elaborates his demand that she shut up and stop interfering. Thus, whereas earlier in the argument Selma rejected Erol’s specific proposal for a story element (line 3), he is now attempting to terminate her very participation in the construction of the story (lines 7 and 10). At this point, Esen begins summing up the story so far, prefacing her narration with a rhetorical question (leaving no time for an answer before proceeding with the story), which serves as a bid for an extended turn space in the subsequent interaction. In fact, the narrative activity that follows spans over eight turns involving both Erol and Selma as co-narrators, and by the time their joint narration has come to an end, their previous controversial suggestions for illustrations seem no longer relevant. In this way, the narrative activity introduced by Esen serves to resolve the conflict between Erol and Selma. This is a rather typical feature of Esen’s attempts to direct the trajectory of interaction towards the task activity. Another example of her using the storyline narration to override subsidiary activities in the group may be found below: Excerpt 3 [Participants are discussing various illustration materials. From Turan (1999: 214).] 1

Selma:

2

Erol:

3

Esen:

4

Selma:

5

Erol:

6

Selma:

7

Erol:

8

Selma:

ay ben êunu bir yerde gördümya hvor (oh I have seen this before but where) ben de gördüm (I’ve seen it too) bagefter skal de på skal de til koncert nu (and then they are going to they are going to a concert now) hallo bak bir (hello take a look) nå ja så skal de også købe en kniv og sådan noget (alright then they also buy a knife and stuff like that) jeg har det her ay bundan bir yerde gördüm (I got it here oh I’ve seen one of these somewhere) Københavns filmfestival filmfestivaline gitsinler (Copenhagen Film Festival let them go to the filmfestival) hold kæft mand filmfestival lige midt i det hele

Social Relations in Bilingual Group Work

9

Esen:

10

Selma:

11

Asiye:

65

(shut up man a film festival in all this) nå det kan godt være (that might well be) o zaman (well then) o zaman êey yapak (well then let us do this)

At the outset of this episode, Selma declares in Turkish that she is looking for a certain illustration. The switch into Danish for the final particle of her turn (‘hvor’/’where’) may be seen as a way to extend the scope of recipients of her talk (cf. Auer, 1984a; Guldal, 1997) to include everyone present – in essence, to invite the group to participate in the search. Esen then aligns with this project, confirming that she has seen the image as well. In line 3, however, Esen demonstrably ignores their search, picking up the storyline with the temporal marker ‘bagefter’ (‘bagefter skal de til koncert nu’/’and then they are going to a concert now’), which is the most typical way of introducing the storyline in this group. However, she finishes her narrative turn with another temporal marker ‘nu’ (‘now), which clearly does not belong to the storyline, and I would suggest that this is a way to solicit the others’ engagement in the storyline narration. In fact, Esen’s entire turn may be seen as a directive aimed at the others to give up their search for an illustration and return to the storyline. This interpretation is confirmed by the responses of Selma and Erol who, although they respond in very different ways, both orient to Esen’s turn as precisely that: a summons to get back to work. Thus, in line 4, Selma challenges Esen’s request using the attention token ‘hallo’ and a code-switched request that the others look at her things. Erol, on the other hand, picks up the story theme (line 5), prefacing his contribution with a token of acceptance (‘nå ja’/’alright’) of Esen’s summons, and leaving it open-ended by means of an ‘et cetera’ particle (‘og sådan noget’/’and stuff like that’). At this, Selma persists in trying to focus the group on the illustration she has in mind, by first declaring that she has got it, which might gain the others’ attention, and then by switching to Turkish and qualifying her previous statement ‘ay bundan bir yerde gördüm‘ (‘oh I’ve seen it here somewhere‘). Ignoring Selma’s attempt to engage the others in her project, Erol (supposedly) produces an illustration of the Copenhagen Film Festival, suggesting that the story characters pay a visit there (line 7). This suggestion is immediately rejected by Selma (line 8), who implies that a visit to the film festival would mess up the entire story. However, her objection is rejected by Esen (line 9), who aligns with Erol’s suggestion, claiming that a visit to the festival may well fit into the storyline. Selma’s response in line 10 (‘o zaman‘/’well then‘) is hearable as a change of state token (Heritage, 1984), suggesting that she is prepared to accept the festival after all. In the final line of the excerpt, Asiye joins the interaction. Tying on to the format of Selma’s response (cf. Goodwin, 1990 on ‘format tying’), she urges the group to execute this idea. This episode shows how Esen exploits the storyline to terminate the subsidiary activities taking place in the group, and to direct the other group members towards her version of a proper working order. Naturally, to accomplish this

Bilingualism and Social Relations

66

task she has to rely on the cooperation of the other participants, and we have seen how her introduction of the storyline invites the others to follow suit. Finding an ally in Erol, she did not directly engage in the potential argument by challenging Selma’s pursuit of an illustration. Rather, not until Selma dropped her search, turning to debate the storyline with Erol, did Esen declare her support for Erol’s idea of the film festival. In other words, Esen engaged in the interaction in such manner as to terminate the other ongoing interactional projects. In effect, by the end of this episode, the group members were unanimously concerned with the creation of the story. Monopolising storytelling In the section above, I have shown how the commonly recognised task of creating an illustrated cartoon-like strip through a jointly narrated storyline is used to organise and indeed, regulate the activities within the group. I have also argued, on the basis of some empirical demonstration, that this is practised chiefly by one of the participants, Esen, who introduced this working order in the first place (cf. Excerpt 1). Specifically, I have shown Esen’s tendency to promote storytelling above other activities, by inviting other participants to co-narrate the story, and by aligning with those who are engaged in storytelling proper and closely related activities. Notably, I have shown that this is not an openly dogmatic practice – not, at least, in the hands of Esen. Rather, her way of focusing the interaction on what she views as the proper working order is one of building alignments across various issues arising in the unfolding activities. This conduct often led the other participants to sooner or later drop whatever projects they had at hand, and turn to the narration of the story. However, this presents a picture far too simple to account for the sort of social manoeuvring that takes place in Conversation 801, and in this section I focus on the monopolising work involved in the narration of the storyline. If co-participation in telling the story is made a virtue at times when other interactional projects are being entertained, the opposite picture emerges in interactional environments where all the participants are focused on the storyline, rendering storytelling as preferably a one-person activity. Our next example therefore illustrates some of the interactional means deployed in the competition for narrative space. Excerpt 4 [Erol has just settled a negotiation with Esen, postponing the writing of the story. Simplified version of transcript from Steensig (2000b: 51–52). Original transcript in Turan (1999: 213)] 1

