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<

fast

»darunter«

very fast

/ver/ste/hen

staccato/syllabic kind of speech

+

pause below one second

(+)

micropause

(h)

onset hesitation

=

fast connection

kom[men [da

overlap and point of overlap

oðlum

Turkish text in italics

Cultural Orientation and Language use among Multilingual Youth Groups: ‘For me it is like we all speak one language’ nci Dirim Kronprinzenstraße 26, D-22587 Hamburg, Germany

Andreas Hieronymus Beckstrasse 4, D-20357 Hamburg, Germany This paper approaches the linguistic setting of a European inner-city, an area with a long history of migration from a sociological and linguistic perspective. After a qualitative-heuristical research and (ethnomethodological) conversation analysis, we are able to present a description of sociolingual conditions and the process of their transformation. In the first part we focus on the relationship between language, culture, identity and social structure. We argue that old structures are breaking up, and that new patterns of speech-communities are developing. We shall come out against common ideas about lingual transformation in the Federal Republic of Germany: against the ‘assimilatonist view’ (the language of origin disappears with the third generation of migrants) and the ‘view of difference’, this is connected with the concept of ‘ethnic revival’ and observes the revitalisation of the languages of origin). Our argument will be put forward in an exploration of phenomenon of ‘mixed language’ (that’s what the interviewees called their way of speaking) in the world of inner-city juveniles. By embedding these linguistic practices in the ethnographical concept of ‘liminality’, we tried to understand the loosening process of the relationship between linguistic practices, identity through nationality and social structure. The usage of language appears as a means of constructing their social position among adolescents in the location of research where social structure is embodied.

Introduction This paper approaches the linguistic setting of a European inner-city, an area with a history of migration, from a sociological and linguistic perspective. After a qualitative-heuristic description (Kleining, 1994) and (an ethnomethodological) conversation analysis (Auer, 1983: 9f.), we are able to present a ‘thick description’ (Geertz, 1994) of sociolingual conditions and the process of their transformation. Our data come from the project Turkish in Mixed Cultural Groups (Auer & Dirim, 2000, 2001) which was carried out between 1995 and 2000 in Hamburg, Germany. In the first part we focus on the relationship between language, culture, identity and social structure. We argue that old structures are breaking up, and that new types of speech communities are developing. We shall argue against common ideas about lingual transformation in the Federal Republic of Germany: against the assimilationist view that the language of origin disappears with the third generation of migrants, as well as the view of difference connected with the concept of ethnic revival and observes the revitalisation of the languages of origin. Our argument will be put forward in an exploration of the phenomenon of mixed language in the world of inner-city adolescents. By embedding these linguistic practices in the ethnographical concept of liminality (Turner, 1995), we try to understand the loosening of the relationship between linguistic practices, national identity and social structure. The use of language among 42

Cultural Orientation and Language Use

43

adolescents appears as a means of constructing their social position in the location where our research was carried out. In this process Turkish stands out against other migrant languages (Auer, 1995). The results of the first part of our study will be supplemented by the second part, which deals with the aquisition and use of Turkish by young people of different backgrounds in Hamburg.

Some Remarks on the Strategy and the Location of our Research This study is based on extensive qualitative data, which were collected in different areas of Hamburg. Since the industrial revolution, districts with a large number of migrants have grown up in the city-scape of Hamburg. People are relatively poor and the areas are affected by gentrification. Our study has its focus on the role Turkish plays in linguistically mixed adolescent groups, and on the use of Turkish by non-Turkish youth. Our methodology was a heuristic one. It was inspired by grounded theory (Glazer & Strauss, 1979) and followed four pragmatic rules for qualitative research: openness of the researchers; openness of the subject of research; maximal variation of perspectives; and analysis of similarities (Kleining, 1994: 23–35). We studied the most heterogeneous classes in three different schools (two Hauptschule and a Gymnasium). The schools differed in social structure, policy toward multilingualism, and location. The classes were known as ‘troublemakers’. We selected the classes to ensure that we would collect a ‘sample of extremes’ (Kleining, 1994: 30). Along with qualitative interviews, participant and non-participant observations, experiments, data on 61 adolescents between the ages of 14 and 16 were collected in the first stage of our study. We included as many girls as boys and covered a wide variety of linguistic and social backgrounds. Beginning with the abstract idea of ‘language’ and the concrete ‘gestalt’ as, say ‘German’, ‘Turkish’ or ‘Yugoslavian’ language, we wanted to study the everyday life of these adolescents. We expected borders to appear, which correspond to Wittgenstein’s comment: ‘The limits of my language mean the limits of my world’ (Wittgenstein, 1990: para. 5.6) and Cihangir, one of the adolescents interviewed, described them like this (all interview quotes are translated into English): ‘I’d learn a language because a language means one person, a differing person. You can go everywhere and then everything is understood’. He refers to a Turkish speaker saying: ‘Bir dil bir insan – iki dil iki insan’. [One language is one person – two languages are two people.] What is the relationship between these limits or borders and the languages in the Hamburg area? To answer this question we will take a closer look at the multilingual areas of St Pauli and Altona.

Local Speech Communities The Ethnologues counts 6703 languages worldwide. In Europe 225 languages are spoken (Ethnologue, 2000). The national German labour agency (Bundesanstalt für Arbeit) 1994 statistics show 91% of the German population to be German citizens and 9% to be immigrants. Of the immigrants, about 2% are Turks, 2% from former Yugoslavia, and 5% others. On the national level we have on the one hand a huge group of German speakers, and on the other hand tiny groups with minority languages. In the three schools where our study was

44

Bilingualism and Social Relations

carried out, we found a different distribution: 42% of the adolescents had a German background, and 36% a Turkish background, 13% had their background in former Yugoslavia, and 9% mentioned other non-German backgrounds. In the German educational system, schools generally recruit their pupils from the local area. Therefore the distribution of languages in an area is pretty well represented in the area’s schools, particularly in this age group (14–16 years). That means that we can vary the social context by choosing schools and classes in socially different areas. The situation in the school classes differs considerably. In one class there is a large group of students with German background, and a group with Turkish backgrounds almost equal in size, followed by the third largest group from a Yugoslav background. The distribution of the languages in the different classes varies with the social context. In the different types of schools the percentage of adolescents with a German background is greater in the Gymnasium (the academically oriented stream), while the percentage with a non-German background is smaller. At the Gymnasium it is therefore less likely that a speech community of minority languages is formed then at a Haupt-, Real-, or Gesamtschool (the trade-oriented streams). In that sense the inequalities are represented in the schools. Our analysis of the social structure in the schools resulted in three types of representations. One Haupt-Real school has a large number of parents who are small shopkeepers. These small shopkeepers are mainly of Turkish background and mainly do business in an ethnic context. The hierarchy of the German labour market forces migrants especially with a Turkish background into self-employment. By law, the Employment Agency has to offer a job first to Germans, then to EU-nationals, and the job is only offered to non-EU-nationals if no one else takes it. The second Haupt-Real school can be characterised as a traditional working-class school, and the Gymnasium as the school for the children of higher qualified and white-collar workers.

The Street as Public and Liminal Space of Youth The basic part of the following analysis uses the logic of an ethnographical description. We move from a description of communication in the home as a community, or kinship, via the street as a public domain, or public space, to civil or national institutions. The street is a space which shows aspects of transit, liminal space, where young people gather. Adolescence is characterised by not being a child anymore (in the sense of being restricted to the home), but not yet being an adult (institutionalised). In ethnology this stage is called liminoid for pre-industrial and liminal for industrial societies (Turner, 1995). Sociolinguistic pragmatics talks of language-communities (Hartig & Kurz, 1971). Language is related to community. This is central from a sociological perspective. Children are socialised not in the abstract, but in real languages. Peer groups appear as networks which have different gestalts and different ranges. Pairs, friendship groups and gangs are speech communities which vary in size and structure. We have mixed-gender friendship groups in our area, but homogeneous male groups were dominant on the streets. The discovery of sexuality and adolescence mark the beginning of youth and alters same-sex and opposite-sex relations. Adolescents develop different ways of being parts of a

Cultural Orientation and Language Use

45

speech community and of getting in contact with other speech communities. Delimitations happen in very different ways. Goffman (1996) uses the term stigma, about the image of discrediting qualities, and he means it as a relational term. Prestige is, according to Goffman, a positive stigma. Stigma and prestige are two sides of the same coin. In our interviews the speakers related stigmata to physical characteristics such as squinting, to national groups such as Serbs, or to groups such as Ausländer (foreigners) or more precisely, groups for which German has no adequate expression. We focus our study on linguistic processes of inclusion and exclusion. It is clear that the difficulties of expression, mentioned by some of the youths, are a common problem for several adolescents from minority linguistic backgrounds, as well as some from a German background. This should not be mistaken for a difficulty with the German language itself, because for nearly all young people the knowledge of German is taken for granted in their world. In contrast the languages of descent appear problematic for the young people in our study. The difference between German and the minority languages becomes a socially shaping power, which results in a special gestalt among young people of different backgrounds. The production of the internal conditions of communication presupposes the existence of a group as a speech community distinguished from the dominant linguistic surrounding force. Such situationally defined speech communities dissociate themselves from other speech communities by specific speech acts (Searle, 1983). The adolescents reported multiple exclusions by a variety of individuals and groups, who are described as ‘acquaintances’, ‘everybody’, ‘German friends’, ‘Turkish girls’ and ‘brothers and sisters’. The strategy of inclusion and exclusion uses this linguistic difference between the dominant language and the minority languages in the adolescent lifeworld.

In- and Exclusion Through Language Choice and Alternation We now introduce some of our data and its analysis. The multilingual adolescents are aware of the including and excluding function of their language choice. Besides deliberate exclusion there is also deliberate inclusion. Gunnar (German background) and Reza (Farsi background) give an example. Interviewer: Did you hear of people speaking another language than German? Or does this not happen? Gunnar: Some speak Turkish, if both of them know Turkish. Reza: Yes. Interviewer: Does it bother you that you don’t understand anything? Do you feel excluded? Gunnar: It would be nice if I could understand something, somehow. Interviewer: So, it bothers you? Gunnar: Yes, a little bit. Reza: When two people meet and I know them and one is a fellow countryman of mine and the other is a German and he addresses me in Farsi, then I always answer in German. I don’t like to speak in my language in front of a German, because it is unpleasant for him, for the German, because he does not understand it. I’d have a unpleasant feeling too, if two other people

46

Bilingualism and Social Relations

were to speak their language in front of me, if they came to meet me and they had a conversation in their language. I wouldn’t like that very much. I’d say that they should have sympathy and that they should have a conversation in a language I understand as well. Interviewer: So, everybody in the group should understand? Reza: Yes Interviewer: You would always be aware of who is here at the moment and you would adjust to it? Reza: Yes. That’s the best. Mainly that is German. (Gunnar, Reza, 1.9.95, Gymnasium) Both agree that in particular speech events all people involved must understand what is happening. For this purpose Reza switches to another language. If he guesses that a person doesn’t speak the language currently in use, he tries to switch to a language that he presumes everybody at least understands. Nathalie, a girl of German background, who says that she speaks English and Danish, describes her fantasies and experiences made with German and Turkish. Interviewer: In which situation, well if you knew Turkish perfectly, well if you are learning it, where could you use Turkish? Nathalie: I would talk with the others. For example, when we’re going out and all the others are Turks, then they would talk definitely only Turkish. They don’t speak German just because of one German. They do, when they talk with me , but when they are among themselves, then they speak Turkish. (…) Interviewer: Does this, for example, bother you? Nathalie: Naw, doesn’t really matter. When we sit together with friends, then most of the time, always German. Just, I don’t really know, when we’re kind of outside, then they sometimes talk Turkish as well. Interviewer: When they talk Turkish, you feel somehow … Nathalie: Yes, kind of deceived. When they kind of talk about you. (Nathalie 15.8.95, Haupt-Realschool A) She thinks that Turks, as long as they are among themselves, speak Turkish and that the presence of one German does not lead to a switch in languages. On the other hand, out of everyday experience she describes the exact opposite. When someone talks to her, he or she immediately switch into German. Nathalie describes how she gets involved in some type of communication and is cut off from some other types of communication. Ufuk has a different perspective. He describes how speaking German is physically enforced in his everyday life. Interviewer: Is it against languages? For example, when you speak Turkish at school or on the street? It isn’t liked very much. Ufuk: Yes, most of the time it is like that. When you see Turks standing in groups, you immediately want to disperse them. (…) Because, I don’t really know what they are afraid of. Interviewer: In school?

Cultural Orientation and Language Use

47

Yes, at school as well. Well you stand there and once in a while you speak with friends, then the teacher comes for example, some definitely has called him or other pupils and then they immediately start making trouble. They kick or do something else, like: ‘Come here, run away!’ Interviewer: Well, it is not liked, when Turks speak Turkish among themselves? Ufuk: Yes, because they are afraid, I don’t really know, what we are talking about anyway. Yunus: Or some think that we are talking about them. Interviewer: Yes. Yunus: They don’t like it very much. Ufuk: That’s why when Germans or other foreigners are there, who don’t understand Turkish, then we speak German most of the time. (Ufuk, Yunus 1.9.95, Gymansium) Ufuk:

Ufuk and Yunus follow a strategy of submission here. They have a sensitivity for the speech event and the degree of legitimacy of the languages in use. They are aware of the speech community norms around them. The composition of German speech communities is not necessarily very homogeneous, as Cennet demonstrates: Interviewer: About what do you talk in your group? Cennet: About us. What we are doing, where are we going tomorrow and what we are going to wear. Interviewer: In what languages mainly? Cennet: German Alex: Always German. Interviewer: And the other girls who belong to your group as well and who also speak another language? Cennet: They speak only German too. That’s why everybody understands. (…) Yes. It is like that, all of us know a different language. Nobody, well there are two Yugoslavs, but they only speak Yugoslavian or Croatian. There are two Polish girls. They all speak only German among each other. (Cennet, Alex, 17.8.95, Haupt-Realschool A) In other reports we found more heterogeneous compositions of speech communities. The choice of German is therefore not only a strategy of submission, but also a pragmatic linguistic strategy to accommodate adolescents of different linguistic backgrounds. In that sense we can talk of German as a lingua franca. To facilitate communication between those basically monolingual speakers of a community, one language is established, which is the mother tongue of another speech community and is not integrated into the social unit in question. (Hartig & Kurz, 1971) Peter (German background) presents us with another strategy of dealing with inclusion and exclusion. He learnt Yugoslavian, that’s what he called his

48

Bilingualism and Social Relations

language, among his friends before he started secondary school (Haupt- und Realschule). Interviewer: And you unterstand it, Yugoslavian? Peter: Yeah, most of it. But I cannot write it. Interviewer: Understanding and speaking? Peter: Yes. (Peter, 17.8.95, Haupt-Real school B) While the learning of the languages of former Yugoslavia has rarely been observed among the majority of adolescents, quite often most adolesecents have acquired some Turkish. Turkish seems to have a positive stigma among our informants (Auer, 1995). The strategy of learning Turkish can be described as one of undermining the processes of inclusion and exclusion. In the data available, we found a lot of evidence of this strategy. Many adolescents spoke about people who understand or even speak Turkish, but have not been brought up in that language. All these people have close contact with adolescents of Turkish background. Volkan: I have a lot of friends who can speak Turkish, even though they are not Turks. And there are some in my street; at school there are some as well in my class too, one or two. (Volkan 15.8.95, Haupt-Realschool A) Volkan (Turkish background) names different places like the street, the school and the school class, where one can find friends and other people who speak Turkish. Armin (Macedonian-Yugoslavian background) explains that one learns Turkish among friends. Armin: I know those Germans, but they have Turkish friends. I know some, they know Turkish as well. (Armin, 15.8.95, Haupt-Realschool A) Those kinds of people have different nationalities. Emre: Some of them do speak normal. I know an Albanian, he speaks… Interviewer: Normal Turkish? Emre: Yes. He learnt it from us. (Emre, 15.8.95, Haupt-Realschool A) Özgür (Turkish background) says, that some such people even live in his neighbourhood. Interviewer: Do you know people who have learnt Turkish somewhere else, outside of their homes? Özgür: Yes, I know one. Our neighbour. Interviewer: Your neighbour? Where does she come from? Özgür: She is German. Interviewer: …is German? Özgür: Yes, and she speaks Turkish very well. Interviewer: And where did she learn it? Özgür: I don’t know. (Özgür, 15.8.95, Haupt-Realschool A)

Cultural Orientation and Language Use

49

Even for Yunus and Ufuk in the Gymnasium (both Turkish background) it is not surprising that other non-native speakers speak Turkish. Interviewer: Them, where did they learn it? Did they learn it from you? Ufuk: They learnt it from friends, from listening, when we speak to our parents, when they are in our houses for a visit, for example. Yunus: Some of them ask ‘ What’s that called?’ as well. Ufuk: Then they also see the way we are living, what we eat. Most of them like our food, then we offer some. Interviewer: They learn it from you, through contact. They don’t learn it at school or somewhere like that? Ufuk: No, through contact with Turkish friends. (Ufuk, Yunus, 1.9.95, Gymansium) The speaking of Turkish by people with another L1 is thus described by our informants as a widespread phenomenon in the world of our youngsters.

The Use of Turkish We now go into the second part of our study where we get a detailed insight into the status and the use of Turkish among young people of non-Turkish background. Altogether 25 young people, most of them born and raised in Hamburg, took part in the second phase. They taped everyday conversations among themselves. These recordings have been supplemented through discussions about topics dealt with in these conversations, and biographical interviews with the informants. Here we give a short description of our second sample. Most of the 25 adolescents (11 male, 14 female) between the ages of 15 and 23 have spent most of their lives in Hamburg. Eight adolescents are of German, four of Tunesian and two of Iranian background. One girl is a Bosnian, and one boy of Bosnian-Croation background. The rest (nine) are Greek, Polish, Jordanian, Afghanian, Moroccan, Capeverdian, Bulgarian-Dutch, Kazak and Spanish-German background. Their educational background differs: three Hauptschool, three Haupt- and Realschool, two Realschool, two Gesamtschool, four Gymnasium, and seven professional schools. One informant dropped education, one was a labourer, and two went to a special school. We have analysed the conversations using the procedure of conversational analysis (cf. Auer, 1983). This means that we were not interested in the structural side of language-mixing (in the mainstream research literature about language alternation, research which analyses this phenomenon grammatically and structurally dominates (cf. Haust, 1993)), but in the use of interactive strategies which made the young people talk in a mixed language. We perceive language alternation as a privileged achievement, as a specific linguistic ability of multilingual persons. In the tapes produced by the young people, they used the possibility of alternation between Turkish and German intensively. The range of Turkish knowledge ranges from nearly perfect monolingual use of Turkish to the transfer of some Turkish expressions into German. We will now present some examples of these two extremes from our data.

Bilingualism and Social Relations

50

Example 1: Conversation in Fikret’s snack bar (for rules of transcription see Appendix) Hans: German, 23 y, m.; Adnan: Turkish, 19 y, m.; Fikri: Turkish, 52 y, m.; Mesut: Turkish, 20 y, m. Hans: ay antonstraße’de biri oturuyo (.) o da galip’in arkadaêÏ galiba ercan bana bugün dedi Oh, somebody lives in Anton Street. It seems to be a friend of Galip, Ercan told me today.

Fikri: Hans:

he Hm. yepyeni ford mondeo duruyo abi yepyeni A brand new Ford Mondeo in front of the house, brand new.

Fikri:

hm Hm.

Hans:

valla yepyeni yani daha yeni aldÏar It is brand new, they’ve just bought it recently.

Fikri: Hans:

Fikri:

( ) en yeni tip yani o da (.) o:de kennzeiche o ciplerde var ya (.) o da o:de The newest model, it had an OD-registration, like those jeeps. It’s got an OD too. he Hm.

Hans:

<> abi o mondeoya bi arkadan geçirmiêler> 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>[<] gül vi skal have orden på dem. are they smile we have to get them in order (they laugh nervously). er du vred Asiye. are you angry Asiye. arkasÏnÏn êeyini yapacaÈÏz mÏ ay. should we do that thing on its behind no. Asiye da pek fena deÈil mi tut êurayÏ # tænker kun på pik. Asiye is pretty bad, too, isn’t she, hold this # she only thinks about dick (shuts her microphone and whispers the last part). nej det er dig der gør. no you do. det hørte du vist heller ikke vel det hørte du vist heller ikke. you didn’t hear that either, did you, you didn’t hear that either.

In this excerpt Asiye challenges Esen’s position in the group. The content of Asiye’s long utterance (line 1–3) is that Esen has shown someone a letter in which Asiye has written a name, apparently a boy’s name. Asiye’s tone of voice indicates that she accuses Esen. The occurrence refers to an episode earlier in the conversation, where Esen and Selma tease Asiye about a boyfriend against her denial that she has one. Esen answers the accusation in this excerpt with a return-utterance, mm, and Asiye continues in line 8–9. After Asiye’s accusation there is silence and she starts humming. In line 13 Esen defends herself in Danish: det var Selma der sagde at den ville komme igen. Hereby Esen draws Selma into the conflict. Esen tries to move the blame away from herself to Selma. She chooses to speak Danish even though she was addressed in Turkish. Selma protests in line 14 claiming that the responsibility lies on Esen, too. She speaks Danish. Danish is a marked choice by Selma and Esen here and it creates a distance not just to Asiye and her preferred language choice but also to her accusations. Esen and Selma both defend themselves in lines 15–16. Then Esen changes the subject to the task: vi skal have orden på dem. During this passage Esen and Selma giggle nervously. Shortly hereafter Esen asks Asiye in Danish if she is angry (line 19). Asiye does not respond but ignores the question and instead she asks a question (in Turkish)

Language Choice in Bilingual Adolescents Conversations

85

to the task (line 20). Esen also ignores Asiye’s question, but she too switches language from Danish to Turkish. She casts Asiye as a listener and the utterance: Asiye da pek fena deÈil mi tut êurayÏ must accordingly be addressed to Selma. After a break Esen adds in Danish: tænker kun på pik (line 22). The switch from Turkish to Danish intensifies both Esen’s teasing and effect of the taboo-word. Asiye protests towards the harsh insult and switches to Danish in line 24. But Esen continues her teasing by referring to the fact that they are being recorded and that the grown man may have heard. In this passage of the conversation Esen is put in an awkward position because of Asiye’s accusation. Her first strategy is to cast Selma as the guilty one, but Selma refuses to take all the blame. Esen and Selma choose to speak another language than Asiye, who has uttered the accusation in Turkish. The fact that they both speak Danish keeps Asiye at a distance. Esen’s gül in line 17 and her question in Danish, er du vred Asiye (line 19), can be considered as an apology or an attempt for reconciliation. When Asiye ignores her, she silently accepts the apology. This is the end of the interchange. Esen’s next move is to insult Asiye. She starts the insult in Turkish, takes a break and when Asiye does not respond, Esen switches to Danish and thereby she increases the provocation. By using different strategies, such as casting, alliance, and code-switches Esen succeeds in manipulating her way back to the powerful role in the conversation. Esen uses her switches between Danish and Turkish in the negotiation of the conflicts in this situation. She switches to the language choice in opposition to the choice made in the resent utterance. The reactions, seen in the example, and the context show, that her choices accentuate the utterance in the context and increase her threats, so that she gets reactions. She increases the power of her own position in the conversation through her language choices. The conversation displays how the informants establish their position in relation to the two languages. But they also manage this through their ability to switch between the languages. This is the only example in the conversation where Esen’s position is threatened. In the rest of the conversation she is in charge and she uses her language resources to have her way and to keep the other girls from having theirs. In this case, where Esen is having problems for the first time, she uses switches between Danish and Turkish in the power negotiations in the situation. She switches to the language opposite of the previous utterance. The reactions seen in the excerpt and in all of the conversation, shows that her choice of language brings out her utterance in the context and increases her threats so that she receives a reaction. Thus increasing her position in the interaction only through her language choice. Esen is the most powerful person in the conversation, and Asiye is without doubt the person with least power. Esen is more capable of managing her languages. She doesn’t just choose the most powerful language, Danish, at all times, but instead she makes her choice of language dependent on the situation. Her use of language choice and code-switching in the situation is not just influenced by relations brought along in the conversation. It is even more influenced by the negotiation in the situation. Local factors – or relations brought about in the situation – influence her language choice. The boys use their languages and language choices quite differently from the girls, as can be seen in the excerpt from conversation 701 which has Hüseyin, Erol, Murat, and Bekir as participants. Erol is often the person who brings the

86

Bilingualism and Social Relations

balance of power to negotiation. All of the four boys have a preferred language – Erol’s preferred language is clearly Turkish (in this conversation). There is as clear a preferred language with the other boys, but Murat and Bekir have more utterances in Danish, while Hüseyin has more in Turkish. Conversation 701 (Murat, Bekir, Hasan, Erol) excerpt transcribed according to the CHILDES conventions, MacWhinney (1995), with Turkish in italics: Erol: English: Bekir: English: Murat: English: Bekir: English: Murat: English: Bekir: English: Hüseyin: English: Murat: English: Bekir: English: Murat: English:

[<] okay jeg prøver. circle mother okay I’ll try. cirkel tamam ah yaptÏm ya bitti jeg prøver siger du så. circle, okay, oh I have made it, it is done, I’ll try, you now say. bu ne yapacak salak. what should this one, you idiot. o salak yapacak # êimdi xxx tekrar. that fool makes it # now repeat xxx (laughs). êimdi bu beêkant yapsÏn. now he should make a pentagon. det kan han ikke finde ud af. he can’t find out. ha siktir len. oh fuck off man. neyse onkant. oh, well, then a decagon. du skal ikke lave trekant det er jeg i gang med. you shouldn’t make a triangle, I am doing that. nej vi gør ikke ben ne yapacaÈÏm. no, we don’t, what should I do.

The excerpt illustrates how Murat and Bekir use their choices of language to control Erol and his interests in the conversation. In the first utterance Erol accepts an order (line 1) given by Bekir. But his response, okay jeg prøver, is ambiguous – he does not say that he will do as Bekir tells him to, only that he will try. This is how Erol starts a playful conflict with Bekir. With first part of Bekir’s response (line 3), cirkel tamam ah yaptÏm ya bitti, he acts as if he was done with the job he was supposed to do. By doing that he strengthens his own position in the negotiation now taking place. At the same time he puts Erol in a position to retort. The next part of his utterance is an attack on Erol, jeg prøver siger du. The utterance amounts to an intrasentential code switch by which Bekir states his own position in Turkish and attacks Erol in Danish. The first part of the utterance is in opposition to Erol’s choice of Danish, and the last part of the utterance has the opposite language choice from the first part. In this way Bekir’s intrasentential code-switch is accentuated and the entire utterance is accentuated in the situation. In line 5 Murat enters the negotiations by attacking the part Erol has in the interaction. His attack is a question and by speaking of Erol in third person he addresses his utterance to Bekir and casts Erol as a listener. This is marked not just by the third person, but also by a code-switch. His language choice – Turkish – is the opposite of the one used by Bekir in his criticism of Erol. And when Bekir answers (line 7), he, too, speaks about Erol in the third person and in Turkish. In

Language Choice in Bilingual Adolescents Conversations

87

this way both Murat and Bekir cast Erol as the listener, and they unite in the casting of Erol and in their language choice. The end of the exchange is Bekir’s response, det kan han ikke finde ud af (line11), to Murat’s suggestion. Bekir’s switch to Danish is the end of his exchange with Murat using the same language – Turkish. The next part of the excerpt involves a negotiation among the two most powerful adolescents in the conversation. Murat gets the last word in line 20. But in this utterance he asks, ben ne yapacaÈÏm. This can be the winner’s attempt to save the loser’s face. Murat and Bekir succeed in keeping Erol under some control. Erol’s status in the interaction is different: he brings power relations to negotiation more often than the others do. He starts conflicts that lead to negotiations of the social relations in the interaction. Even though he starts most of the conflicts, his position in the conversation does not seem to be affected in any positive direction. The relative status of a speaker can be brought into every situation in the interaction, and then reiterated, and this conversation has several passages of overt negotiations. The reactions Erol gets on his initiatives, and the lack of change in his position in the conversation, suggest that his status could be brought along into the conversation. To influence the conversation he has to challenge the balance of power over and over again, and he does that by starting a lot of more or less playful controversies: partly by giving orders, reprimands, and suggestions, and partly by choosing the opposite language as the one used in the previous utterance. The other participants also use language choice to accentuate utterances in negotiations and controversies. The reactions and the contexts in this conversation and its negotiations suggest that the opposite language choice is used for emphasis, a tool to set the agenda and gain more power in the conversation. Furthermore language choice can be used to cast other participants as a listening or overhearing part in the conversation. Thus language choice alone can be a means to change or re-negotiate the balance of power. In the girls’ conversation language choice functions differently in the conversation. Esen was the winner because she was capable of using the switch between Turkish and Danish to manipulate the other participants. She chooses language and switches language influenced by the negotiation itself. Thus her language choice is attuned to local factors more than outside evaluations. Esen is the only one of the young bilinguals whose bilingualism enables her to use the languages to constitute power no matter what status they have outside the conversation. This makes her even more powerful. The adolescents in Grade 7 are able to establish a position for themselves by using Turkish and Danish. They choose language and switch between languages in the conversations. They use both their languages and thus establish – to a larger degree than in the previous grades – themselves as bilingual as opposed to the surrounding society. At the same time they use their languages – the language choices and the language switches – in the social interplay of the interaction. Thus the two languages are not just a way of establishing a position on the outside but also a means to define and re-negotiate relations and hierarchies within the group. I have observed power negotiations taking place in the conversations. Each of the conflicts that I have studied can be seen as a negotiation of power relations between the participants. I have also found that language choices can be used in

88

Bilingualism and Social Relations

these negotiations. The informants choose, when they are younger, to a very large extent their native language – Turkish – in group conversations. In their later grades they increasingly use both Turkish, Danish, and switches between the languages. In this way they establish their identity as bilinguals. The point is that they establish themselves in opposition to an outer society which to a large extent understands itself as monocultural and monolingual. We have seen how the adolescents use their language choices and language switches as means to solve controversies, and thereby as a power resource in the negotiations. Language choices and language switches are used increasingly as a pragmatic power resource over time. The significance of this power resource has changed through the years. The language use has changed from being determined by differences in status between the languages so that one language was the power language, and the other language was the solidarity language. Now the language choices and the language switches are made as parts of negotiations in the situation, and under the influence of what happens in the situation. The bilingual adolescents not only use their two languages to establish themselves in opposition to an outer society but also to establish themselves within the group, e.g. to establish a hierarchy. Language choice and the choice of two languages are used in a social interplay within the group. Both languages, the choices, and the switches between them are used in the conversations as a part of a power play where all participants accept the rules. From my analysis I have found that the rules apply generally. Thus the adolescents master the rules and know them already in Grade 1, but we cannot say from where the adolescents have this knowledge. This indicates that some rules are brought along the conversation from the outer society. At the same time Danish, Turkish and the switches between them are used to define identity and to (re-)negotiate the power relations in the conversation. Thus the rules are not invariable, but they are made and redefined over and over again in the running negotiation in the situation. With these analyses the view of the social structures as causes of linguistic variation is weakened – especially on behalf on modern sociolinguistic research. According to this research the relations between the outer reality and social relations and speech community are constructed in the interaction due to negotiations and re-negotiations. Correspondence Any correspondence should be directed to Trine Esdahl, Department of Nordic, University of Copenhagen, Njalsgade 120, DK-2300 Copenhagen S, Denmark ([email protected]). References Auer, P. (1995) The pragmatics of code-switching: A sequential approach. In L. Milroy and P. Muysken (eds) One Speaker, Two Languages. Cross-disciplinary Perspectives on Code-switching (pp. 115–135). Cambridge: Cambridge University Press. Auer, P. (1998) Introduction: Bilingual Conversation revisited. In P. Auer (ed.) Code-switching in Conversation. Language, Interaction and Identity (pp. 1–24). London: Routledge.

Language Choice in Bilingual Adolescents Conversations

89

Boyd, S. (1985) Language survival: A study of language contact, language shift and language choice in Sweden. In Gothenburg Monographs in Linguistics 6. Göteborg: Department of Linguistics, University of Göteborg. Cameron, D. (1997) Performing gender identity: Young men’s talk and the construction of heterosexual masculinity. In S.A. Johnson and U. Meinhof (eds) Language and Masculinity (pp. 47–64). Oxford: Basil Blackwell. Coates, J. (1993) Women, Men and Language (2nd edn). London: Longman. Esdahl, T. (2001) Fokus på kodevalg og magtudøvelse – empirisk baserede samtaleanalyser af tosprogede unges sproglige adfærd. Unpublished graduation thesis, Copenhagen University. Goffman, E. (1972) On face work. An analysis of ritual elements in social interaction. In Interaction Ritual (pp. 5–45). New York: Pantheon. Gumperz, J.J. (1982) Discourse Strategies. Cambridge: Cambridge University Press. Hudson, R.A. (1996) Sociolinguistics. (2nd edn). Cambridge: Cambridge University Press. Johnson, S.A. (1997) Theorizing language and masculinity. In S.A. Johnson and U. Meinhof (eds) Language and Masculinity (pp. 8–26). Oxford: Basil Blackwell. Jørgensen, J.N. (2001) (ed.) En køn strid. Københavnerstudier i tosprogethed, Køgeserien bind K10. Frederiksberg: Roskilde Universitetsforlag. Jørgensen, J.N. (1998) Chidren’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. MacWhinney, B. (1995) The CHILDES Project: Tools for Analyzing Talk. (2nd edn). Hillsdale: Lawrence Erlbaum. Madsen, L.M. (2002) De som har kan få. Københavnerstudier i tosprogethed. Køgeserien bind K12. København: Danmarks Pædagogiske Universiteit. Møller, J. (2001) Identitet og kodevekslen hos unge tosprogede med dansk- tyrkisk baggrund. Københavnerstudier i tosprogethed. Køgeserien bind K8. København: Danmarks Pædagogiske Universitet. Ng, S.H. and J.J. Bradac (1993) Power in Language. Verbal Communication and Social Influence. London: Sage. O’Barr, W.M. (1982) Linguistic Evidence: Language, Power, and Strategy in the Courtroom. New York: Academic Press. O’Barr, W. and B. Atkins (1980) ‘Women’s language’ or ‘powerless language’? In S. McConnell-Ginet, R. Borker and N. Fulman (eds) Women and Language in Literature and Society (pp. 93–110). New York: Praeger. Rampton, B. (1995) Crossing. Language and Ethnicity Among Adolescents. London: Longman. Rampton, B. (1998) Language crossing and the redefinition of reality. In P. Auer (ed.) Code-switching in Conversation. Language, Interaction and Identity (pp. 290–317). London: Routledge. Romaine, S. (1994) Language in Society. An Introduction to Sociolingvistics. Oxford: Oxford University Press. Scotton, C.M. (1988) Code switching as indexical of social negotiations. In M. Heller (ed.) Codeswitching. Anthropological and Sociolinguistic Perspectives (pp. 151–186). Berlin, New York, Amsterdam: Mouton de Gruyter. Sebba, M. and Wootton, A. (1998) We, they and identity: Sequential versus identity-related explanation in code-switching. In P. Auer (ed.) Code-switching in Conversation. Language, Interaction and Identity (pp. 262–286). London: Routledge. Turan, F. (ed.) (1999) A text collection of Turkish-Danish bilingual grade school students’ conversation. Copenhagen Studies in Bilingualism. The Køge series K6. Copenhagen: Royal Danish School of Educational Studies.