Asiye:

2

Esen:

3 4

Erol:

5

Esen:

[öbür xxx bunlarÏn iêi ne oluyor] (the other xxx what are they gonna do) [hun kommer over til biografen ] køber popco:rn (she arrives at the movie buys popcorn) (.6) og c[ola ] t[o col]a (and cola two cola) [o::g] (.) [cola ] sammen me (a::nd) cola together with)

Social Relations in Bilingual Group Work

6 7

Selma:

8 9

Erol:

10

Esen:

11

Erol:

12 13

Selma:

14

Esen:

15

Selma:

16 17

Erol:

18

Esen:

19

67

si[n kæreste og ser Sp:eed] (her boyfriend and watches Speed) [og så ska hun tifrisør fris]ør (and then she’s going to the hairdresser hairdresser) (2) Speed ­êey mi ø::hm (is Speed eh) hi (what) Speed êey deÈil ki korku filmi (Speed is not it is not a horror movie) (1.2) [(det er sgu) ] (damn right it is) [den er sgu da] lidt uhygge[lig °mand°] (it is a little bit scary man) [ja: ] og (Gud (yeh and God) [ske) ] [neresi] uhyggelig kÏz si [zde] (where is it scary girl) [den] er lidt uhyggelig (it is a bit scary) så lidt (just a bit)

The overlap at the outset of this transcript comprises two very different simultaneous actions: Asiye’s inquiry about the details of the storyline, and Esen’s narration of the story proper. As we can see in lines 4 and 7, both of the other participants orient their actions to the narration. Thus, when Esen’s story reaches a short pause, Erol begins inserting his previous contribution, namely that the girl should also buy cola (cf. Excerpt 1). However, before Erol has finished, Esen resumes her narration with a turn-competitive articulation of the continuation marker ‘o::g’, that is, by competitively recycling Erol’s turn beginning. This may be a way of displaying her intention not just to go on with the narration, but also to incorporate Erol’s projected contribution into the storyline. However, Erol makes a second insertion through which he changes his original suggestion, now proposing that the girl should buy two Coca-Colas. While this occurs in overlap with Esen, it does not disrupt her continuation of the story in any way. Shortly thereafter, Esen’s narration is challenged again, this time by Selma’s clearly turn-competitive incoming (French & Local, 1983) in line 7. The first competitive feature of Selma’s turn is of course its interjacent placement within Esen’s narration, which at this point is nowhere near a recognisable ending. The high-paced delivery of the main part of Selma’s turn constitutes its second competitive feature, and the final aspect of her attempt to secure the overlap is the repetition of ‘frisør’ (‘hairdresser’), that is, the key element of her suggestion. Thus, Selma exploits an array of conversational techniques to contest Esen’s storytelling (see Schegloff’s (2000) account of various resources for

Bilingualism and Social Relations

68

overlap-competitive activities). Notably, Esen orients to the competitiveness of Selma’s turn, as shown in her prosodically marked (sound-stretch and emphasis) delivery of the last narrative unit (‘Sp:eed’). In short, she ratifies the competition for the narration as a shared project. This highly coordinated turn competition ends in a relatively extended pause. The silence is not broken by the speakers taking part in the narration, but by Erol who begins by asking something about the film that the fictive couple is supposed to watch at the cinema (line 9). In so doing, he orients to Esen’s part of the narration rather than Selma’s, thus construing Esen as the ‘winner’ of the competitive passage (cf. Schegloff, 1987), and we may see Esen further promoting the topic by soliciting an elaboration of Erol’s question. Now, as Erol suggests that Speed is not a horror movie (line 11), Selma strongly opposes that statement in overlap with Esen, who also rejects this notion but in a mitigated manner. In other words, the narrative activity is temporarily suspended in favour of a negotiation of the film’s status in terms of horror. We have seen that narrative activity can engage the best part of the group. On such occasions, however, storyline narration often displays monopolistic tendencies. Thus, the example above shows how Esen successfully keeps the other group members from participating directly in the narration of the story, by incorporating Erol’s already accepted contribution (cf. Excerpt 1), on the one hand, and by contesting Selma’s unsolicited, and potentially disruptive, attempt to introduce a new element, on the other. The next excerpt is taken somewhat later in the recording and shows another instance of Esen overriding another participant’s attempt at co-narration. Excerpt 5 [Selma is searching for an illustration. From Turan (1999: 214).] 1

Selma:

2

Esen:

3

Selma:

4

Esen:

5 6

Selma:

7

Esen:

mål for hvor fanden er det der måltid henne (meal where the hell is that meal) her (here) ja så kan de(well then they can) -bagefter går de over på en café og spiser kage (and then they go to a cafe and eat cake) og drikker kaffe (and drink coffee) ja (yes) og snupper en drink (and have a quick drink)

In brief outline, as Esen provides a solicited illustration of a meal, Selma turns to suggesting further activities for the storyline (line 3). However, as soon as Selma’s turn is projectable as a contribution to the story, Esen picks up the narration, causing Selma to drop her contribution in mid-turn. In contrast to the jointly competitive work we have seen in the previous extract, this time when Esen’s narration reaches a temporary completion, Selma simply confirms the story with