Power Relationships, Interactional Dominance and Manipulation Strategies in Group Conversations of Turkish-Danish Children Lian Malai Madsen Institut for Nordisk Filologi, University of Copenhagen, Denmark This paper deals with interactional dominance and power wielding in bilingual conversation among school children. I find that different pragmatic strategies are used by bilingual children as a mean of negotiating power relationships and identities but that the social relations and the power bases brought into the conversations by the interactants make the outcomes of the negotiations predictable. Furthermore, I have found differences related to gender in the use of interactive dominance strategies and the handling of conflicts. I conclude that the traditional view of gender as fixed social categories does not fit my data and that it is more appropriate to regard gender as a linguistic and social construction.

Introduction A conversation is a social event in which the participants perceive themselves as both individuals and as members of a group. In a conversation the interactants negotiate social relations and identities. Thus, every linguistic act can be seen as a social act. My study concerns the construction of gender-identity and linguistic power wielding. My data consists of conversations between bilingual children and adolescents. In bilingual conversations one of the pragmatic linguistic means of negotiating power relations and identities is of course the choice of language. However, code switching is not my main object of study. I focus on the distribution and the types of speaking turns as a measure of the power-processes in the negotiations and furthermore I focus on the outcome of the conflicts in the conversations as a measure of power. My research (Madsen, 2001) concerns both linguistic and social parameters. Thus I consider the relationships between linguistic variation and social structures. As a starting point I see linguistic variation as a mean of negotiating power-relationships and identities. The linguistic choices bring about social relations in the conversation. My work, however, suggests that linguistic behaviour in the conversation to some extent depends on brought-along social factors (Rampton, 1995).

Language and Power In the social sciences the concept of power is used to describe an aspect of human social behaviour. Basically, power is the potential to carry out one’s own will despite the opposing interests of others. Language is the primary means of power wielding in our society. Olson and Cromwell (1975) divide the concept of power into three dimensions: power bases, power processes and power outcomes. The term power bases refers to the resources possessed by a person which improve his or her ability to influence or control other people or prevail in 90

Linguistic Power in Conversations

91

a conflict. These power bases can be institutionalised positions or personally acquired. The term power processes refers to a the use of power bases, the power holders self-assurance and control in given situations. In a conversation the power processes concern, for example, the degree of interactive dominance. Influence and control of a conversation can be achieved in different ways. Taking the floor, controlling the themes by keeping or changing the focus of the conversation, and taking or casting different conversational roles are all ways of power wielding. Finally the term power outcomes concern the results of disputes and discussions in the conversation. To have one’s way or to avoid losing face in a face-threatening situation (Goffman, 1972) are power outcomes. These three dimensions of power are useful in the study of language and power. First and foremost because the distinction between power processes and power outcomes is an important one in the study of linguistic power wielding and power relationships in a conversation. Many studies of language and power are in an stylistic perspective (Bradac & Mulac, 1984; Hosman, 1989; O’Barr, 1982). In a study of power relations among Turkish family members in the Netherlands, Huls (2000) studies competitive turn-taking patterns and she asks the family members about who makes the final decisions in the family. In that way she considers both power processes and power outcomes. In this work I focus on the pragmatic use of language. I study the power processes through an initiative-response analysis and, instead of asking the participants, I study the power outcomes by analysing the occurring conflicts in the conversations to decide who gets their way. Inspired by Jørgensen (1993) who studies code-switching as a power tool and manipulation skills in the conversations from the Køge Project, I also describe some of the principles of manipulation used by the children in my data.

Language and Gender Linguistic differences between the two genders is explained by scholars in different ways. The main views are based either on the premise that women are a repressed group in society or the premise that men and women belong to different social subcultures. The first view claims that the fact that men are dominant while women are submissive explains the differences between the languages of men and women (Coates, 1993). The other view claims that the differences between the languages of men and women reflect the subcultures and social reality (Romaine, 1994). Sørensen (1989) studies girls’ groups and boys’ groups, and she finds that boys’ groups are hierarchical. The boys compete, and it is important for them to obtain and hold a powerful position within the group. Girls’ groups have flat structures, and here it is important to maintain a position as a member of the group. The girls try to be friends with everyone in the group so that they never end up as outsiders. It is better if someone else gets excluded from the group. Sørensen finds that boys’ groups are based on action, while girls’ groups are based on talk. Goodwin (1980) likewise concludes that the boys exert physical dominance: they compete in sports and fights. Girls, on the other hand, exert dominance through language. They gossip instead of fighting. They have a more

92

Bilingualism and Social Relations

indirect way of manipulating than boys and they often criticise girls who are not present. Johnson and Meinhoff (1997) show a less categoric view of the relations between language, gender and social reality. They consider gender a linguistic construction. Social structures are thus a result of linguistic variation. The way one speak can signify gender in the same way as, for example, choice of clothing. In this way linguistic variation can be used to build social differences and similarities.

Data My data is part of a longitudinal study of the linguistic development of bilingual Turkish-Danish children in a Danish school (the Køge Project, see Turan, 1999). The project has collected a large amount of data from the children’s school start until they finished school nine years later in 1998. The data used are the group conversations. The conversations involve three to four participants each and are either pure boys’ conversations, girls’ conversations or gender-mixed conversations. At the opening of each conversation the children are given an assignment. The conversations dure around 45 minuts, they have been audio-recorded and transcribed according to the CHILDES-conventions (MacWhinney, 1995). I have selected two to three conversations from each of the first eight grades. The selection of data was made in connection with a larger project (see Jørgensen, 2001; Madsen 2001).

The Quantitative Study of Linguistic Power Wielding in Conversations In operationalising linguistic power wielding I consider the distinction between power bases, power processes and power outcomes. I study the power processes by analysing the interactional dominance and the conflicts in the conversations. In the study of interactional dominance I combine a quantitative initiative-response analysis with studies of turn-taking related to attachment, reception and competition. Thus I have been able to get a picture of the quantitative, topical and interactional dominance of the children, and furthermore I have been able to compare the power processes with the outcomes of the conflicts.

Coding and Categories Every utterance in 18 of the conversations taken from the Køge Project is categorised in several dimensions. One set of categories is developed from the initiative-response concept from Linell and Gustavsson (1987), adapted for group conversations. I have added the two dimensions of reception and turn-taking. This gives me six variables: The regular initiative-response categories, scope of links (local vs. non-local), speaker links (alter- or selflinked responses) and focality of links (focal vs. non-focal), reception, and finally turn-taking. Thus I have been able to categorise on a relatively complex level although my set of categories is less differentiated than Linell and Gustavsson’s.

Linguistic Power in Conversations

93

Initiatives and responses I use a simplified set of the initiative-response categories defined in Linell et al. (1988). Initiatives point forward toward the next turn and the response point backwards to preceding turns. Responding initiatives point both back and forward in the conversation. Minimal responses only respond to an initiative. I use these concepts and furthermore I register whether responses are locally or non-locally linked, and whether they are focal or non-focal. Whether an utterance is local or non-local depends on what it responds to. The local response is linked to an utterance which is not further away than the number of participants in the conversation minus one, namely the speaker (Linell, 1990: 47). The focal response is linking up with focal aspects of an interlocutor’s turn as opposed to the non-focal response which links up with peripheral aspects. It is possible for a participant to link a response to both the speaker’s own preceding turn or to one of the interlocutors’ turns. Finally I note whether or not an utterance receives verbal response. An utterance is considered an interruption if one participant begins to speak while another participant is still speaking; open turns that allow all participants to take the floor are scored as such. In case the nominated next speaker takes the floor, this is registered as taking one’s own turn, whereas a turn is a stolen turn if another participant than the nominated speaker takes the floor. New initiatives and non-focal responses are in this connection considered strong attempts to influence the conversation, because when a participant often contributes new initiatives and non-focal responses he or she exerts topical dominance. Those participants who succeed in making the other participants follow up on the subjects introduced are thereby in control of the conversational subjects. Stolen turns and interruptions can be seen as direct means to take over the conversation. By stealing the turn or by interrupting, the participants exert power to change the intended distribution of speaking-turns. A great amount of received responses indicates actual dominance in the conversation. Linguistic behaviour like initiating or responding with non-focal linked responses and thereby introducing new subjects indicates intended dominance, but it is not successful if the attempt is ignored by the other participants. To be ignored by the other participants is like being shut out of the conversation, thus the less dominance-characteristic utterances are the ones that do not get any response.

The Results of the IR-analysis The IR-analysis resulted in two main points. Firstly, we see that the girls in their linguistic behaviour appear to be more competitive than the boys in the second to fourth grade, and on the other hand the boys’ linguistic behaviour is more coherent than that of the girls. The competitive behaviour of the girls can be observed in a greater amount of dominance-characteristic linguistic features, such as new initiatives and non-focally linked utterances. Conversely, the boys have a greater amount af focally linked utterances as well as responding initiatives (see Table 1). Secondly, my results indicate that the girls who participate in both single-sex and gender-mixed conversations change their linguistic behaviour in the direction of a more coherent behaviour similar to that of the boys when they participate in conversations with the boys. Thus in the mixed conversation we see that

94

Bilingualism and Social Relations

Table 1 The average percentage of the responding initiatives, new initiatives focaland non-focal links from girls and boys in the non-mixed conversations from the second to fourth grade Variant (average percentage of the total amount of utterances) Responding initiatives New initiatives Focal links Non-focal links

Girls 69.1% ** 16.6% * 67.6% ** 11.8% **

Boys 76.9% 9.9% 78.2% 8.1%

Note: ** indicates significance (p < 0.01) and * indicates (p < 0.05).

the girls in gender-mixed conversations make less use of new initiatives, non-focally linked utterances, and self-addressed turns than in the girl-conversations. The boys show no statistically significant differences between their linguistic behaviour in the mixed- and the non-mixed conversations. Finally we find no statistically significant differences between the boys and the girls who participate in the gender-mixed conversations. Thus it is only the girls who seem to act differently when the two gender interact. Their behaviour in the mixed conversations is less dominance-characteristic than in the girl-conversations but this is not caused by the boys’ conversation being dominant (see Table 2). Table 2 The average percentage of new initiatives, non-focally linked utterances, own turns, and focally linked utterances with the same five girls in mixed and non-mixed conversations Variant (average percentage of the total amount of utterances) New initiatives Non-focal links Own turn Focal links

Girl conversations

Mixed conversations

14.1% * 12.4% ** 19.6% ** 69.7% *

8.4 % 7.7% 11.6% 77.0%

Note: ** (p < 0.01); * (p <0.05).

Thus I have found a gender related connection between the frequency of the use of dominance-characteristic and coherence-characteristic linguistic variants.

Power Outcomes To compare the linguistic power processes and the power outcomes I have supplemented with a quantitative analysis of the conflicts in the conversations. Identifying a conflict, however, involves a qualitative and time-consuming analysis. As a consequence of this I have selected for this analysis eight of the conversations. I have chosen these so that each informant participates in at least two of the conversations, and furthermore so that we have conversations from different grade levels. Gumperz (1982) defines a conflict in the conversation as a speech event in which the participants place themselves in opposition to each other. What I am able to observe in the conversations is not the speech event, but a situation in which opposition is expressed. I define a conflict as a situation in the conversation where opposing interests or values are explicitly expressed. A conflict then is a speech event in which a participant by his or her utterances opposes to one or

Linguistic Power in Conversations

95

more of the other participants, their utterances, or their actions. In my definition a conflict always implies face-threatening acts (Goffman, 1972). According to this definition I have identified and counted all the conflicts in the conversations. I have analysed every conflict by focusing on the outcomes. Every conflict in our data ends up in some kind of solution, unless the participants simply change the subject of the conversation. The conflicts rarely end in a solution which is equally acceptable to all the participants involved. Most of the conflicts result in the loss of face for one or more of the participants. Therefore, it has been possible to count the won and lost conflicts for every informant. I consider a conflict won if it is evident in the conversation that the situation has changed in the direction preferred by one or more of the participants, the winner(s). Likewise I consider it a victory if a participant succeeds in building up his or her own face. This can be expressed by remarks from other participants, or it can happen by another participant’s acceptance (of face-loss) without resistance. A conflict does not necessarily end with only a single winner or loser. The participants can ally against other participants.

Results I have studied conversations from Grades 2, 3, 5 and 7. One issue I wanted to look at was the question whether power relationships change over time, or whether they stabilise at a certain time and then do not change. Table 3 shows the conflict outcomes for three girls and three boys twice in their school career, Grade 2 and Grade 7. Table 3 Conflict outcomes for the participants in the girl- and the boy-conversations in Grade 2 and Grade 7 Grade 2 Esen Selma Asiye Murat Bekir Erol

Total 11 11 9 9 15 15

Won 10 7 7 7 7 5

Lost 1 4 2 1 6 8

Agreement – – – 1 2 2

Total 10 6 8 19 20 20

Won 9 3 0 13 9 5

Grade 7 Lost 1 3 8 4 7 12

Agreement – – – 2 4 3

My results indicates that the power relationships are fixed already by Grade 2 in the sense that power outcomes among both the boys and the girls do not change through the whole period of primary school. Table 3 is one illustration of this. The conversations furthermore seem more asymmetric in power distribution when the participants are girls than when they are boys. There is a greater difference between the girls who often win the conflicts and those that do not, than there is among the boys. It is also characteristic that none of conflicts in the girl-conversations end in a compromise.

Power Wielding and Manipulation Strategies in the 2nd Grade In Table 1 we saw that the girl conversations in the beginning of the school period were more competitive than the boy conversations. In Table 3 we saw that

96

Bilingualism and Social Relations

the power relationships according to power outcomes likewise are more asymmetric in the girl conversation in the second grade and that none of the girls’ conflicts ends in an agreement. In the following we shall look at two examples from the second grade which show how the girls and the boys handle conflicts in conversation. In the first excerpt the four participants in the conversation are the girls Esen, Selma, Asiye and Eda. Their assignment is to prepare a family to go on vacation in Turkey by cutting out pictures of suitable stuff from a catalogue. The assignment is the main subject of the conversation and also the main cause of conflicts. Another recurrent conflict is the three other girls’ bullying of Eda. This example begins with a sound she makes and the other girls’ teasing her by saying that she farted. Excerpt 1 from Grade 2: Eda farted (italic =Turkish) Selma: English: Asiye: English: Esen: English: Eda: English: Selma: English: Eda: English: Selma: English: Esen: English: Selma: English: Eda: English: Esen: English: Asiye: English: Esen: English:

oh Eda osur +… Oh Eda fart +… oh osur osurmuþ oh Eda osurdu # Eda osurdu. Oh fart she has farted oh Eda farted Eda farted (Eda laughs). Eda osurdu. Eda farted. Esen osurdu oh. Esen farted oh. (Eda laughs) ssch terbiyesiz konuþmayýn mikrofonda ben konuêmam. Hush you should not speak naughty in the microphone I don’t speak. Esen osurdu Selma osurdu. Esen farted Selma farted (Eda laughs). hiç de deðil asýl sen. Not at all actually it was you. bir daha seninle birþey olmaz. Another time we don’t want to be with you. he. Yes. ben böyle dedim bak cart öyle ettim biliyor musun sizde söylediniz Eda osurdu dediniz buda cýk cýk seni xxx xxx. I said like this look cart I did like this you know you said that you said Eda farted and this one here cýk cýk you xxx xxx. oh hiç birþey yok. Oh there is nothing at all. benimkide de yok oh bu ne güzel mi Esen. There is nothing in mine either oh how this is beautyful is’nt it Esen. evet bana baþka birþey + … Yes something else for me + …

In line 1 Selma comments on the sound Eda makes oh Eda osur and the two other girls follow up in line 3 and 6. Eda laughs and responds in line 8 by saying exactly the same about Esen. Eda is still laughing which indicates that she has fun. In line 11 Selma tells the others to be quiet and reminds them that the conversation is

Linguistic Power in Conversations

97

recorded, and therefore they should not ‘speak naughty’ even though it was Selma herself who brought up the subject in the first place. Eda continues anyway and includes Selma in the teasing Esen osurdu Selma osurdu (line 13) while she is still laughing. Selma defends herself and is supported by Esen (line 16 and 18). Eda stops laughing and this time she does not return the insult. Instead she tries to explain the sound she made ben böyle dedim. She claims that it was the other girls who said that she farted söylediniz Eda osurdu dediniz(line 22). The others ignore her defence. In line 26 Esen changes the subject by commenting on her catalogue and Asiye follows up the new subject in line 28. The conversation is focused on the assignment for a short while. Then Esen takes up the bullying of Eda again. Excerpt 2 Grade 2: Eda farted 2 Eda: English: Selma: English: Esen: English: Esen: English: Selma: English: Asiye: English: Esen: English: Esen: English: Selma: English: Eda: English: Selma: English: Eda: English: Esen: English: Asiye: English: Eda: English: Asiye: English: Selma:

bende baþka bir ev keseceðim çabuk ol. I also have to cut another house, hurry up. Itamam. Okay. þehirler neler. Cities and so. Eda osurdu. Eda farted (whispering). nasýl da keseceðiz daha [/] oh Eda ben o zaman þu +… How should we cut out more oh Eda so I’ll take it. oh nasýl kesiyor. Oh how is she cutting. Eda osurdu. Eda farted. (Esen is whispering Eda laughs) # ne oldu ne oldu. What happened what happened. oh oh.[>] Oh oh. [<} sizi güldürmek için cart dedim <sizde lâfý büyütüyorsunuz.[>] And I just said cart one time to make you laugh and then you are making so much of it.
98

English: Esen: English: Asiye: English:

Bilingualism and Social Relations

It’s the truth. ya hiç bir êehir bile yok. This isn't even a city. benimkide var Esen aha bak bunlarýý buldum tek. There’re some in mine Esen look I only found these.

Eda, Selma and Esen are speaking about what to choose from the catalogue when Esen whispers Eda osurdu (line 7). None of the others respond to her remark. In line 12 Selma comment on the way Eda cuts out the pictures. In this utterance she speaks about Eda without addressing her directly. In that way she casts Eda in the conversational role as the an overhearer (Clark & Carlson, 1982). A similar casting takes place in the utterance ‘Eda farted’. Esen repeats her remark in line 14, and this time Eda responds by laughing. But in line 21 Eda is defends herself. She expresses how she thinks that they are making too much of her attempt to have fun. In the following part the three girls continue to make fun of Eda, and they continuously cast Eda as an overhearer. The casting functions as a means to establish an alliance between the three other girls and in that way exclude Eda from the group. The alliance of three against one is a strong strategy and Eda does not have the power to defend herself. The conflict ends when Selma states dogrusu (line 33), and Esen changes the subject again. The examples from the girls’conversation shows the asymmetric power relationships among the four girls. Eda is the one who loses the conflicts. The other three girls control the subjects of the conversation and ally against Eda. These examples also shows how different manipulation strategies are used in the linguistic power wielding, alliances being one of them. We see that Esen successfully changes the subject several times. Change of subject is another strategy which diverts the others’ attention. In the first example Edas defence is ignored by the other girls. By ignoring Edas utterance the other girls show that they do not recognise it as a threat. What you do not hear (or seem to hear) you do not have to consider. Another strategy which is often used in the conversation of the children is referring to an authority. In most cases it is the adults outside the room. This strategy is indirectly used by Selma in the first excerpt when she refers to the fact that they are being recorded and says that they should not speak naughty into the microphone. Finally Eda is tries to defend herself by the principle of ‘rather you than me’ though not successfully, when she returns the insult by just changing the names. The strategies found in the conversation relate to the general principles of manipulation described in Kjøller (1991) which can be useful in the study of linguistic power wielding (see Jørgensen, 1993; and Madsen, 2001). The following example is taken from the boys’ conversation in the second grade. The four participants are Murat, Bekir, Erol and Ali. They have been given the same assignment as the girls. Excerpt 3 Grade 2: You are both idiots Bekir: nerde atalým ben alamýyorum þunlarýý atalým bari. English: Let’s throw this one out I can’t take these we had better throw these out. Erol: valla ben kesiyim. English: By God I want to cut.

Linguistic Power in Conversations

Bekir: English: Murat: English: Bekir: English: Murat: English: Ali: English: Bekir: English: Erol: English: Murat: English: Erol: English: Bekir: English:

99

o zamanda ben ala+/. So I can’t ta+/. bize kýsýncata biz yapýêtýramýyoruz. If you get angry with us we can’t glue it on öbürünü ben de alamyorum kafasýz # ama. I can’t take the other one you idiot # but. ama. But. herhalde ikinizde kafasýz. Clearly you’re both idiots. oh. Oh. o zaman size ben veriyim size siz yapýþtýrýn o zaman. In that case to you I give to you so that you can glue it. kendin yapýþtýrsana biz yapýþtýrmayýz. You can glue yourself we don’t glue. dövüþüyorsunuz valla benim için hiç farketmez keserim sussanýz. You’re fighting by God it makes no difference to me at all I can cut if you’ll be quiet. kes ama düzgün kes çünkü baksana bu ev ta bulutlarý bile almýþ. Cut but cut straight because just look this house you even got clouds too.

From the turns right before this excerpt it appears that Erol and Bekir are occupied with finding suitable pictures in the catalogues, and they are connected by this activity. Ali and Murat are glueing, and Murat has just asked Bekir and Erol to remove some catalogues from the table to give more space. In the beginning of the excerpt it looks as if a conflict arises between Bekir and Erol. In line 1 Bekir suggests that they throw out something (probably the catalogues that they were asked to move). Erol responds in line 3 with the intensified expression valla that he wants to cut. Bekir begins objecting but is interrupted by Murat in line 8 bize kýsýncata biz yapýêtýramýyoruz. By saying you get angry with us he marks a distance between you and us. His utterance is directed to Erol. Bekir repeats and finishes his former objection and calls Erol an idiot (kafas9z, line 10). In line 14 Ali joins in the conversation he claims that both Bekir and Erol are idiots. His remark seems to make an impression on the two boys. Bekir cries out oh but makes no further response. Erol agrees to negotiate by suggesting a division of tasks. He will cut out the pictures for the others to glue (line 18). Murat does not accept his suggestion. In line 20 he claims that Erol can do his glueing himself and words the end of his utterance we don’t glue like he is speaking for both Ali and himself. Erol closes the dispute in line 22. He expresses that there is no need to fight (dövüêüyorsunuz). He will cut if they will be quiet. Murat and Ali keep silent and Bekir continues the conversation about what to cut out. The four boys have come to some kind of agreement, and they can take up the work again. This example does not leave us with a clear sense of the power relationships among the boys. As opposed to the girl conversation the power relations seem less asymmetric. The conflict in this example ends in agreement. Furthermore the boys’ use of manipulation strategies seems less complex and less competitive.

100

Bilingualism and Social Relations

Like the girls they ally by casting other participants but in this excerpt we only see argumentation used as a strategy in the negotiation.

Conclusion One of my interests has been the possibility of occurring differences in linguistic behaviour between boys and girls attending Primary and secondary school. An important conclusion is that the girls apparently are characterised marked by more linguistic dominance and competitive behaviour than the boys are. There is no simple conclusion to my study, however. Considering the results, I am able to establish that the gender-related differences in language use can not be described as in classical feministic sociolinguistics. The fact that the girls’ conversations are more marked by competition than that of the boys and that the girls change their language use in the gender-mixed conversations supports a view of gender-related linguistic differences quite similar to one seen in most recent sociolinguistic research. According to this research gender is considered to be a construction and linguistic variation is regarded as means of constructing gender. The children in my study show different ways of constructing their conversation identity – as either boys or girls. The means used by the girls to construct gender are more flexible than the means used by the boys. I can also observe power-negotiations taking place in the conversations. Each of the conflicts that I have studied can be seen as a negotiation of power-relations between the participants. Nevertheless, in the light of my study of power-results, we have to conclude that the power-relations between the informants are unaltered in the eight conversations. The children winning most of the conflicts in second grade also win in third, fifth and seventh grade. The children who most often get their way are those with the highest rank of power. All of these results point to the fact that social structures play an important part in the outcome of linguistic negotiations of power. This isEspecially seen in relation to the discussion of either linguistic variation being a result of social structures, or language being a resource in negotiations and constructions of social reality. The micro-community of the bilingual children of two classes in a Danish school involves certain social structures, which predict the results of power-negotiations. The children use different strategies in the negotiations of their mutual power-relations, but not everything can be negotiated or re-negotiated. Our study confirms that it is not possible to fully depict language as a social phenomenon without considering the social luggage of the language user. The negotiation itself and the language as a resource in the negotiation are both interesting. Nevertheless, the linguistic construction of identity or gender taking place in a social interaction does not count for the entire social reality. People engaged in a social interaction will – if they have met before – be influenced by brought-along experiences when placing themselves and others in mutual relations.

Linguistic Power in Conversations

101

Correspondence Any correspondence should be directed to Lian Malai Madsen, (Stud. Mag), Institute for Nordisk Filologi, University of Copenhagen, Njalsgade 80, 2300 Copenhagen S, Denmark ([email protected]). References Bradac, J.J. and Mulac, A.(1984) A molecular view of powerful and powerless speech style: Attributional consequences of specific languages features and communicator intentions. Communication Monographs 51, 307–319. Clark, H.H. and Calson, T.B. (1982) Hearers and speech acts. Language 58 (27), 274–398. Coates, J. (1993) Women, Men and Language (2nd edn). London: Longman. Goffman, E. (1972) On face work. An analysis of ritual elements in social interaction. In Communication in Face to Face Interaction (pp. 319–363). Penguin. Goodwinn, M.H. (1980) Directive-response speech sequences in girls’ and boys’ task activities. In S. McConell-Ginet, R.A. Borker and N. Furman Women and Language in Literature and Society (pp. 157–173). New York: Prager. Gumperz, J.J. (1982) Discourse Strategies. Cambridge: Cambridge University Press. Hosman, L.A. (1989) The evaluative consequences of hedges, hesitations and intensifiers: Powerful and powerless speech styles. In Human Communication Research 15, 383–406. Huls, E. (2000) Power in Turkish families. In Discourse and Society 11 (3), 345–372. Johnson, S.A. (1997) Theorizing language and masculinity. In S.A. Johnson and U. Meinhof (eds) Language and Masculinity (pp. 8–26). Oxford: Basil Blackwell. Jørgensen, J.N. (1993) Children’s code switching in group conversations. In European Science Foundation Network on Code-switching Summer School (pp. 165–181). Paris: European Science Foundation. Jørgensen, J.N. (2001)(red.) En køn strid. Københavnerstudier i tosprogethed, Køgeserien bind K10. Roskilde: Roskilde Universitetsforlag. Kjoller, K. (1991) Manipulation enhindbog (2nd edn). Copenhagen: Borgansforlag. Linell, P. (1990) Om gruppsamtalets interaktionsstruktur. In U. Nesselbladt and G. Håkansson (eds.) Samtal och språkundervisning (pp. 39–53). Linköping: Universitetet i Linköping. Linell, P. and Gustavsson, L. (1987) Initiativ och Respons Om dialogens dynamik, dominans och koherens. Linköping: Universitetet i Linköping. Linell, P., Lennart G. and Paivi, J. (1988) Interactional dominance in dyadic communication: A presentation of initiative-response analysis. In Linguistics 26, 415–442. MacWhinney, B. (1995) The CHILDES Project: Tools for Analyzing Talk (2nd edn). Hillsdale, NJ: Lawrence Erlbaum Associates. Madsen, L.M. (2001) De som har kan få. Unpublished Graduation Thesis. Copenhagen: Copenhagen University. O’Barr, W.M. (1982) Linguistic Evidence: Language, Power, and Strategy in the Courtroom. New York: Academic Press. Olson, D.H. and Cromwell, R.E. (1975) Methodological issues in family power. In R.E. Cromwell and D.H. Olson (eds) Power in Families (pp. 131–150). New York: Wiley. Rampton, B. (1995) Crossing. Language and Ethnicity Among Adolescents. London: Longman. Romaine, S. (1994) Language in Society. An introduction to Sociolinguistics. Oxford: Oxford University Press. Sørensen, A.S. (1989) Pigekultur og pigepædagogik. Odense: Odense Universitets Trykkeri. Turan, F. (1999) A Text Collection of Turkish-Danish Bilingual Grade School Students’ Conversation. Copenhagen Studies in Bilingualism, the Køge series volume K6. Copenhagen: Royal Danish School of Educational Studies.

Adolescents Involved in the Construction of Equality in Urban Multicultural Settings Erica Huls Discourse Studies, University of Tilburg, PO 90153, 5000 LE Tilburg, The Netherlands

Ad Backus and Saskia Klomps Babylon: Center for Studies of Multilingualism in the Multicultural Society, University of Tilburg, PO 90153, 5000 LE Tilburg, The Netherlands

Jens Normann Jørgensen Institut for Nordisk Filologi, University of Copenhagen, Njalsgade 80, 2300 Copenhagen S, Denmark This study makes use of Brown and Levinson's (1978; 1987) theory of politeness. More specifically, it focuses on how the model they lay down as guiding language users is actually exploited in various ways by these users. Therefore, four different operationalisations of the basic hypothesis of politeness theory are proposed, ranging from the possibility that linguistic choices are totally determined by social norms to one allowing considerable freedom of choice for individual language users. The article reports on a questionnaire study carried out with adolescents in two urban multicultural areas: Rotterdam in The Netherlands and Køge in Denmark. In both cities, two groups of adolescents took part: one with a Turkish migrant background and one without a migration background. The questionnaire was based on Hill et al. (1986). Only one of the four groups, the Turkish adolescents in Rotterdam, showed an answering pattern that confirms politeness theory. The Køge adolescents with no migration background were the most extreme in not doing what the theory predicts them to do. Instead, they seem to be claiming and constructing equality. The other two groups of respondents also appear to be involved in a process of levelling; with their linguistic choices, they soften the effects of social hierarchies. All four groups of adolescents attenuate the linguistic encoding of social hierarchies, as far as they even acknowledge them, and thus cast doubt on the universality of the determinative relationship between social context and language, a relationship which has generally been assumed to be a necessary cornerstone in any theory of sociolinguistics.

Introduction Suppose you are having dinner with some people. The food is a bit bland. You think that adding a bit of salt might help, but the salt is at the other end of the table. What will you do? Some of the possibilities are: (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

Gimme the salt. The salt! Can you pass me the salt? Could you pass me the salt? Do you mind passing me the salt? Would you mind passing me the salt? I was wondering if you could pass me the salt? Can I use the salt? May I have the salt? Do you see the salt? The food is a bit bland, isn’t it? 102

Adolescents Constructing Equality

103

(12) I need the salt. (13) May I pass you the salt? The numbers 1 to 12 do not need to be clarified here: non-linguists will accept them easily as requests for passing the salt, while linguists can find the pragmalinguistic foundation for this acceptance in the work of Searle (1998). Number 13, however, is strange at first sight. It comes from a joke told by the British sociolinguist W.P. Robinson at the opening of the Second Conference on Language and Social Psychology in Bristol in 1983. The scene is a British upper-class boarding school, where the pupils are enjoying their meal. Robert wants the salt, but it is out of his reach, and in front of Michael. Robert says to Michael: May I pass you the salt? Michael notices that Robert’s offer makes no sense – given that the salt is out of Robert’s and within his own reach. Michael infers that he should not interpret Robert literally and constructs an indirect meaning: Robert might be alluding to his own need for the salt, thereby indirectly requesting the salt. Michael accepts this indirect request and offers the salt saying: Of course. May I? This anecdote shows that, although it takes a lot of reasoning to interpret number 13 as a request for passing the salt, in certain circumstances it may definitely be one. Thus, the range of possibilities for performing requests is great, much greater than a list we may come up with sitting at our desks. The anecdote points to a phenomenon that also plays a role in number 1 to 12 and one that is central to this study: the choice of a specific request form in a specific social situation, out of an enormous array of possibilities, is indicative of the social relationships between, and the identities of, the parties involved. Here, the pupils of a British upper class boarding school are described as extremely careful, polite and indirect (and also humorous and stiff).1 Although it seems reasonable to suggest that speakers design their utterances to invoke certain social relations and personal identities, the study of the means by which this is done is still in its infancy (Nofsinger, 1991). There are some conversation analytic studies on this phenomenon (e.g. Goodwin, 1987; Maynard & Zimmerman, 1984), but there is no clear insight into how the design of requests contributes to this process. Intuitively, the options in 1 and 2 seem to be linked either to a relationship of equality, or to a relationship in which the speaker is superior to the addressee in status, but not to an unequal relationship in which the speaker is in the hierarchically lower position. Someone who utters 1 or 2 presents himself as straightforward and efficient, rather than as hesitant and roundabout. Option 7, on the other hand, seems to suggest the speaker is a careful person who shows respect for his addressee and allows him a lot of freedom of choice. Although we can make these kinds of intuitive interpretations, the relationship between linguistic forms and their social meanings is dependent on the social context and is not fixed at all: the complex interplay of forms, meanings, relationships and identities is an essential and inherent aspect of human interaction. The present study is aimed at revealing a part of this interplay. In the present study, the focus is on adolescents in two urban multicultural areas: Rotterdam (The Netherlands) and Køge (Denmark). In both cities we worked with two groups of respondents: adolescents with a background of

104

Bilingualism and Social Relations

migration from Turkey (from now on referred to as ‘Turkish’, although they are also Dutch respectively Danish) and adolescents without a migration background. We collected data by means of a questionnaire containing three sets of questions (in correspondence with Hill et al. (1986) and a previous study by one of the authors (Huls, 1991)). The first set was aimed at detecting which part of the total range of possibilities for performing requests is actually in use by the various groups of respondents. The second set provides us with insight into the hierarchies that the respondents observed in their social environment. The answers to the third set of questions show how the adolescents present themselves visàvis others, along a cline of ascribed hierarchical positions. Comparison of the answering patterns of the four groups of respondents allows for conclusions about similarities and differences as to their identity formation. Before we treat the design of the study and the data collection in more detail (see Design of the Study below), we will elaborate upon the theoretical framework of the study.