Social Relations in Bilingual Group Work

69

a single ‘ja’ (line 6), allowing Esen to add another storyline component. In other words, Esen’s narration not only disrupts and substitutes Selma’s potential contribution to the story, but also solicits Selma’s alignment with Esen’s version of the order of events. This is one of the examples of Esen sustaining her role as key narrator of the story, a role that is frequently recognised, and oriented to, by the other participants in the group. This observation notwithstanding, Esen’s attempts to present a coherent narration by herself are sometimes obstructed by the other participants’ attempts to introduce new ideas, as for instance in Excerpt 4. However, Esen’s monopolisation of the story’s narration is not restricted to the exclusion of new storyline components. The final example shows how Esen, through highly coordinated, locally sensitive delivery of storyline elements, manages to remain the sole author of the story’s recapitulation. The transcript begins as Erol suggests that they write down the story. Excerpt 6 [Participants are discussing the layout of the cartoon. From Turan (1999: 217).] 1

Erol:

2

Esen:

3

Selma:

4

Erol:

5

Esen:

6

Erol:

7 8

Selma: Esen:

9

Selma:

10

Esen:

11

Selma:

12

Esen:

13

Selma:

14

Esen:

15

Selma:

16

Erol:

altina yazalim be (let’s just write it down here) hvad skal hun hedde (what’s her name going to be) ay dur (oh stop) aman yapÏêtÏr gitsin ya yaparlar (oh just glue it they will do it) êunlarÏ yapalÏm da her er Lena (let’s do these here is Lena) oh bok gibi isim buldun ya (oh what a shitty name you found) ((laughs)) eski sevgilin mi (is she an old girlfriend of yours?) ja det er rigtig nok (yeh that’s it probably) hun ringer til biografen (she calls the theater) biografen og bestille billet ja (theater and book a ticket yes) for at bestille to billetter eller sådan noget (to book two tickets or something like that) ja for kæ-og så for kæresten (yes for the bo-for the boyfriend too) hun (she) nej for så er det lige(no then it is just) mobiltelefondan etsin

Bilingualism and Social Relations

70

17

Esen:

18

Erol:

19 20 21

Esen: Selma:

(she can call from a mobile phone) ringer (calls) bak (xxx) personligt altså personlighed (look xxx personal eh personality) to the movie ((laughter)) ((to Erol?)) hold kæft mand (shut up man)

Erol’s attempt to focus the group on writing fails as Esen asks for the story girl’s name, thus proposing a somewhat different trajectory of action on behalf of the group. Note that Esen’s disaffiliative turn diverges from Erol’s on the level of language choice, just as Selma’s subsequent rejection of this proposal in turn diverges from Esen’s choice of Danish. In effect, Selma’s protest suggests an incipient alignment with Erol on the level of action as well as language choice. Upon Erol’s second attempt to direct the group towards some progress (line 4), Esen begins narrating the story with the preface êunlarÏ yapalÏm‘ (‘let’s do these‘). Note that the story preface is delivered in the language of immediately preceding talk, allowing for a code-switch to Danish at the outset of the narration to set off the story proper from previous interaction. Another notable aspect of Esen’s beginning narration is that she has given the story girl a name, Lena. This allows Erol to heckle the storyteller (Sacks, 1992, Vol. 2) by ridiculing her choice of name. Again, his disaffiliative action involves a contrastive language choice. Although Selma’s laughter in line 7 may be seen as another bid for alignment with Erol, he does not seem to ratify the alliance. Thus, when Esen counters the heckling by suggesting (now in Turkish) that Erol’s dislike of the name Lena is motivated by his bad experience of a girlfriend with the same name, Selma breaks up the incipient alignment with Erol by supporting Esen’s counterattack (now in Danish). Once this dispute is settled, leaving Erol with no ally, Esen proceeds with her recapitulation of the story (line 10). However, as her first narrative turn reaches completion, Selma picks up its last lexical unit starting to co-narrate the story (line 11). Notably, she ends her co-narrative turn with a token of agreement (‘ja’), thereby casting her own turn as a continuation, or even completion of Esen’s turn. On the level of action, we might view this move as a way of soliciting co-narrative alignment with the storyteller. The alliance is not achieved, however, as Esen clearly rejects Selma’s contribution in line 12. This is accomplished chiefly in two ways: first, Esen tailors her turn beginning in such a way as to preserve the continuity of the narration from her own turn in line 10. Thus, she observes the interturn tying rules (Sacks, 1992, Vol. 1) to the degree that lines 10 and 12 form not just a coherent story part, but indeed a syntactically valid sentence. In this sense, Esen’s turn in line 12 disqualifies Selma’s contribution as part of the story proper. The second aspect of Esen’s rejection concerns content, namely her pointing out that the girl orders two tickets for the movie constitutes a case of other-initiated other repair (Schegloff et al., 1977), orienting in this way to Selma’s contribution as problematic. We should further note that the open-ended et cetera-type closure of this turn may be a way of marking this type

Social Relations in Bilingual Group Work

71

of repair as a dispreferred activity. In response, Selma orients to the repair aspect of Esen’s action, volunteering an explication that the second ticket is for the boyfriend, which again might serve to solicit the other girl’s alignment. However, the alignment is not mutually ratified this time either, as Esen simply continues narrating the story (line 14), this time producing a single word ‘hun’ (‘she’) as if she were writing it down (there is some evidence, occuring about 10 turns later, to suggest that this may be happening here). At this, Selma starts protesting (‘nej for så er det lige’/’no then it is just’), but is interrupted by Erol, who suggests in line 16 that the girl call the movie from a mobile phone. In so doing, he manages to anticipate Esen’s next turn (line 17) by projecting the story’s continuation (of course, if indeed Esen is writing it down, he may simply be using visual information to anticipate her turn at talk). Again, Esen’s turn is a single word ‘ringer’ (‘calls’; ‘rings’), which ties directly onto her previous turn in line 14. Erol’s subsequent attempt to disrupt the storytelling is prefaced with a bid for attention ‘bak‘ (‘look‘), upon which he raises some issue of personality (marked in the transcript as partly inaudible). In any event, during the last few turns, Esen’s recapitulation of the story has been challenged by different contributions from two other participants. We should also note that both Turkish and Danish have been used for these turns. And it is precisely with this consideration that we should understand Esen’s switch to English in line 19, in which she produces another storyline element, minutely tied to her previous turn in line 17 (apart from the code-switch, of course). I would therefore suggest that, by switching to English, Esen is able to continue her recapitulation in a bilingual sequential environment that presents a threat to her monopolistic storytelling. It is possible that the ensuing laughter is a reaction to this unfolding of events. In any case, Selma’s demand that Erol shut up (although perhaps not dead serious, as it occurs in a laughing environment) seems to display her orientation to Erol’s actions as potentially disruptive of the storytelling. In sum, the final excerpt provided further illustration of the sort of work involved in maintaining an internally coherent storyline in a conversational environment in which several participants are bidding for participation in the narrative activity. For lack of a better term, we have called this practice monopolisation of the storyline narration. Several aspects of this practice have been discussed at some detail, showing how structural matters of turn organisation are part and parcel of the discursive accomplishment of social relations. Finally and in conclusion, let us consider some issues of participation in the group’s work in terms of interactional asymmetry and dominance.