Politeness Theory According to politeness theory (Brown & Levinson, 1978, 1987), social interaction is characterised by the dilemma of being either too close or too distant to one’s interlocutor. The theory explains some details of language usage as being the result of strategies that people follow in solving this dilemma. The notion of ‘face’ (compare expressions such as ‘To show one’s face’ or ‘To lose one’s face’ (Goffman, 1955, 1967; Lakoff, 1975)) is central to politeness theory. Two aspects of face are distinguished: ‘positive face’ refers to the human need for closeness, sympathy and solidarity; ‘negative face’ refers to the need for distance, respect and individuality. All humans have face wants, both positive and negative, and these face wants can only be satisfied by social interaction. In general, it is in the mutual interest of interacting persons to maintain one another’s face. Politeness is the notion that covers all communicative behaviour that is directed at maintaining face. Some acts, however, intrinsically threaten face. A request is an example of an intrinsically face-threatening act: the speaker intrudes upon the addressee by asking him or her to perform an act. In such a situation, the speaker will wish to minimise the threat. He or she can do this in a lot of different ways, for example by ‘embellishing’ the request with terms of endearment (Close the door, my dear), by paying attention to the addressee’s need for not being intruded upon (Would you mind closing the door?) or simply by hinting, thereby offering the addressee a lot of possibilities for interpretation, one of which is escaping the request (There is a draft here.) In other words, when people approach one another, they make strategic choices: they can be direct or indirect, they can pay attention to the addressee or show respect, they can be straightforward or hedge. In total, Brown and Levinson distinguish 40 different politeness strategies, which can be ordered in five main groups: (1) direct and without redress; (2) direct, but with redress aimed at creating solidarity; (3) direct, but with redress aimed at showing respect;

Adolescents Constructing Equality

105

(4) indirect; (5) abandoning the idea to say something. These five main groups show an order in terms of directness, but they are also ordered on another dimension, one of degree of closeness, which can be represented as a continuum between the extremes representing ‘to approach’ and ‘to keep one’s distance’. When someone chooses a strategy with a low number, she or he tries to come closer; when someone uses a high-numbered strategy she or he keeps her distance. Whatever choice one makes from the range of possibilities, every choice has certain advantages and disadvantages. When one chooses a direct formulation, there is the risk of being seen as rude. Indirect means are open for misunderstandings. Showing respect can be interpreted as ‘stiff’. Being kind can slide down to being too kind and thus become embarrassing. Thus, strategy choice is a delicate matter. It is not simply a question of selecting an alternative out of the whole range of possibilities, but also a question of how to be sensitive to the social context and how to present oneself. According to Brown and Levinson’s politeness model, three contextual variables play a role in strategy choice: (1) The power relationship between the speaker and the addressee. If the speaker has less power than the addressee, she or he will be relatively indirect. If, on the other hand, the speaker holds a status that is superior to that of the addressee, she or he can be relatively direct. (2) The imposition the act entails. When you ask someone if you can borrow a pen, the imposition is much smaller than when you ask for her car. (3) The social distance between the speaker and the addressee. When speaker and addressee meet each other regularly, they can approach each other in a relatively direct way; with an unknown interlocutor, an indirect alternative is more probable. These three contextual variables together cumulatively determine the weight of a verbal act: Weight = Power + Imposition + Social Distance. People start their interactions with an assessment of W. Depending on this assessment, they will start to say something or abandon this idea. If they are going to say something, they can do this either directly or indirectly. If they choose to be direct, they can present the message as direct as possible or make use of ‘embellishment’. In case of ‘embellishment’, they can choose between solidarity means or respectful means. Figure 1 presents politeness theory in a nutshell.

Strategy Choice and Social Identity Essentially, politeness theory can be interpreted as a deterministic model: three variables in the nonlinguistic context are assumed to determine linguistic choices. This interpretation of politeness theory fits well with the way of observing and analysing language behaviour that was dominant in sociolinguistic practice until quite recent (see e.g. Holmes, 1992; Labov, 1972; Trudgill, 1974). However, even a small-scale confrontation of the theory with empirical data makes clear that people are creative, rational, and innovative beings who do not let social differences determine all they say and do. By using

Bilingualism and Social Relations

106

1 Without ‘embellishment’

Small Direct

W = P + + SD

2 Solidarity means Say something

With ‘embellishment’ 4 Indirect

3 Respectful mean

5 Do not say something Great

Figure 1 Politeness theory in a nutshell (After Brown & Levinson, 1978)

the slogan ‘Sociolinguistics should be applied pragmatics’, Brown and Levinson (1978: 286) admit this. It is important to realise that their politeness theory, with its roots in pragmatics and anthropology, focuses primarily on regularities in patterns of language production and does not have the intention to fix people and their relationships firmly in social contexts. Politeness theory is flexible in at least two respects. First, though the contextual variables ‘power’, ‘imposition’ and ‘social distance’ influence strategy choice in a universal way, the assessment of these variables is culturally and subculturally specific. They do not predict the use of a specific strategy, but a range from which people can make choices. Politeness patterns that appear to be vastly different, such as the Japanese and the American systems, can thus be explained in the frame of the same universal politeness theory (Hill et al., 1986). Politeness theory has the ambitious goal of offering an ‘etic’ set of concepts in terms of which politeness can be analysed in ‘emic’ terms for any particular society or social group (Brown, 2001). Second, the influence of ‘power’, ‘imposition’ and ‘social distance’ is a joint one. Once the strategy has been chosen, one cannot reconstruct the context in terms of the separate variables. This offers people the opportunity to play with these contextual variables by means of their strategy choice. In this play, these contextual variables can be changed (Huls, 2001; Scollon & Scollon, 1995). The strategic choices people make can thus, in turn, redefine social and cultural aspects of the context. Relationships change and develop, and strategy choice plays a role in this. Brown and Levinson (1978: 233–234) treat this aspect of communicative behaviour under the heading ‘exploitation of the model’. Furthermore, this process of changing and developing relationships affects the individual participants in it. Strategy choice becomes the playground for claiming identity and for the formation of social identity. A final point is that people can cross the borders of the ranges that politeness theory predicts for their behaviour. Is politeness theory flexible enough to deal with ‘rule-violations’, ‘non-occurrence of expected forms’, ‘exceptions’ or ‘breaches’ (Brown & Levinson, 1978: 287)? Brown and Levinson’s answer to this question is a conditioned ‘yes’. The condition is that one can make sense of the non-occurrence of expected forms by analysing the rationale behind them. The question can now be posed differently:

Adolescents Constructing Equality

107

which communicative behaviours are really contradictory to politeness theory? In this respect, politeness theory is not clear at all, which has made critics of the theory conclude that it cannot be refuted empirically, and thus cannot be a ‘real’ theory (Brown & Levinson, 1987). In this study, however, we take a different course. We make use of the politeness framework, with a special interest in so-called ‘borderline behaviour’, i.e. less-, or even non-determined behaviour. We want to explore the broader ranges from which adolescents choose language strategies and we want to explore the phenomenon of strategy choice as an arena for changing relationships and for presenting identities. Our study takes place in an urban multicultural setting, and, as other studies with similar subjects show (e.g. Rampton, 1995; Sebba, 1993; van Alphen, 1999), this setting is particularly interesting because both traditional gauges for social identity such as gender, race and class, as well as the patterns of language use based on these are in a state of flux in such communities. Because of this vantage point, this study ties in with the overarching theme of this volume. Social structures and social differences are no longer considered solely as determinants of linguistic behaviour. Instead, we assume that, although linguistic choices can reflect the social context, they can also help to change relationships and construct identities. With respect to gender, Johnson (1998: 21) denotes this vantage point as: ‘Gender is a verb’. Until a couple of decades ago categories such as race, social class, gender, and ethnicity were treated as static entities, originating in nature (or, indeed, in society as such), i.e. they were dealt with as ‘nouns’ (Street, 1993). However, in recent years, sociolinguistics (as well as other sciences) has come to regard these as variables which are the results of social construction in which human beings assume the roles of agents, i.e. they are ‘verbal’. As said before, the theoretical framework of this study is politeness theory. More specifically, the study focuses on the predictive adequacy of this theory. Before we go into the method and results, we will first present the results of a previous study, which uncovered that different patterns exist in superficially similar European countries. Most patterns confirm the hypothesis that is central to politeness theory, but one did not. Two assumptions in politeness theory are: (1) People differentiate social contexts on a dimension of ‘weight’ or ‘delicateness’, resulting in relatively careful or relatively uninhibited behaviour. (2) People order language forms such as requests along a dimension of politeness: some are seen as intrinsically more indirect, while others are interpreted as intrinsically more straightforward. The basic hypothesis, then, is: people choose relatively highly ranked request forms towards relatively highly ranked addressees and in relatively highly ranked situations, and they choose relatively lowly ranked request forms with relatively lowly ranked addressees, and in lowly ranked situations. A strict and deterministic interpretation of this hypothesis is presented in Figure 2. At the top of Figure 2 addressee/situation categories (the dots) have been ordered from most careful (left) to most uninhibited (right). At the left side of the figure, language forms (the scribbles) are ranked from most careful (top) to most straightforward (down). The diagonal in the figure represents the hypothesis:

108

Bilingualism and Social Relations

Figure 2 A deterministic interpretation of the hypothesis in politeness theory

the highest ranked combination of addressee and situation receives the highest ranked request form, a lower ranked combination receives a lower ranked request form, etc. This pattern can function as a point of reference for the interpretation of empirical data: if the empirical facts yield this diagonal pattern, the hypothesis is confirmed. Figure 3 portrays a second, and a less deterministic interpretation of the hypothesis. In this case, highly ranked addressee/situation combinations can receive a number of alternative, though relatively highly ranked, request forms. The same holds for relatively lowly ranked addressee/situation combinations: they can be approached with a number of alternatives, but all of them will be ranked relatively low. If the empirical facts yield the pattern of Figure 3, politeness theory is confirmed. Nevertheless, the language users who produce this pattern have more freedom of choice than those for who Figure 2 holds. A pattern

Figure 3 A second version of the hypothesis

Adolescents Constructing Equality

109

like the one in Figure 3 was found to describe the behaviour of adolescents in the United Kingdom (Huls, 1991). Figure 4 also portrays a possible interpretation of the politeness hypothesis. In this case one can see that, although the language offers a lot of alternatives for performing requests, only a very limited number of these are in use. Because the highest ranked request form is used towards the higher ranked addressee/situation combinations, and the second highest ranked request form is used towards lower ranked addressee/situation combinations, and so on, politeness theory is again confirmed, although the pattern is widely different from that in Figures 2 and 3. West-German adolescents produced a pattern resembling the one in Figure 4 (Huls, 1991). Figure 5, however, if found in actual data, refutes politeness theory. Here, addressee/situation combinations are hierarchically ordered, just as request forms are, but, as the rectangle in the lower part of Figure 5 indicates, highly ranked addressee/situation combinations are approached in the same way as lowly ranked addressee/situation combinations. That is, they receive low-ranked request forms. This means that language forms are used to level out

Figure 4 A third verison of the hypothesis

Figure 5 A refutation of the hypothesis

110

Bilingualism and Social Relations

a social hierarchy, an outcome that is at odds with politeness theory. The pattern of Figure 5 has indeed been found among Danish adolescents (Huls, 1991).

Design of the Study As stated before, we collected data by means of a questionnaire, consisting of three parts (compare Hill et al., 1986; Huls, 1991). Part I contained a list of items one might use when one wants to borrow a pen: • I wonder if I could borrow your pen for a minute? • Do you mind if I borrow your pen for a minute? • Would you mind if I borrowed your pen for a minute? • May I borrow your pen for a minute? • Would you lend me your pen for a minute? • Do you have a pen I can use for a minute? • Can you lend me your pen for a minute? • Could you lend me your pen for a minute? • Give me your pen for a minute. • I need a pen. • A pen. Some of the items in this list are direct without ‘embellishment’, others are direct and embellished with respectful means, and yet others are indirect. No direct forms embellished with solidarity means were presented: this type of embellishment overwhelmingly involves major changes in the grammatical structure of the item and this would impede further analysis. The items were presented in Dutch (Rotterdam) and Danish (Køge). If applicable, the list contained both a T-form (derived from Latin tu) and a V-form (derived from Latin vos) of the item, resulting in a total of 20 items presented. The respondents had to cross out the items that they would not use, and rate the other ones with respect to carefulness on a five-point scale. We presented the items in a random order that was identical for all respondents. Part II consisted of the following list of people and situations: • Your tutor in his or her office. • A younger sibling with whom you are talking at home. • The secretary who makes an appointment for you with your school principal. • A younger teacher sitting with you in the cafeteria. • A fellow student with whom you are doing your homework at home. • Your school principal at school. • An older sibling with whom you are talking at home. • A classmate with whom you are waiting for class to begin. • Your best friend in your own room. • Your mother with whom you are talking at home. We asked the respondents to cross out the items that were totally foreign to their experience, and rate the other ones with respect to carefulness on, again, a five-point scale. The order of presentation was random, and identical for all respondents.

Adolescents Constructing Equality

111

In Part III, we asked the respondents to keep in mind a certain addressee in a specific context, and choose one or more items of the list for requesting a pen from these specific addressees in these specific contexts. The respondents lived in urban multicultural settings. In Rotterdam in the Netherlands, 48 Turkish adolescents currently attending Grades 3 or 4 of a form of lower secondary education (mavo/vbo) took part in the study. Their mean age was 15.0, with a minimum age of 14 and a maximum of 17. As for gender, 30 respondents were female, and 18 were male. All respondents were competent in Dutch. The majority of them was born in the Netherlands, while their parents were born in Turkey. The other group of adolescent respondents in Rotterdam had no migration background: they were all born in the Netherlands, as were their parents. Of the total number of 50, 30 were female and 20 male. These adolescents attended the same form and level of education as the Turkish respondents. Their age ranged from 14 to 16, with a mean of 14.9 years. In Køge in Denmark, the Turkish group consisted of 22 adolescents attending Grades 8 or 9 of extended primary education. Their mean age was 14.4, with a minimum age of 13 and a maximum age of 15; 15 respondents were female, and 7 were male. All respondents were competent in Danish. The majority of them was born in Denmark, while their parents were born in Turkey. As in Rotterdam, we also investigated a group of adolescents in Køge that had no background of migration. They were all born in Denmark, just as their parents were. This group consisted of 37 adolescents, of which 21 were female and 16 male. These adolescents attended the same form and level of education as the Turkish group. Their age varied from 13 to 16, with a mean of 14.4 years. The questionnaires were filled in during school hours. Depending on the wishes of the school principals involved, this happened in class or in small groups outside the classroom. There was always one of us present to provide explanations and to answer questions. Someone fluent in Turkish was also present in order to explain things in Turkish if necessary. We told the respondents that they were participating in a study into the question of how people use language in different situations and gave a short oral instruction in advance. Then we asked the adolescents to read the written instructions that accompanied the questions carefully. Although for some respondents it took some effort to understand the instructions, in the end, all appeared to have understood them, and the completion of the questionnaire was smooth.

Results Tables 1 and 2 contain the results of the respondents in Rotterdam (The Netherlands). Table 1 concerns the adolescents with no migration background; Table 2 concerns the Turkish adolescents. Similarly, Tables 3 and 4 contain the results of the respondents in Køge (Denmark). On the left in each table, the request forms have been ordered from most careful (‘Ik vraag me af of … etc.’ or ‘I wonder if …’) to most uninhibited (‘Een pen’ or ‘A pen’). Each request form is followed by what we will refer to as its carefulness score.

112

Bilingualism and Social Relations

The top part of the tables contains the addressees and their situational contexts. They are also ranked from most careful (the school principal in his or her office = SP) to most uninhibited (a classmate with whom you are waiting for class to begin = C). Here too, every item is followed by its carefulness score. In the central part of the tables, the frequencies of specific request forms directed at specific addressees in specific situations are represented. The following abbreviations are used: SP = your school principal at school; T = your tutor in his or her office; S = the secretary who makes an appointment for you with your school principal; M = your mother with whom you are talking at home; YT = a younger teacher sitting with you in the cafeteria; OS = an older sibling with whom you are talking at home; FS = a fellow student with whom you are doing your homework at home; F = your best friend in your own room; YS = a younger sibling with whom you are talking at home; C = a classmate with whom you are waiting for class to begin. The respondents with no migration background in Rotterdam The mean ranking of request forms varies from 4.5 to 1.3. The standard deviations vary from 0.47 to 1.43. The mean standard deviation that is found for 20 request forms is 0.92. This is relatively low (if we take the mean standard deviations that were found in the other groups, as well as in a comparable study of adolescents in different countries of the European Community (Huls, 1991), as the point of reference). This means that the consensus among the adolescents in the assessment of the request forms is rather high (see Table 1). The highly ranked forms are nearly all forms with two or more hedges, while the use of the V-pronoun also appears to contribute to a high ranking. Request forms ranked low are relatively short and characterised by use of the T-pronoun. These rankings correspond to common intuitions about politeness. Nine request forms obtain a mean ranking between 4 and 5. This means that one judges a relatively high number of request forms as ‘not standard’ or ‘not neutral’, but as very careful. This might be a remarkable phenomenon, signalling a different attitude of this new group of respondents towards respect and politeness. The mean ranking of addressees in specific circumstances varies from 4.6 to 1.7. The mean standard deviation of this ranking is 1.05, which is intermediate, neither indicating a high, nor a low degree of agreement. The classmate, who is ranked lowest, obtains a relatively high position with 1.7. The different addressees occupy dispersed positions on the scale. The hierarchy appears to reflect a combined effect of power and social distance. Remarkable in this hierarchy (at least in comparison to previous research with Dutch adolescents (Huls, 1991)) is the position of the mother: she is judged relatively high by the respondents from Rotterdam (3.4 vs. 1.8). This might be attributed to an increase in her status, but it cannot be excluded that ‘affection’ plays a role too: affection can influence the carefulness of behaviour in otherwise egalitarian and close relationships. Further evidence for some role of ‘affection’ in the construction of the social hierarchy by this group of respondents comes from the results for the classmate and for the friend. The classmate has the lowest position in the hierarchy, which cannot easily be interpreted as solely reflecting power and/or social distance. Perhaps a certain lack of interest, which might be trans-

Adolescents Constructing Equality

113

Table 1 Adolescents with no migration background in Rotterdam SP 4.6 1

Ik vraag me af of ik uw pen even zou kunnen lenen. (4.5)* Zou u mij uw pen even kunnen 5 lenen? (4.5) Zou u er bezwaar tegen hebben 2 als ik uw pen even leende? (4.5) Zou u mij uw pen even willen 3 lenen? (4.4) Hebt u er bezwaar tegen als ik 2 uw pen even leen? (4.3) Ik vraag me af of ik je pen even zou kunnen lenen. (4.2) Heb je er bezwaar tegen als ik je pen even leen? (4.1) Zou je er bezwaar tegen hebben als ik je pen even leende? (4.1) Kunt u me uw pen even lenen? 7 (4.1) Mag ik uw pen even lenen? (4.0) 19 Hebt u een pen die ik even kan 4 gebruiken? (4.0) Heb je een pen die ik even kan gebruiken? (3.6) Zou je me je pen even willen lenen? (3.2) Zou je me je pen even kunnen 1 lenen? (3.1) Geef me uw pen even. (3.0) Kun je me je pen even lenen? (2.8) Mag ik je pen even lenen? (2.6) 3 Ik heb een pen nodig. (2.0) Geef me je pen even. (1.8) Een pen! (1.3)

T 4.5

S* 4.0 1

5

M 3.4

YT OS FS 3.2 2.4 2.1

1

2

2

1

1

2

1

2

F 1.9

YS 1.9

C 1.7

1

1 1 2 1

1 1

1 1 1

3

1

23 2 3

1

1 3

2 1

2 7

8

1

2

1

1

2

2

2 13 4 12 2

7

2

2

1

2

12 10 4

2

1

2

1

4 16 2 16 2

2 22 2 18 2

5 14 3 13 3

5 24 3 13 1

* Item that was crossed out by more than 75% of the adolescents.

lated as lack of affection, can be due to this low ranking. On the other hand, the best friend – a close and equal interlocutor – is ranked relatively high. Just as in the high ranking of the mother, affection may be playing a role here. The deterministic interpretation of politeness theory predicts a clustering of the observations along the diagonal (see Figure 2). This is not found. The bulk of the frequencies is in the lower half in Table 1. This means that the social hierarchy is not reflected in the choice of linguistic forms as sharply as is theoretically possible: a kind of linguistic levelling, comparable to the levelling by the Danish adolescents in 1991 (see Figure 5), has taken place. However, the pattern is not equal to the previously found Danish one. Within the lower part of Table 1, one

114

Bilingualism and Social Relations

can see the social hierarchy doing its work: although highly ranked addressees (4.5 and 4.6) are pulled down with a request form ranked relatively low (4.0), these addressees still receive higher ranked request forms than lower ranked addressees such as the fellow student, the friend, the younger sibling and the classmate. Looking at the frequency of request forms towards specific addressees in specific circumstances (the central part of Table 1), one sees that three forms, with very different positions on the scale, are favoured: ‘Mag ik uw pen even lenen?’ (‘May I borrow your-V pen for a minute?’), ‘Mag ik je pen even lenen?’ (‘May I borrow your-T pen for a minute?’) and ‘Geef me je pen even’ (‘Give me your-T pen for a minute’). Some other forms are also used with some frequency with highly ranked addressees: ‘Zou u mij uw pen even kunnen lenen?’ (‘Could you-V lend me your-V pen for a minute?’) and ‘Kunt u me uw pen even lenen?’ (‘Can you-V lend me your-V pen for a minute?’). Similarly, ‘Kun je me je pen even lenen?’ (‘Can you-T lend me your-T pen for a minute?’) is chosen once in a while with persons ranked low, but many forms are not chosen with a frequency that is worth mentioning. This pattern is similar to Figure 4. This means that the range of possibilities that we offered for performing requests is not exploited fully by the respondents. To summarise, the respondents show a choice pattern that is a mixture of Figures 4 and 5: they choose a limited number of forms with dispersed, but generally low, positions on the scale. The respondents are fairly united in their answers. With their choices they level a hierarchy without completely eliminating it. They know about the politeness variable asked about in the questionnaire, but they do not use it very much as a relevant category for their language behaviour. Instead we could describe their behaviour as the construction of equality. The Turkish respondents in Rotterdam In the ranking of request forms there is a variation from 4.2 (‘Zou u er bezwaar tegen hebben als ik uw pen even leende?’ or ‘Would you-V mind if I borrowed your-V pen for a minute?’) to 1.8 (‘Geef me je pen even’ or ‘Give me your-T pen for a minute’). This means that these respondents make less use of the extremes of the scale then their placemates with no migration background. The standard deviations in the assessment of request forms vary from 1.04 to 1.76. The mean standard deviation that was found for 20 request forms is 1.24, which is relatively high. The mean standard deviations found in a previous study (Huls, 1991) were all lower; the highest standard deviations were 1.1, and were obtained for the Italian and the West-German adolescents. This high mean standard deviation signals a relatively low consensus among the Turkish respondents with respect to the ranking of request forms. In other words, the Turkish respondents are not uniform in ascribing degrees of carefulness to the various forms. V-forms were generally rated higher than the corresponding T-forms. And although so-called embedded forms were generally rated higher than imperatives, a counterintuitive ranking is given to ‘May I borrow your-V pen for a minute?’ This item ends up in second position, which is relatively high. The expression ‘A pen’ occupies a relatively high and ‘I wonder if I could borrow

Adolescents Constructing Equality

115

your-V pen for a moment’ a relatively low position, but the fact that these items were crossed out by more than 75% of the adolescents, limits the significance of this finding. A relatively great number of request forms (12) obtained a rather ‘neutral’ ranking between 3 and 4. This finding, together with the great standard deviations and some counterintuitive ranking, might indicate that the Turkish respondents do not attach very precise social meaning to several request forms. The limited use of these forms in their social environment (see the other Rotterdam group) might add to an indifferent attitude towards them. The Turkish respondents start and end their social hierarchy slightly lower than the respondents with no migration background. They give the school principal the highest position (4.4). The younger sibling is ranked lowest. The standard deviations were intermediate (mean 1.04). The different addressee/ situation combinations are found dispersed on the scale. The social ranking is similar in the two Rotterdam groups. For both groups, the ranking reflects power and social distance, but it might also reflect affection. That affection plays a role was also confirmed by some respondents when they gave spontaneous comments on the task. A clustering of numbers can be observed in the left upper part and the right lower part of the table. This means that ‘high’ addressees are approached with ‘careful’ request forms and equals with more uninhibited forms. This pattern was also found in Belgium and the Netherlands in the 1991 study, and confirms the predictions of politeness theory. In informal contacts (OS and the other addressees to the right of this item) two request forms are preferred (‘Give me your pen for a minute’ and ‘May I borrow your pen for a minute?’). In formal contacts (school principal and tutor) two forms are preferred, but a lot of other forms are also chosen. The mother is the addressee who allows most freedom of choice: various request forms, with dispersed positions on the scale, are chosen with her. The younger teacher occupies a position in the social middle, as she or he is approached with relatively careful request forms. The right upper rectangle in Table 2 only contains some incidental observations, which leads to the conclusion that social desirability/over-politeness was not very operative in the response behaviour of this group (nor is it in the non-migration group). The left lower rectangle features some higher frequencies, however, especially in the column for the mother. As stated before, the highly ranked mother was approached with means from the whole range, so including informal forms. Although the presence of significant frequencies in the left lower part of the table in general indicate a levelling process, in this case they can be related to the specific relationship between the mother and her adolescent child, in which power, social distance and affection play a complex role. To summarise, the choice pattern of the Turkish respondents in Rotterdam is hierarchical. It resembles Figure 3. Relatively many request forms are in use, and these can be applied in a number of situations. Every addressee can be approached with various forms. Politeness theory is thus confirmed, but only in the interpretation which offers individuals a certain freedom of choice.

Bilingualism and Social Relations

116

Table 2 Turkish adolescents in Rotterdam SP T M S* YT OS FS F C YS 4.4 3.9 3.7 3.4 3.2 2.4 2.0 1.9 1.7 1.5 7 2 2 1

Zou u er bezwaar tegen hebben als ik uw pen even leende? (4.2) Mag ik uw pen even lenen? (3.9) 12 Zou u mij uw pen even kunnen 6 lenen? (3.7) Zou u mij uw pen even willen 5 lenen? (3.7) Kunt u mij uw pen even lenen? (3.7) 5 Hebt u er bezwaar tegen als ik uw 3 pen even leende? (3.7) Zou je er bezwaar tegen hebben als ik je pen even leende? (3.5) Ik vraag me af of ik je pen even zou 1 kunnen lenen? (3.5) Heb je er bezwaar tegen als ik je pen 2 even leen? (3.5) Hebt u een pen die ik even kan 6 gebruiken? (3.2) Heb je een pen die ik even kan gebruiken? (3.2) Zou je me je pen even kunnen lenen? 1 (3.0) Zou je me je pen even willen lenen? 1 (3.0) Kun je me je pen even lenen? (2.9) 1 Een pen! (2.7)* Mag ik je pen even lenen? (2.6) Geef me uw pen even. (2.6) 1 Ik vraag me af of ik uw pen even zou kunnen lenen. (2.6) Ik heb een pen nodig. (2.4) Geef me je pen even. (1.8)

15 6

13 1

3 1

8 3

1

1

8

5

1

2

1

12 1

3

3

1

1

1 1

1 1

3

3

1

4

1

1

2

2

2

1

2

2

9

1 1

1 1

3

3

4

2

1

4

3

3

1

1

3

5

4 1 11

5 1 19

4 1 14

6 3 19

1 5 13

1 8

5 9

6 17

3 17

6 24

1

1

3 5

3 1

* Item that was crossed out by more than 75% of the adolescents.

The respondents with no migration background in Køge In ranking the request forms, the Køge adolescents without migration background did not use the extremes of the scale (the highest score is 4.0 and the lowest 1.6). All items but four obtain a ‘neutral’ ranking between 2.9 and 4.0. This might indicate that this group feels rather indifferent about the social meaning of request forms. The standard deviations in the rating of the request forms varied from 0.65 to 1.38. The mean standard deviation was 1.01, which is higher than the comparable group in Rotterdam, but lower than the Turkish group there. If they are indifferent, they are not erratic. The Danish adolescents that were previously studied (Huls, 1991) did not show this indifference.

Adolescents Constructing Equality

117

Some counterintuitive rankings add to this picture of indifference. ‘Ville du have noget …’ (an embedded T-form) takes the first position. ‘Kunne du …’ and ‘Kan De …’ occupy the same place. In ordering the addressee/situation combinations, only two parts of the whole scale are used by these Køge adolescents. There is a cluster of five addressees towards the top, around 3.7 (the school principal, the secretary, the tutor, the younger teacher and the mother). A second cluster is found somewhat lower, around 2.5 (consisting of the fellow student, the classmate, the friend, the older sibling and the younger sibling). The standard deviations of these rankings vary between 0.71 and 1.42, while the mean is 1.10. The social order constructed by the Køge adolescents with no migration background, is very different from the orders constructed by the adolescents in Rotterdam, and also diverges considerably from the order that was constructed by the comparable group of Danish adolescents in 1991. The key word to denote the difference is: levelling. The social ladder no longer exists for this group. Instead, they carve up their social environment, as far as it is represented in the questionnaire items, into two groups, which have a relatively small hierarchical difference between them. Most remarkable in the central part of Table 3 is the high frequency with which one request form with an intermediate ranking, ‘Må jeg lige låne din kuglepen?’ or ‘May I borrow your-T pen for a minute?’ is chosen (123 times). If this had actually been the only form chosen, we could have claimed that a process of linguistic as well as social levelling in optima forma had been observed. However, this claim cannot be made in such strong fashion, because four other forms are also chosen, though far less frequent (‘Gør det noget, hvis jeg lige låner din kuglepen?’ or ‘Do you-T mind if I borrow your-T pen for a minute?’ with a frequency of choice of 20; ‘Har De en kuglepen, jeg kan bruge et øjeblik?’ or ‘Do you-V have a pen I can use for a minute?’ also with a frequency of 20; ‘En kuglepen!’ or ‘A pen!’ with a frequency of 22; and ‘Stik mig lige din kuglepen’ with a frequency of 29). Their choice implies a diagonal pattern which starts relatively low (at 3.7). One can argue that the pattern of Figure 4, although changed in certain respects, is also operative to some degree in the data. The statement ‘everybody treats everybody equal’ or ‘we treat everyone as we treat our mother’ describes this group of adolescents well, more so than any of the other groups that took part in the study. This group falsifies the central hypothesis of politeness theory because they do not manipulate a variety of linguistic forms in the light of social relationships. Instead, they claim equality. The Turkish respondents in Køge The Turkish respondents in Køge use the extremes of the scale more than the other groups: the highest score is 4.5 (‘Ville De have noget imod, at jeg lige lånte Deres kuglepen?’, or ‘Would you-V mind if I borrowed your-V pen for a minute?’) and the lowest 1.0 (‘Stik mig lige din kuglepen’, or ‘Give me your-T pen for a minute’). The standard deviations in the rating of the request forms varies from 0.15 to 1.53. The mean standard deviation was high (1.15), although not as high as among the Turkish respondents in Rotterdam. With one exception (‘Må jeg lige låne din/Deres kuglepen’ or ‘May I borrow your-T/V pen for a minute?’), the V-forms are rated higher than the T-forms. So-called embedded forms are rated higher than the imperative and elliptical

Bilingualism and Social Relations

118

Table 3 Adolescents with no migration background in Køge SP S T YT M FS C F OS YS 4.0 3.9 3.7 3.7 3.4 2.6 2.6 2.5 2.5 2.3 3 1 1 1

Ville du have noget imod, at jeg lige lånte din kuglepen? (4.0) Jeg spekulerer på, om jeg lige kunne 1 låne Deres kuglepen. (4.0)* Ville De Have noget imod, at jeg lånte Deres kuglepen? (4.0) Vil De lige låne mig Deres kuglepen? 2 (3.9) Gør det noget, hvis jeg lige låner Deres 1 kuglepen? (3.8) Gør det noget, hvis jeg lige låner din 5 kuglepen? (3.7) Har De en kuglepen, jeg kan bruge et 2 ¢jeblik? (3.7) Har du en Kuglepen, jeg kan bruge et 3 øjeblik? (3.5) Må jeg lige låne Deres kuglepen? (3.5) 2 Kunne De lige låne mig Deres 2 kuglepen? (3.2) Jeg spekulerer på, om jeg lige kunne låne din kuglepen. (3.1)* Må jeg lige låne din kuglepen? (3.0) 10 Kunne du lige låne mig din kuglepen? 2 (3.0) Kan De lige låne mig Deres kuglepen? (3.0) Kan du lige låne mig din kuglepen? (2.9) Vil du lige låne mig din kuglepen? 4 (2.9) Stik mig lige Deres kuglepen. (2.3) Jeg har brug for en kuglepen. (2.1) En kuglepen! (1.8) Stik mig lige din kuglepen. (1.6)

2

2

1

1

1

1

1

1

1

2

5

2

1

2

6

3

2

1

2

2

2

5

2

1

2

2

2

1

2 2

3

1

1 1

1 3

1

1 1

1

15

14 1

1 8 1

1 1

17

3

6 3

18

14

1

1

1

1 1

14 2

1

2

2

4

3

1

2 2

1

2 1 2 5

2 1 5 4

2

6 5

7

1

1 1 2 7

1 7 6

* Item that was crossed out by more than 75% of the adolescents.

forms. Thus, the ranking reflects canonical notions about politeness to a great extent. A relatively large number of request forms (13) obtained a ranking on the scale between three and four. This finding, together with the high standard deviations and the one counterintuitive ranking, might indicate that these respondents are sometimes not confident when assessing the carefulness of request forms. They are not as indifferent as the other Køge group. In the ordering of addressee/situation combinations, more positions on the scale were used than by the Køge respondents without a migration background.