Some Remarks on the Practice of Interactional Dominance in Conversation 801 The analysis above has highlighted some relational aspects of the students’ engagement in construing an illustrated cartoon strip, organised in important parts around the collaborative narration of a storyline. It was argued that the verbal production of this storyline was established in the early phases of this conversation as a sort of procedural agenda for the group. In support of this claim, I have discussed at some length several negotiations of activities in which

72

Bilingualism and Social Relations

this agenda is sustained, showing that the participants indeed orient to the storyline as informing their conduct. In effect, most of the interaction taking place within the group relates in different ways to the unfolding storyline. Apart from informing the various activities in which the students engage, the storyline – or rather its narrative production – has been exploited to regulate the interaction in the group. Specifically, it was shown that this was practised chiefly by Esen, who frequently engaged in storyline narration to suppress alternative activities within the group. This engagement relied for its success on her efforts to build alignments across a variety of issues evolving during the group’s work, and the analysis highlights her preference for topics and activities closely related to the storytelling. In other words, the interactional project Esen is demonstrably pursuing is that of keeping the storyline going, and the obvious consequence of this is that in pursuing this project, she engages in regulating the group’s activity. Now, her attempts to focus the group on storyline narration do not in themselves warrant the description of her as interactionally dominant, as this still allows the others to take an active part in the narrative activity, and Conversation 801 entails several instances of co-narration. However, we also see numerous examples of attempts at co-narration being either ignored or otherwise disqualified, or indeed directly contested by Esen, and the analysis has shown various techniques used by her to monopolise the storytelling. I wish to suggest that it is Esen’s tendency to impose her version of a proper working order upon the group, on the one hand, and to monopolise participation in the very activity constituting that order, on the other, that we may view as interactionally dominant. While in principle, it leaves the other students little opportunity to contribute meaningfully in the group’s work altogether, in practice it works to sustain her once established (in Excerpt 1) status as the key narrator. Note that the other participants orient to this status in various ways. For instance, it is common that the other students attempt to clear their narrative contributions with Esen regarding content (cf. Excerpt 1, line 25; Excerpt 3, line 7; Excerpt 6, lines 11 and 13), whereas she seldom seeks the others’ approval in this regard. Also, in cases of simultaneous, disruptive or otherwise competitive narration involving Esen as of the actors, the subsequent speakers pervasively orient to Esen’s version of the story (cf. Excerpt 4, lines 2 and 6; Excerpt 5, lines 4–5). In this way, the other participants do not merely display their orientation to Esen as key narrator, but also reflexively sustain this status. Considering that the narrative activity constitutes the backbone of the group’s work with the assignment, acting (and being recognised) as the principal narrator is very much akin to being the supervisor of the group’s interaction. This may well account for Esen’s monopolising tendencies in the narrative activity. But let us consider the present analysis of asymmetrical participation in light of some previous studies of the group conversations in the Køge data (Jørgensen, 1998; Maegaard, 1998; Steensig, 2000a). Very roughly, these studies characterise Asiye as the interactionally most withdrawn participant, Erol as a rather talkative but relatively powerless one, Selma as a participant who gets her way with all the others except Esen, and Esen as the person who dominates the interaction most of the time. By and large, the present analysis seconds these conclusions, possibly with a reservation for the distinction between Erol and Selma, which seems less clear-cut in the present analysis.

Social Relations in Bilingual Group Work

73

However, any comparison of participants in terms of interactional asymmetry and dominance, or any other aspect that may be glossed as ‘power’, begs the question of how this power is accomplished in the situated flow of events, that is what types of resources are used to dominate the interaction. Here, Jørgensen (1998) specifically points to the students’ (and particularly Esen’s) use of code-switching as a means of domination and control. In the present study, language choice and code-switching have been discussed as interactional techniques for soliciting and ratifying alliances and for highlighting oppositional actions. However, the present account of interactional asymmetry relates primarily to the storytelling activity. This activity in turn implies a specific organisation with respect to language choice. In the present conversation then, perhaps some of the code-switching practices related to dominance and control (‘power wielding’) can be accounted for in terms of the storyline narration and the linguistic division of labour involved in that activity. The issue of powerful or dominant participation in the group’s work with the cartoon would then not be, in the first place, a matter of language choice or code-switching per se, but of the act of imposing storytelling upon other activities taking place in the group and of monopolising participation in the story’s narrative production. Acknowledgements The author wishes to thank Normann Jørgensen, Fadime Turan and other persons involved in the Køge project for making their data available to the author as well as other researchers. This also includes Jakob Steensig, who provided a detailed transcript of a portion of Conversation 801. The author is also indebted to Asta Cekaite for insightful comments on a late draft of this article. This article was prepared with financial support by the Committee for Educational Research of the Swedish Research Council. Correspondence Any correspondence should be directed to Dr Jakob Cromdal, Department of Child Studies, Linköping University, S-58183 Linköping, Sweden ([email protected]). References Aronsson, K. (2000) Address, affect and audience design in bilingual multiparty talk. In J.N. Jørgensen and A. Holmen (eds) Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe (pp. 91–100). Copenhagen Studies in Bilingualism, Køge Series, K7. Copenhagen: The Royal Danish School of Educational Studies. Auer, J.C.P. (1984a) Bilingual Conversation. Amsterdam: John Benjamins. Auer, J.C.P. (1984b) On the meaning of conversational code-switching. In J.C. P. Auer and A. di Luzio (eds) Interpretive Sociolinguistics. Migrants – Children – Migrant Children (pp. 87–112). Tübingen: Gunter Narr. Chaudron, C. (1988) Second Language Classrooms. Cambridge: Cambridge University Press. Cromdal, J. (2000) Creating a monolingual story in bilingual conversation. In J.N. Jørgensen and A. Holmen (eds) Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe (pp. 57–75). Copenhagen Studies in Bilingualism, Køge Series, K7. Copenhagen: The Royal Danish School of Educational Studies. Cromdal, J. (2001) Managing bilingual overlap: Some implications of code-switching for overlap resolution. Research on Language and Social Interaction 34, 421–451.