Adolescents Constructing Equality

119

Table 4 Turkish adolescents in Køge T M SP S FS F YT C OS YS 4.2 4.0 4.0 3.7 3.1 3.1 3.1 2.7 2.6 2.0 Ville De Have noget imod, at jeg lånte Deres kuglepen? (4.5)* Vil De lige låne mig Deres kuglepen? 2 (4.1) Jeg spekulerer på,om jeg lige kunne låne din kuglepen. (4.0)* Har du en Kuglepen, jeg kan bruge et 6 øjeblik? (3.9) Gør det noget, hvis jeg lige låner 1 Deres kuglepen? (3.9) Har De en kuglepen, jeg kan bruge et 1 øjeblik? (3.6) Ville du have noget imod, at jeg lige 1 lånte din kuglepen? (3.6)* Kan De lige låne mig Deres kuglepen? (3.5) Gør det noget, hvis jeg lige låner din 1 kuglepen? (3.5) Må jeg lige låne din kuglepen? (3.4) 10 Jeg spekulerer på, om jeg lige kunne låne Deres kuglepen. (3.3)* Kunne De lige låne mig Deres kuglepen? (3.3) Må jeg lige låne Deres kuglepen? (3.3) Vil du lige låne mig din kuglepen? (3.3) Stik mig lige Deres kuglepen. (3.2) Kunne du lige låne mig din kuglepen? (2.7) Kan du lige låne mig din kuglepen? (2.6) Jeg har brug for en kuglepen. (2.5) 1 Stik mig lige din kuglepen. (1.8) En kuglepen! (1.0)

1

3

1

6

4

1 1

1

2 1 1

2 1

1

1

1

1 1

10

1

1

1

8

5

13

2 1

1

1

1

9

2

11

1

2 1

7

8

2 1

2

1

2

2

1

1

1 6 3

5 2

2 1 1

4 4

1 1

2 3

1 4 5

* Item that was crossed out by more than 75% of the adolescents.

Still, since this group varies between 2.0 and 4.2, they make use of a smaller range than the two groups of adolescents in Rotterdam did. There is a highly ranked cluster of the tutor, mother and the school principal. Then comes the secretary. There is a cluster of three addressees around 3: the fellow student, the best friend and the younger teacher. Two addressees end up slightly above 2.5: the class mate and the older sibling. Low in the hierarchy is the younger sibling. It is evident that these respondents acknowledge a hierarchy, although the parameters which determine it are less clear. Power and social distance appear to play a role in it, but the high positions of the mother and the friend suggest that affection is also a contributing factor.

120

Bilingualism and Social Relations

When we look at the choice of forms for specific addressees (the central part of Table 4), we see that there is one form that is predominant: ‘Må jeg lige låne din kuglepen?’ or ‘May I borrow your-T pen for a minute?’. If the Turkish respondents in Denmark do not choose this form, they choose ‘Har du en Kuglepen, jeg kan bruge et øjeblik?’ or ‘Do you have a pen I can use for a minute?’ when addressing a high-placed person and ‘Stik mig lige din kuglepen’ (‘Give me your pen for a minute’) or ‘En kuglepen’ (‘A pen!’) when addressing a low-placed person. This means that, when these respondents do not behave in the most likely manner, they behave according to politeness theory. More specifically, their answering pattern shows similarities with Figure 4. There is not much freedom of choice. To summarise, the language behaviour of the Turkish adolescents in Køge is basically oriented to the construction of equality. The differences they acknowledge in both the social hierarchy and among the various request forms that are offered in the questionnaire, are apparently not exploited in actual language use. The repertoire that is in use for performing request forms is limited. Addressees are overwhelmingly treated in the same way. Where they are not, the choice of request form seems fixed.

Conclusion and Discussion Three of the four groups of respondents participating in this study show an attenuated politeness pattern. The exception is the group of Turkish adolescents in Rotterdam. Their answering pattern is schematised in Figure 6. The pictures for the other three groups are presented in Figures 7, 8 and 9. The pattern of the Turkish adolescents in Rotterdam is hierarchical and non-determined, and it confirms politeness theory. It is similar to the patterns found by Hill et al. (1986) in the US and Japan, and the patterns foundby Huls (1991) in The Netherlands, France, Belgium, the United Kingdom and Italy. The Køge adolescents with no migration background (represented in Figure 7) are the most extreme in refuting politeness theory: they are claiming and constructing equality. The framework of politeness theory with its presumed social hierarchy and its subtle treatment of the ‘minutiae in social interaction’ (Brown & Levinson, 1978) is no relevant point of reference for the language

Figure 6 The Turkish adolescents in Rotterdam

Adolescents Constructing Equality

121

Figure 7 The adolescents with no migration background in Køge

Figure 8 The Turkish adolescents in Køge

behaviour of this group. The importance of traditional norms for politeness has faded away. The language behaviour of the Turkish adolescents in Køge (Figure 8) is also oriented towards the construction of equality. They differ from their compatriots who have no migration background in three respects: they do construct a social hierarchy, they do try to differentiate language forms as to their social meanings, and they do not show an exclusive preference for one or a few request forms. On the other hand, they do align with the other Køge group in not using politeness in request forms as a means for building relationships and for distinguishing oneself from the other. Finally, the adolescents in Rotterdam with no migration background (Figure 9) appear to be involved in a process of levelling as well: they do not exploit the possibilities of language for social differentiation. Instead, they soften the effects

Adolescents Constructing Equality

123

tructed. The limited repertoire that we found in request behaviour also reflects attenuation of the traditional repertoire of linguistic forms used for requesting things. However, that doesn’t mean this study necessarily reflects the current reality of request behaviour and/or identity formation in its totality. The ways in which adolescents actually negotiate identity in interaction with one another (Rampton, 1995), falls outside our scope. Although we have reported standard deviations and have commented on the degree of consensus among the respondents within the four groups, it is difficult to address individual differences between respondents based on this questionnaire. To answer such questions, other methods of data collection, such as observation of spontaneous speech and more elaborate elicitation techniques, are required (Rampton, 1995; Sebba, 1993). Finally, it is interesting to note that this study points to the relevance of the variable ‘affection’ for linguistic politeness. While the diminishing importance of a social status hierarchy might be a recent phenomenon in circles of adolescents in urban multicultural settings, the finding that ‘affection’ plays a role in strategy choice confirms what Brown and Gilman (1989) already found for Early Modern English (1500–1700). In their empirical test of politeness theory in the four major tragedies of Shakespeare, they found power and imposition to be good predictors of politeness behaviour, but the results for social distance were not so good. They explain this as follows: ‘the two components of D (social distance), interactive closeness and affect, are not closely associated in the plays’ (Brown & Gilman, 1989: 159). Other studies of social distance also suggest that it has been conceptualised in too simple a manner in politeness theory, in particular because it ignores the role of affect (Baxter, 1984; Slugoski, 1985; Slugoski & Turnbull, 1988). In their later work, Brown and Levinson (1987) admit this. A model which puts a great deal of weight on affection as a factor in accounting for differences in politeness behaviour is Wolfson’s (1988) ‘bulge’ model of interaction. It claims that we behave similarly to those at the two extremes of social distance, i.e. to intimates and strangers. We do not bother to use a great deal of linguistic politeness towards these types of interlocutors. There is a bulge in the amount of politeness paid to people in between these extremes – acquaintances and friends. However, although this ‘bulge’ model certainly explains some aspects of the hierarchies of carefulness made by the adolescents in this study, it is not reflected in their amount of linguistic politeness. The main value of this study is that it has revealed that there are limits to the degree to which social hierarchy determines language behaviour, and even casts doubt on the universality of the construct ‘social hierarchy’ itself. This finding legitimates the growing interest in sociolinguistics for the linguistic behaviour of individuals: in addition to aggregating them into groups on the basis of the traditional categories used in the field (social class, gender, ethnicity, etc.), we should also pay more attention to how they exploit, and sometimes alter, what we often conceive of as deterministic models. What shows up as idiosyncratic language use, may actually be a sign of ongoing change. Note 1. When told by someone who is not British, the anecdote becomes an ethnic joke: it plays on the stereotype that exists of British culture. However, when told by a well-established British linguist, his self-mockery adds to the fun.

124

Bilingualism and Social Relations

Correspondence Any correspondence should be directed to Dr Erica Huls, Discourse Studies, University of Tilburg, PO 90153, 5000 Le Tilburg, The Netherlands ([email protected]). References Alphen, I. van (1999) Variatie in verbale interactie (v/m); een sociolinguïstisch onderzoek naar de vorm en de functie van het taalgebruik van adolescente meisjes en jongens. (Variation in verbal interaction (f/m); a sociolinguistic study of form and function of language use of adolescent girls and boys). Amsterdam: Academische Pers. Baxter, L.A. (1984) An investigation of compliance gaining as politeness. Human Communication Research 10, 427–456. Brown, P. (2001) Politeness in language. In N.J. Smelser and P.B. Baltes (eds) International Encyclopedia of the Social and Behavioral Sciences. Amsterdam: Elsevier. Brown, P. and Levinson, S. (1978) Universals in language usage: Politeness phenomena. In E.N. Goody (ed.) Questions and Politeness: Strategies in Social Interaction (pp. 56–289). Cambridge: Cambridge University Press. Brown, P. and Levinson, S. (1987) Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Brown, R. and Gilman, A. (1989) Politeness theory and Shakespeare’s four major tragedies. Language in Society 18 (2), 159–212. Bruggen, C. van (1974) Prometheus: een bijdrage tot het begrip der ontwikkeling van het individualisme in de literatuur. (Prometheus: A contribution to the understanding of the development of individualism in literature). Amsterdam: Van Oorschot. Goodwin, C. (1987) Forgetfulness as an interactive resource. Social Psychology Quarterly 50, 115–130. Goffman, E. (1955) On face-work: an analysis of ritual elements in social interaction. Psychiatry 8, 213–231. Goffman, E. (1967) Interaction Ritual: Essays on Face to Face Behavior. New York: Anchor Books. Hill, B., Ide, S., Ikuta, S., Kawasaki, A. and Ogino, T. (1986) Universals of linguistic politeness; quantitative evidence from Japanese and American English. Journal of Pragmatics 10, 347–371. Holmes, J. (1992) An Introduction to Sociolinguistics. London and New York: Longman. Huls, E. (1991) Een landenvergelijkend onderzoek naar de formulering van een verzoek. (A comparative study of request formulations in various countries). In R. van Hout and E. Huls (eds) Artikelen van de Eerste Sociolinguïstische Conferentie (pp. 241–258). Delft: Eburon. Huls, E. (2001) Dilemma’s in menselijke interactie; een inleiding in de strategische mogelijkheden van taalgebruik. (Dilemmas in human interaction; an introduction to the strategic options in language use). Utrecht: Lemma. Johnson, S. (1998) Theorizing language and masculinity: A feminist perspective. In S. Johnson and U.H. Meinhof (eds) Language and Masculinity (pp. 8–26). Oxford: Blackwell. Labov, W. (1972) Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Lakoff, R. (1975) Language and Women’s Place. New York: Harper Colophon. Maynard, D.W. and Zimmerman, D.H. (1984) Topical talk, ritual and the social organization of relationships. Social Psychology Quarterly 47, 301–316. Nofsinger, R.E. (1991) Everyday Conversation. Newbury Park: Sage. Rampton, B. (1995) Crossing: Language and Ethnicity Among Adolescents. London and New York: Longman. Scollon, R. and Wong Scollon, S. (1995) Intercultural Communication; a Discourse Approach. Cambridge MA/Oxford: Blackwell. Searle, J.R. (1998) Indirect speech acts. In A. Kasher (ed.) Pragmatics: Critical Concepts Vol. 5, Part 8: Indirect Speech Acts (pp. 639–657). London: Routledge. Sebba, M. (1993) London Jamaican. London: Longman.

Adolescents Constructing Equality

125

Slugoski, B.R. (1985) Grice’s theory of conversation as a social psychological model. PhD thesis, Oxford University. Slugoski, B.R. and Turnbull, W. (1988) Cruel to be kind and kind to be cruel: Sarcasm, banter, and social relations. Journal of Language and Social Psychology 7 (2), 101–121. Street, B.V. (1993) Culture is a verb: Anthropological aspects of language and cultural processes. Language and Culture. British studies in Applied Linguistics 7 (pp. 23–43). Clevedon: Multilingual Matters. Thomas, J. (1995) Meaning in Interaction; An Introduction to Pragmatics. London and New York: Longman. Trudgill, P. (1974) The Social Differentiation of English in Norwich. Cambridge: Cambridge University Press. Wolfson, N. (1988) The bulge: A theory of speech behaviour and social distance. In J. Fine (ed.) Second Language Discourse: A Textbook of Current Research (pp. 21–38). Norwood, NJ: Ablex.

Languaging Among Fifth Graders: Code-switching in Conversation 501 of the Køge Project J. Normann Jørgensen Department of Nordic, University of Copenhagen, Denmark Conversation 501 from the Køge Project is analysed for the four adolescent speakers’ use of five to six different langauges or varieties. The code choice patterns are shown to function in a range of different ways, all of which can be described as phenomena that are not inherently different from the language use of youth as it is described in youth languages studies. In the conclusion, the article argues that ‘bi-lingualism’ – understood as something which involves two units of which ‘mono-lingualism’ only involves one – is meaningless. The multifaceted behaviour of the Turkish-Danish adolescents is languaging, which is in principle the same as what other human beings do.