74

Bilingualism and Social Relations

Durán, R.P. and Szymanski, M. (1995) Cooperative learning, interaction and the construction of activity. Discourse Processes 19, 149–164. French, P. and Local, J. (1983) Turn competitive incomings. Journal of Pragmatics 7, 17–38. Garfinkel, H. (1967) Studies in Ethnomethodology. Cambridge: Polity Press. Goodwin, M.H. (1990) He-said-she-said. Talk as Social Organization Among Black Children. Bloomington: Indiana University Press. Guldal, T.M. (1997) Three Children, Two Languages: The Role of Code Selection in Organizing Conversation. NTNU Trondheim, Norway. Heritage, J. (1984b). A change-of-state token and aspects of its sequential placement. In J.M. Atkinson and J. Heritage (eds) Structures of Social Action: Studies in Conversation Analysis (pp. 299–345). Cambridge: Cambridge University Press. Holmen, A. and Jørgensen, N. (1997) Aspects of the linguistic development of minority children in a majority school. In J.N. Jørgensen and A. Holmen (eds) The Development of Successive Bilingualism in School-age Children (pp. 129–146). Copenhagen Studies in Bilingualism, 27. Copenhagen: The Royal Danish School of Educational Studies. Hutchby, I. and Wooffitt, R. (1998) Conversation Analysis. Cambridge: Polity Press. Jefferson, G. (1990) List-construction as task and resource. In G. Psathas (ed.) Interaction Co mp e te nce (p p. 63 –9 2). Was hi ng t o n, DC : Inte rnatio nal Ins titute fo r Ethnomethodology and Conversation Analysis and University Press of America. Jørgensen, J.N. (1998) Children’s acquisition of code-switching for power-wielding. In P. Auer (ed.) Code-switching in Conversation: Language, Interaction and Identity (pp. 237–258). London: Routledge. van Lier (1988) The Classroom and the Language Learner: Ethnography and Second-language Classroom Research. London: Longman. Jørgensen, J.N., Holmen, A., Gimbel, J. and Nørgaard, I. (1991) From Köy to Køge: A longitudinal study of the bilingual development of Turkish immigrant children in Danish schools. Language and Education 4, 215–217. Maegaard, M. (1998) Sprogvalget i gruppesamtaler (Language choice in group conversation). In J. Møller, P. Quist, A. Holmen and J.N. Jørgensen (eds) Tosproget Udvikling (Bilingual Development) (pp. 21–40). Copenhagen Studies in Bilingualism, Køge Series, K4. Copenhagen: The Royal Danish School of Educational Studies. MacWhinney, B. (1995) The Childes Project. Tools for Analyzing Talk. Hillsdale, NJ: Lawrence Erlbaum. Møller, J. (1998) Identitet og kodevalg hos unge piger med tyrkisk-dansk baggrund (Identity and code choice in young girls with a Turkish-Danish background). In J. Møller, P. Quist, A. Holmen and J.N. Jørgensen (eds) Tosproget Udvikling (Bilingual development) (pp. 41–70). Copenhagen Studies in Bilingualism, Køge Series, K4. Copenhagen: The Royal Danish School of Educational Studies. Sacks, H. (1992) Lectures on Conversation. Vols 1&2. (Gail Jefferson, ed.) Oxford: Blackwell. Schegloff, E.A. (1987) Between micro and macro: Contexts and other connections. In J.C. Alexander, B. Giesen, R. Münch and N. J. Smelser (eds) The Micro-Macro Link (pp. 207–234). Berkeley: University of California Press. Schegloff, E.A. (2000) Overlapping talk and the organization of turn-taking for conversation. Language in Society 29, 1–63. Schegloff, E.A., Jefferson, G. and Sacks, H. (1977) The preference for self-correction in the organization of repair in conversation. Language 53, 361-382. Steensig, J. (2000a) Notes on some uses of code-switches and other interactional devices in Conversation 801. In J.N. Jørgensen and A. Holmen (eds) Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe (pp. 9–30). Copenhagen Studies in Bilingualism, Køge Series, K7. Copenhagen: The Royal Danish School of Educational Studies. Steensig, J. (2000b) CA transcript of Conversation 801, Køge Project. In J.N. Jørgensen and A. Holmen (eds) Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe (pp. 31–55). Copenhagen Studies in Bilingualism, Køge Series, K7. Copenhagen: The Royal Danish Scool of Educational Studies.

Social Relations in Bilingual Group Work

75

Turan, F. (ed.) (1999) A Text Collection of Turkish-Danish Bilingual Grade School Students’ Conversations. Copenhagen Studies in Bilingualism, Køge Series, K6. Copenhagen: The Royal Danish School of Educational Studies. Tuyay, S., Jennings, L. and Dixon, C. (1995) Classroom discourse and opportunities to learn: An ethnographic study of knowledge construction in a bilingual third-grade classroom. Discourse Processes 19, 75–110. Willett, J. (1995) Becoming first graders in an L2: And ethnographic study of L2 socialization. TESOL Quarterly 29, 473–503.

Appendix: Transcription key (2) (.) (( )) [ = (x) (xxx)

numbers in single parentheses represent pauses in seconds micropause, i.e. pause shorter than (.5) investigator’s comments indicates start of overlapping speech indicates latching between utterances inaudible word inaudible words

® : og cola

highlights a particular feature discussed in the text prolongation of preceding sound

EXACTLY °( )° ­¯ ? . >< hi; ha; he; hö; hh vay yavrum

sounds marked by emphatic stress are underlined capitals represent markedly increased amplitude embeds talk markedly lower in amplitude indicates rising/falling intonation in succeeding syllable(s) indicates rising terminal intonation indicates falling terminal intonation embeds talk that is faster than surrounding speech embeds talk that is slower than surrounding speech indicate varieties of laughter italics mark speech in Turkish

Language Choice as a Power Resource in Bilingual Adolescents’ Conversations in the Danish Folkeskole Trine Esdahl Department of Nordic, University of Copenhagen, Njalsgade 120, DK-2300 Copenhagen S, Denmark This paper deals with the negotiations of social relations among adolescents at a crucial point in their development of a bilingual identity. Based on a general study of bilingual Turkish–Danish Folkeskole pupils’ development of language choice and code-switching, we find that the seventh grade is a pivotal stage in their linguistic development. Sudden abrupt changes in the language choice patterns of girls in particular seem to take place in the seventh grade. We take a qualitative look at the negotiations in a group conversation among seventh grade boys and a group conversation among seventh grade girls, and we find both similarities and differences. Among the similarities is the fact that social relations seem to determine language variation (in casu language choice) with both genders – considerably more so than at younger ages. Among the differences it is found that the power struggles among the girls are surprisingly tougher than among the boys. The girls’ group has a clear winner and a clear loser, whereas the boys’ group deals with the power struggle in a more playful way.

Introduction The relationship between social variation and linguistic variation has always been the main focus of sociolinguistics. Early sociolinguistics studied the influence of social variation on language variation, including the language choices made by bi- or multilingual speakers. Modern sociolinguistics tends to study language variation as a cause of social variation in that social relations are seen to be constructed and negotiated through variation in language use. However, this issue is, of course, not as simple as the question of whether language is an effect or a cause of social structures. In terms of bilingual language use, the relations particularly between societal evaluations of the languages involved and specific language choices made by bilingual speakers in specific situations are complex. Speakers use their shared knowledge of generally held beliefs and evaluations to establish rapport and a sense of community by commenting on them (directly or indirectly), and by ironising over them. In this paper I shall present some of the results of Esdahl (2001). I will describe the general development of language choice patterns of a group of bilingual pupils of the Danish Folkeskole, and I will show how language choice is used as a power resource. I focus on a specific phase of the development by analysing two conversations between bilingual adolescents in Grade 7. I will show differences between boys and girls in the way they use their two languages and the switches between these languages as means of getting their way and as resources to gaining power in the conversation. The material is taken from the Køge Project, a longitudinal study of the bilingual development of Turkish-Danish pupils in the Danish Folkeskole (see Turan, 1999). My choice 76

Language Choice in Bilingual Adolescents Conversations

77

Figure 1 Conversation types and proportions of Danish

of Grade 7 is made because there is a change in the language choice patterns compared with Grade 6, see Figure 1. Figure 1 shows the language choice patterns in group conversations with different compositions of participants, namely girls’ conversations, boys’ conversations, and gender-mixed conversations. The change between Grade 6 and Grade 7 is characterised by the girls at this time having a steep rise in they percentage of Danish utterances. Generally the boys – when they participate in pure boys’ conversations – from Grade 3 to Grade 7 show a slow increase in the percentage of Danish-based utterances; after grade 7 a slight fall starts. The girls’ percentages of Danish-based utterances are very low until Grade 6 in pure girls’ conversations. From Grade 6 to Grade 7 a sudden, steep increase starts, continues to Grade 8 and levels out in Grade 9. The change in the girls’ language choice pattern between Grade 6 and Grade 8 is the most profound change of all in the developments that one can observe. The boys’ percentages of Turkish-based utterances fall slowly from Grade 1 to Grade 7. From Grade 7 this fall levels out. Like it is the case with the percentages of Danish-based utterances in pure girls’ conversations where the change came in Grade 7, the change in the girls’ percentages of Turkish-based utterances also accelerates in grade 7. Their share of Turkish-based utterances is the main part of all their utterances until Grade 7. From Grade 7 there is a sudden steep fall in their percentage of Turkish-based utterances. There is of course no great surprise in the simultaneous rise in Danish and fall in Turkish, but as one observes by

78

Bilingualism and Social Relations

Figure 2 Conversation types and proportions of Turkish

comparing Figure 1 and Figure 2, the development of choice of Turkish and Danish is not absolutely parallel. This is due to the fact that mixed utterances appear with different strengths at different times. In Grade 7 we find a great difference between the development of the girls’ and the boys’ language choice. Furthermore, as we shall see, there is a marked difference between how the girls and the boys use their languages as a power resource at this grade.

Bilingualism Research The relations between language variation and societal variation with respect to bilingualism are the focus of Gumperz’ (1982) distinction between they- and we-codes. He describes how the two languages of bilinguals have gained different prestige in society. These differences in the languages’ prestige are the result of structural differences in society. Gumperz distinguishes between languages with high status and languages with low status. The choice of language by bilinguals may in specific situations be determined by their sense of what is appropriate, and in such cases a change in the situation (participants, theme, or otherwise) may lead to a code-switch, a so-called situational code-switch. However, the interlocutors may also choose to switch codes in order to convey a meaning, which will to a large extent depend on the societal evaluation of the languages (and consequently on the languages’ function as a

Language Choice in Bilingual Adolescents Conversations

79

we- or they-code of the speakers). In this case, the code-switch is not situational, but metaphorical. Rampton (1995, 1998) further develops the realisation that speakers may refer to the societal evaluations of languages and varieties. According to Rampton, all social relations can be negotiated in social interactions. Thus the social interactions also contribute to the creation and negotiation of identities. With his concept of crossing, Rampton introduces a new dimension in bilingualism research. A speaker has the possibility of using a language that is not considered his ‘own’ and thereby causing a re-negotiation of the relations brought along in the interaction (1998: 291). Crossing gives the speaker the opportunity to explore the identity and ethnicity of others and to re-define her or his own identity (1998: 300). It also provides the speaker with an opportunity to define and re-define the relations between the speech community and the outside reality. In recent years, studies on bilingualism such as Sebba and Wootton (1998) and Auer (1998) have included code choice patterns and code-switching in a similar perspective. Negotiations and re-negotiations of social relations define and re-define not only social relations in the group of the interaction, but the outside society is no longer considered to be the determiner of language use. For instance, the use of two languages in the conversation can be a pragmatic power tool. Many factors of the interaction influence the choice of language. The social relations brought along in the conversation can cause a speaker to choose a certain language. But the possibility of re-negotiating the social relations can cause new social relations to be brought about, and these brought about relations influence the language choice. Language choice can be made either as a result of factors brought along or brought about in the conversation.

Power and Language Ng and Bradac (1993: 4) define power as ‘the probability that one actor within a social relationship will be in a position to carry out his or her own will despite resistance, regardless of the basis on which this probability rests’. Power is thus based on the possession of resources that provide the speaker with the ability to realise his own intentions and interests despite the interests of other participants. All conversations are also battles over power, and in the battles of the conversations linguistic resources can be used as tools. These resources can be brought along and by that be defined by outer factors such as social background. They can also be brought about, however, and thus won or acquired as competence or personal character. The resources can be negotiated and re-negotiated in the interaction and thus bring about new resources. The new resources can be specific for a single conversation, but they can also be brought along to the next interaction as well. Power in a conversation is about gaining control of the conversation. Every participant in a conversation can be interested in influencing interlocutors and carrying out her or his own interests. Control of the conversation is not necessarily enough to realise one’s aims, however. The most important way to gain power in a conversation is through language. The power of language can depend on stylistic features in the language such as the difference between powerful and powerless styles (O’Barr, 1982; O’Barr & Atkins, 1980). The difference between

80

Bilingualism and Social Relations

powerful and powerless styles depends on the fact that some features in the language are perceived as stronger than other features. The interpretation will depend on the situation. A specifically bilingual power tool is language choice. However, the effectiveness of a power motivated language choice may depend on the difference in status between the two languages. These differences in status are again caused by outer factors. They are not brought about in every conversation, but exist as predetermined language attitudes prevailing in each single individual. Differences in status or prestige are nevertheless not entirely due to outer influences from society. In an interaction these differences can be re-negotiated as every other power resource, and linguistic resources achieve a new status. Thus new resources are brought about in the conversation. The difference between powerful and powerless styles is not restricted to varieties of a language. One language can be more powerful than another, and bilinguals thus have a powerful and a powerless language. Bi- and multilinguals have the possibility of choosing between their languages and, by doing that, use them as power resources. The choice of language can indeed be based on a principle of ethic (Boyd, 1995) when a speaker chooses the language that is most considerate to the other interlocutors, i.e. which offends the interlocutors the least in the situation. On the other hand language choice can depend on a power principle. The participant being most influential in the conversation has the power to choose language in the conversation. Having the power to choose the language spoken in the conversation does not necessarily mean that one uses this power, however. And it does not imply that one’s interlocutor follows the principle of ethic. A new situation arises when the most powerful person is challenged and the power relations are brought to negotiation. In that case the societal (global) distribution of power no longer solely determines the language choice, but local factors also influence the conversation. Specific factors in the conversation caused by values and relations created in the conversation become factors that can influence the language choice. These values determined by the situation are of great importance in the practice of pragmatic code-switching. The code-switch itself can be a means of gaining power independent of the direction of the switch (Jørgensen, 1998). Differences in status between the languages, Gumperz’ weand they-codes, and other outer circumstances are present as a shared knowledge, but this knowledge can be challenged and rejected. Speakers substitute them with locally negotiated relations and values, i.e. which are created in the conversation.

Gender and Language Differences The differences in linguistic behaviour between men and women is a phenomenon which has been described by classical feminist sociolinguistics as a power relation. Men are seen as suppressors, being assertive and competitive. Women, on the other hand, are seen as complacent and cooperative (Coates, 1993: 12). Another view of gender-related linguist differences postulates that men and women belong to different societal subcultures. This causes the differences in their languages, and these differences reflect the subcultures and social reality (Coates, 1993: 13).

Language Choice in Bilingual Adolescents Conversations

81

Men’s linguistic behaviour is more power-oriented than that of women. Particularly in conversations involving only women the style of speech is described as one which reflects solidarity. Women in social interaction with other women use language to establish interactional cooperation. They do not allow one person to dominate the group and do not compete against each other. The subject of conversation is often a personal or emotional one. Quite contrary to this, conversations between men are seen to not involve personal issues. The subject of conversation is changed more often among men. Men’s social interactions are focused on competition within the group. The language is used as a means of gaining power, for instance through interruption and other aggressive features. Children gain knowledge of society’s stereotypical view of gender differences by having their parents as role models. They then signal their identity as boys or girls through their language behaviour in ways similar to their parents, and societally determined linguistic differences are reproduced. In other words, these views regard linguistic differences as results of social differences. Cameron (1997), however, contributes considerably to the discussion of linguistic behaviour being caused by social phenomena or vice versa. She reviews recent studies that show a less categorical view of the relations between gender, language and reality. Solidarity and power are all variable concepts, which depend on the context and are created in the context. Furthermore, Johnson (1997) considers gender as a linguistic construction and social structures as results of linguistic variation. The way one speaks can signify gender in the same way as gender can be expressed by e.g. choice of clothing. In this way linguistic variation can be used to build social differences (and similarities). In my analysis I compare the use of language choice as a power tool among boys and girls. I do not think that linguistic behaviour, in this case language choice or linguistic variation, is necessarily given in advance by societal structures, or that they can be understood only as results of influence from society. This does not mean that the societal structures do not influence linguistic behaviour. The relations between linguistic variation and societal structures go in both directions. In my analysis, I aim to determine whether language choice is used by the speakers to influence the negotiations in such a way that negotiations are influenced not only by circumstances in the conversation, but also by the speakers’ taking outside factors into consideration. If one takes for given that circumstances outside the conversation to a great extent determine linguistic variation, then the adolescents in my study should be influenced by the Danish school policies and the input regarding language they receive from teachers, school authorities, and public statements about appropriate language choice. The fact is that the bilingual pupils are pressed to speak Danish if they want to be a part of the society outside the family. If the outer circumstances do influence language choice the Køge data should consequently show an increasing use of Danish over time. Figure 3 shows the language choice patterns of the bilingual pupils as a group. The figure depicts a development towards an increased use of Danish and a decreased use of Turkish, and a slight increase in the use of mixed language. The conversations until Grade 3 contain very little Danish. After Grade 3 there is a gradual increase of Danish which continues until Grade 6, after which the use of

82

Bilingualism and Social Relations

Figure 3 The development in language choice

Figure 4 Gender and proportions of Danish and Turkish

Language Choice in Bilingual Adolescents Conversations

83

Danish accelerates. All of this confirms that there may in fact be a societally influenced development in language choice patterns. On the other hand, the decrease in the share of Danish in Grade 9 and the simultaneous increase in the share of Turkish are an unexpected development. Perhaps the increase of Danish from Grade 1 to Grade 8 can be seen as a sign of society’s influence on the young bilinguals’ language choice in group conversations which take place in the school, and the change between Grade 8 to Grade 9 as a change towards a status levelling between the two languages, or a change reflecting a more consciously bilingual attitude to language choice. I have also compared the language choice patterns of the boys and the girls. There is very little Danish until Grade 4, in which the boys begin to use some Danish. In Grade 5 the girls use more Danish. This is probably a result of the composition of the groups – in Grade 5 there are only gender-mixed groups. After Grade 6 the girls use more Danish than the boys. In Grade 9 the girls’ percentage of Danish is 68% of all utterances, while the boys choose Danish in only 33% of their utterances. The proportion of Turkish-based utterances generally decreases until Grade 8 – this goes for both genders. From Grade 8 to Grade 9 the use of Danish by the girls levels out. All of this combined suggests that the girls generally lead the way in the development of language choice patterns. Apparently the girls and the boys develop differently, but in fact we cannot know if it is a result of the girls’ earlier development, or if it is a difference between genders that will be maintained into adulthood. Nevertheless it is remarkable that the girls show greater variation in their linguistic behaviour than the boys do. We also find greater variations in language between the girls individually than between the boys individually. Certain girls lead the way in the development of language choice patterns. This leads to the important conclusion that the girls apparently are more characterised by linguistic inequality than the boys are. This implies that there is a more uneven distribution of linguistic power resources among girls than among boys, and as we shall see, some of the girls know very well how to take advantage of these resources. An example of the linguistic behaviour of the girls can be found in the following excerpt of a Grade 7 conversation. Conversation 702 takes place between the girls Esen, Selma and Asiye. The conversation is characterised by being mostly in Danish. Two of the girls, Esen and Selma, speak Turkish less than the third girl, Asiye. More than 50% of all Asiye’s utterances are Turkish, and she only speaks Danish in 29% of her utterances. Esen and Selma on the other hand choose to speak Danish in more than half of all their utterances. The three girls choose their language differently, but they also use their choice and their switches differently in the interaction. Conversation 702 (Esen, Selma, Asiye), excerpt transcribed according to the CHILDES conventions, MacWhinney (1995), with Turkish in italics: Asiye:

Esen Esen hani sen dedin ya o Ali xxx sana geri gelir dedinya hani ben Ïonun adÏnÏ yazdÏmya hani gstermiyelim o mektubunu at [//] attÏm diye kÏz bir tane mektup gösterdinya.

84

English:

Esen: Asiye: English: Esen: English: Selma: English: Esen: English: Selma: English: Esen: English: Esen: English: Asiye: English: Esen: English: Asiye: English: Esen: English:

Bilingualism and Social Relations

Esen Esen you said it that this guy Ali xxx. It will come back to you, you said so yourself. I wrote his name, you know, let us not show anyone your the letter, throw – which you had thrown, girl, you showed a letter. mm. o var mektup varmÏê buraya sen dedin varmaz kendine geri gelir dedin xxx ama geri varmÏê oraya. that reach- the letter came here you said, it would not get there, it will come back to yourself again, you said, xxx, but it came back there (hums). # det var Selma der sagde at den ville komme igen. it was Selma who said that it would come back again. åh ja også dig ikke kun mig. oh yeah you too not just me. jeg sagde at det kunne vi ikke vide. I said that we could not know. jeg sagde da også det ikke ]. I also said that it would not. <er de>[] and [ [>]

Languaging Among Fifth Graders

Eng: Erol: Eng: Esen: Eng: Selma: Eng: Erol: Eng: Ali: Eng: Selma: Eng: Esen: Eng: Ali: Eng: Erol: Eng: Ali: Eng: Erol: Eng: Ali: Eng: Selma: Eng: Esen: Eng: Erol: Ali: Eng: Erol: Eng: Esen: Eng: Erol: Eng: Ali: Eng:

135

I have not been there < ben de > [] we spend some time in stanbul, then we are in Uêak, and then we are in Ankara < biz eskiden [/] > [< >] in the old days we [/] < Australien > [] dindi I lost my Turkey again [ [ [>] SAS (an airline) [ [>] come on, Erol (complaining) <manyak.>[] you can put it into that one you can try # then you will have money xxx [< >] cut it off xxx [>] <no I am Danish>[[>] >[ [>] Sealand dos (with exaggerated intonation, ‘dos’ equivocal as ‘there are also’ or ‘fool’) [ [>] yes I am is that not xxx. [[>] this is not Africa, kikicikcik (singing) this is not Africa [