The linguistic behaviour of adolescent speakers has been the focus of an increased number of studies in Northern Europe over the last decade of the 1900s (e.g. Androutsopoulos, 1998; Androutsopoulos & Scholz, 1998; Kotsinas, 1994; Kotsinas et al., 1997; Kotsinas et al., 1999; Stenström et al., 2000; Drange et al., forthcoming, and others). To a certain extent the studies have focused on multilingual behaviour, particularly code-switching and code-mixing. These studies deal with code alternation in (at least) two different perspectives. One perspective concerns the use of elements (words and phrases) from traditional school-languages, particularly English, by majority adolescents in otherwise monolingual mother tongue speech (e.g. Preisler, 1999; Språknytt, 2000: 1, and several contributions to Stenström et al., 2000). According to these studies, the use of English among teenagers has increased in the 1990’s. English words and phrases seem to function as signals of group identity, for instance as hip-hoppers or death metal enthusiasts (Preisler, 1999), i.e. as markers of an urban youth identity. As Møller and Jørgensen (forthcoming) specify, the inclusion of English elements into the mother tongue of the young people indeed has such a function, whereas wholesale substitution of the mother tongue by English is not appreciated by the teenagers. The high status of the English language is, at least in Northwestern Europe, assumed to be a worldwide phenomenon. Terms such as ‘an international lingua franca’ often appear, particularly in public debates about school languages, and in this light it is not difficult to understand why English elements can carry a flavour of international orientation. The use of English expressions thus signals modern urban youth group membership, and at the same time it functions as a provocation against adult self-appointed linguistic purists. The other perspective deals with the simultaneous use of majority language elements and minority language elements. This practice has been found particularly, but certainly not only, among linguistic minority teenagers (Jørgensen, 1998; Sebba, 1993; Turan, 1999; to mention just a few). In these studies, code-switching is generally described in terms of negotiation of social relations: 126

Languaging Among Fifth Graders

127

Within a single interaction – say, a conversation – participants need not maintain an unvarying relationship among each other, the linguistic event, and the context (both wide and narrow) in which it takes place. Indeed, if we take that conversation, as it unfolds, as constantly changing and renewing its own context, then the relationship just described must be constantly changing as well. (Sebba, 1993: 130) A specific aspect of multilingual behaviour among teenagers is the so-called crossing, i.e. the introduction of elements from languages which the speakers do not command, into their conversations. Rampton (1995) has coined the term and described the phenomenon among young speakers of English, Caribbean Creole, Punjabi, and other varieties, in a London youth centre. The crossing performed by these young speakers has implications for their mutual relationships, the way they draw boundaries to the outside world, and they use crossing to comment on their relations to society at large. Similarly, youth language use involving a majority language and elements from minority languages has been described in Utrecht (Nortier, 2000), Mannheim (Kallmeyer & Keim forthcoming), Copenhagen (Quist, 2000), Hamburg (Auer & Dirim, 2000), and Stockholm (Kotsinas, 1985, 1988a, 1988b). The implications of these studies is that the scope of majority languages becomes wider, for instance in the fact that features from the involved minority languages may be adopted by young majority speakers. In addition, these features may perhaps over time work their way into and thereby enrich the majority languages. Several language political issues are involved here. The general tendency among gate-keepers and opinion-makers in our societies (teachers, personnel directors, politicians, newspaper editors, etc.) is very negative towards the specific youth language features (see Kotsinas, 1994; Kristiansen, 1993; Kunøe, 1991). Some of the features of youth language are favourite targets of negative attitudes. This is doubly relevant for the use of English expressions among adolescents. Such use has been the target of modern day purism, at least in some of the Scandinavian countries. Linguists (e.g. Haberland et al., 1991; Haberland, 2000) insist on a distinction between on the one hand English loans in the language use of teenagers, and on the other hand whole societal functions being taken over partly or entirely by English (air traffic control, advertising, etc.), but politically motivated purists do not always make such a distinction. The ensuing debate is frequently reduced to a discussion for or against English as a medium of instruction in the public school systems. In this atmosphere, the minority languages sometimes disappear out of sight. However, when they do attract attention, it is usually as targets of budget reductions or other negative steps. With respect to the school systems, all of the Scandinavian countries have or have had some measure of mother tongue teaching of minority languages, and in recent years the countries have all cut down on its funding. Also in other respects, the minority languages, and especially immigrant minority languages, have been under pressure by the majority languages (see e.g. Hvenekilde, 1994; Jørgensen, 2000). These differences in the values ascribed to the languages by society at large form a hierarchy of language evaluations (see Jørgensen, 2000), and this is no secret whatsoever to the young speakers. But this is not the only language atti-

128

Bilingualism and Social Relations

tude which is widely held in society. Grade school students are also met with a view of bilingualism which is very negative towards simultaneous language use (see Boyd et al., 1994; Jørgensen & Holmen, 1997). There is a lot of pressure on grade school students to learn several second or foreign languages. But at the same time they are expected to keep their languages separate. It is – sometimes – described as a positive quality that minority children acquire their parents’ mother tongue. But it is almost always also emphasised that this must not be at the cost of not learning to speak the majority language exactly like the majority speakers do. This double monolingualism view of bilingualism is very strong among adults, including decision makers, and it is massively enforced on grade school students (Holmen & Jørgensen, 1997; Laursen, 2001). The teenagers in modern, urban Northwestern Europe see and hear several languages every day. Majority adolescents can hardly avoid English in the media, entertainment, computer games, etc. which most of them consume on a daily basis. Minority teenagers, in addition to the two societal languages experienced by the youth, also hear their mother tongue. There are, as we have seen, at least three strong societal forces which are determined to control and regulate the adolescents’ use of these languages: • the language hierarchy; • the double monolingualism view; and • the generally negative evaluations of teenagers’ speech. Youth is the period between what is unequivocally childhood and what is unequivocally adulthood. It is characteristic of childhood (and late childhood in particular) that the child orients itself towards the adult-formulated tasks to be fulfilled in school, at home, or elsewhere, and sometimes identifies with the tasks. In the youth age the person is engaged in sometimes fashion-oriented attempts to establish a youth subculture (Erikson, 1971) which will often appear to be a counterculture to the adult world: the young people would rather behave shamelessly in the eyes of the adults than be forced to do something that appears shameless in the eyes of their peers. Thus the signals of belonging to the youth group can be very conveniently chosen among the behaviours deemed by the adults as not comme il faut. Eckert’s (2000) study in a white suburban high school showed the students to orient themselves in relation to two opposing cultural categories. One of these was a counterculture which was in opposition to the culture approved and encouraged by the adult school authorities. Eckert further found that the students who seemed to most favour membership of the extreme counter-authority group, were also those with the most advanced level of specific sound changes in their speech. Eckert argues that the spread of linguistic features among the students is closely related to the local groupings and the evaluations of these groupings, the linguistic variation is a means to manipulate group membership and group boundaries. The students at the Danish boarding school Herlufsholm similarly showed a tendency to use their local, and unique, argot as a means to distinguish themselves from others (Jørgensen, 1997). The features of the argot are used to show social boundaries and signal group membership. However, at the same time these signals can function as internal instruments of discipline. The Herlufsholm argot is used not only as a criterion of group membership, but also as a means to

Languaging Among Fifth Graders

129

uphold hierarchies within the group. The students do not have equal rights to the argot, and similarly they do not have equal responsibility of maintaining it (Møller & Jørgensen, forthcoming) There is no difference in principle between bilingualism in the traditional sense and the use of different styles and phonetic features as documented in Belten High and Herlufsholm. Any linguistic difference, however small, that can be appreciated by the speakers, can also be used for the speakers’ conversational purposes. In the following I will analyse the use of different languages and varieties by the participants in a grade 5 group conversation among four bilingual Turkish-Danish students. The aim of my analysis is to find out how the varieties are used in relation to the three societal norms for adolescent language use: the language hierarchy, the double monolingualism view, and the generally negative evaluations of teenagers’ speech. The adolescents’ uses of the varieties make sense, although the exact meaning of a language choice may not be readily available to the analyst (see also Hinnenkamp, this volume). This sense we can generally only understand if we also involve the societal norms. But it is also true that to appreciate the meaning of specific conversational items, we must realise that they are precisely that – contributions to conversations. Therefore we must be prepared to understand ad-hoc establishment of themes and evaluations. In the conversation we shall study here we will see both how values are taken from society at large and played into the conversation, and how values are created on the spot without particular reference to specific societal values. The material is Conversation 501 from the Køge Project, a longitudinal study of the bilingual development of linguistic minority students in the Danish grade school (Turan, 1999). This project has provided a wide range of data which enable us to study the fifth grade conversation not only in light of the linguistic development of the individuals as well as the group, but also with reference to the social structures in the group. In the Køge Project, group conversations with the bilingual students have been recorded every year of their nine-year school career. The recordings have been transcribed according to the Childes conventions (MacWhinney, 1995), and the individual utterances have been categorised according to language (see also Jørgensen, 1993; and Esdahl, this volume). With the Childes programmes it is possible to calculate how many utterances in each conversation are formed in a language which is different from the language of the preceding utterance, i.e. the number of intersentential code-switches. Figure 1 shows for each grade level the percentage of utterances which are intersentential code-switches. We can observe a high degree of intersentential code-switching which is unique for Grade 5. There is a steep rise between Grade 4 and Grade 5, and there is a clear fall between grade 5 and grade 6. This indicates that something special happens with the students’ intersentential code-switching practices precisely in Grade 5. In addition to the group conversations with bilingual students, the project also collected group conversations that involved both young majority speakers of Danish and young bilinguals. In these conversations Turkish was rarely used after Grade 5. There is little doubt that code-switching as an efficient linguistic tool is something that the speakers have to acquire and develop. Widely used code-switching is a skill, and it probably depends on a minimum of command of

130

Bilingualism and Social Relations

Figure 1 Utterances which are intersentential code-switches (pct. of all utterances), based on 30 conversations, grade 1–9

the two languages involved. One indication of this is the relatively low frequency of crossing in Rampton’s (1995) study. The instances of crossing he found among his adolescent informants are highly salient – but not very numerous. In the Køge Project the bilingual students learn to – or force themselves to – not speak the language that is not understood by everyone present. This is of course a social and not a linguistic skill, as there is no reason why the bilinguals should not now and then make a comment to each other in their specifically shared langauge. It happens in all the conversations that some remarks are clearly not addressed to everyone present, but only to one, and sometimes there are even two simultaneous conversations. Even in such circumstances, the bilinguals do not speak Turkish to each other after Grade 5. This is not really a surprise, as there is a tremendous pressure on the bilinguals not to use their mother tongue in class, or in front of majority Danes, and this affects their behaviour. Turkish is simply eliminated from situations where monolinguals are present. This indicates that the bilingual students by Grade 5 have acquired so much Danish that they can handle all situations in such group conversations in that language. Their linguistic skills now allow them to abide by the social rules that are set up for them by the outside world. This downgrading of Turkish even spills over into the group conversations with only bilingual participants where they could get by with only Turkish. In these conversations Danish plays a still more important role with growing age (see Figure 2). Figure 2 shows the percentage of Danish-based utterances for each grade level. Danish takes over more and more from Turkish during the eight years. Again there is a peak in Grade 5, and there is a fall in the use of Danish in Grade 6. The graph does not show a continuous increase in the use of Danish. Some factor, uncontrolled by the project design, causes either too much Danish in Grade 5 or

Languaging Among Fifth Graders

131

Figure 2 Danish-based utterances (pct. of all utterances), based on 30 conversations among bilingual students, grade 1–9

too little Danish in Grade 6. The gender composition of the groups is a likely explanation. The boys in boys-only groups begin to use Danish in Grade 4, and they continue to use some Danish, but still mostly Turkish. The girls in girls-only groups use almost no Danish at all until Grade 7. From Grade 7 and on they almost only use Danish. But this is not because the girls do not know Danish well enough to use it (if nothing else, we know this from the face-to-face conversations with adult Danes which were also collected in the project). The girls do in fact use Danish in Grade 5. As it happens, in Grade 5 we only have gender-mixed groups, and here all of the bilingual student use a certain amount of Danish – not very different from the amount of Danish used by the boys in Grade 4 and 6. In other words, when girls are among girls, they speak Turkish – until Grade 7 when they shift into Danish. But when girls are among boys, they do like the boys do. Boys on the other hand, do not seem to change their way of speaking nearly as much as the girls do. This finding fits nicely with the findings of dominance-oriented feminist linguistics: female speakers adjust their linguistic behaviour according to the norms of male speakers. What we do not know, however, is whether boys actually dominate the gender-mixed conversations in terms of initiatives and responses or in terms of nominating the topics, etc. There are indications that this is not at all the case (see Madsen, this volume, Jørgensen, 2001). Another interesting observation can be made around Grade 8: The opposite of Grade 5 seems to be the case here, namely that the boys in mixed-gender conversations adapt a little to the way the girls speak at that age: they speak more Danish. The implication of that observation may be that there is a tendency for the gender that speaks least Danish in other circumstances, to speak more Danish in company of the other gender. There is probably identity work involved in the language use reflected in Figure 2. A sociogram of the adolescent group that these speakers belong to was drawn when they were in the eighth grade. Quist (1998: 113) finds that the most

132

Bilingualism and Social Relations

important boundary within the group is not ethnic, but gender-based. There are many connections between monolinguals and bilinguals, but only one single connection involving a boy and a girl. This may be part of the explanation that the use of Danish, the adolescents’ L2, increases among girls in Grade 5 and among boys in Grade 8, as we observed in Figure 2 above. There are also clear differences in the ways the two gender organise themselves. The boys are united in one large inclusive network, which is hierarchically organised. The bilingual boys are evenly distributed in this hierarchy, as a group they are not particularly central or particularly peripheral. The girls, on the other hand, are organised in friendship pairs or small friendship groups, which are exclusive, but apparently not hierarchical. Except for two bilingual girls, the monolingual girls and the bilingual girls form separate friendships. These particular two girls are both members of a three-person bilingual friendship group and an otherwise monolingual friendship group. They are living bridges between monolingual and bilingual friendship groups, but there are no other girl connections across the ethnic boundary. In other words, the ethnic boundary seems to have some importance among the girls, but not among the boys. Considering the language choice patterns of the girls it is conceivable that what happens between Grade 6 and Grade 7 is related to the relative importance of ethnicity among the girls compared to the boys. The girls simply stop acting Turkish by Grade 6 and start acting Danish by Grade 7. This is in line with the findings of Møller (2001) that age is a highly relevant criterion of identity construction among the bilinguals. In face-to-face interviews they will often place themselves opposite to the teacher and parent generation, in some cases also the grandparent generation. ‘We’ are more often ‘our age group’ than ‘we’ are ‘the bilinguals’. Akkari et al. (1998: 259) also found that ‘youth have multidimensional identities that may be expressed in “youth culture” (music, clothes, sports)’ in their study among youth in a multicultural neighbourhood in Geneva, Switzerland. This gives us another good reason to focus on Grade 5. The identity work involved seems to strengthen the tendency for the speakers to behave differently in the Grade 5 conversations than otherwise. It is likely that the skills of most of the students until grade 5 have not developed to the degree where frequent code-switching is possible, but it is also likely (cf. Erikson’s description of the transition from childhood to adolescence) that there is no social need for it until Grade 5. We may not be able to determine precisely the effect of each of these factors, but we can analyse what happens. There is no doubt that code alternation sets in dramatically by Grade 5. The code profiles of the individual conversations demonstrate that clearly, as one can observe by comparing the profiles of Conversations 401 and 501. In the profiles, ‘5’ marks an entirely Turkish utterance, ‘4’ a Turkish utterance with a Danish loan, 3 ‘other’ utterances, ‘2’ a Danish utterance with a Turkish loan, and ‘1’ an entirely Danish utterance. The flow of utterances in Conversation 401 is quite straight, with an occasional switch from the mainly Turkish-based production. In Conversation 501, however, there is constant oscillation between Danish-based and Turkish-based production. Switches happen with very short intervals all through the conversation. As a result, one cannot unequivocally determine this as either a Danish or a Turkish conversation. Nevertheless it is noteworthy that there are relatively

Languaging Among Fifth Graders

133

Figure 3 Code profile of Conversation 401

many Turkish-based utterances with Danish loans, but only few Danish-based utterance with Turkish loans. This in line with the general distribution of utterances on the two languages (see below). Conversation 501 is worth a closer look. The text has been published in Turan, 1999: 92–107, with English translations. The conversation has also been studied by Jacobsen (2002), Madsen (2002) and Reiff (2002). The participants are two boys: Erol and Ali, and two girls: Esen and Selma. During the conversation they were alone in a room (not a classroom) at the school, wearing buttonhole microphones. They were seated around a table, sitting in comfortable chairs. They had a large (size A1) photocopied map of the world (without names), a stack of travel catalogues, pens, scissors and gluesticks. Their task was to identify the places shown on some of the pictures in the catalogues, cut out the pictures, glue them on the map, and connect them to the right place, be it very local or not so local – KuêadaêÏ would be as fine as Brazil. In the conversation as a whole, roughly 60% of the utterances are in Turkish, about 20% in Danish, and the remaining 20% other (either in a third language, a mix, or unclassifiable). From these figures it appears as if we are indeed dealing with a conversation in Turkish. Nevertheless, as we can also observe in profile 501, the inclusion of non-Turkish elements is so frequent that we would oversimplify if we just considered Conversation 501 a Turkish conversation with an occasional loan or code-switch. The use of non-Turkish elements is not evenly distributed on the speakers. Only Selma meets the average of the group. The boys are the most frequent users of Turkish with roughly 70%, 10% Danish, and 20% other (Ali using slightly more Turkish and less Danish), while Esen has about a third of her utterances in each category. In this conversation, the girls speak less Turkish than the boys.

Bilingualism and Social Relations

134

Figure 4 Code profile of conversation 501

During the conversation a variety of themes appear, get developed, are left and reintroduced. The task that the group is assigned to do is always there as a possible frame of reference. It leads to several sub-themes, such as particular places. Africa appears in six different connections. In one, Esen comments on a picture, in another they all look for Africa on the map, in a third (see Excerpt 4 below) they discuss how to illustrate Africa as part of their task. Turkey is another theme that appears – and it leads further to a different theme, namely holidays spent in and outside Turkey (see Excerpt 1). Similarly Paris leads to a discussion of the Euro-Disney amusement park there, and again vacations. Sweden is a theme of Erol’s who brings it up eight times in the course of the conversation. First he looks for it, then he thinks he has found it, and then again he asks the others whether he has indeed found the right place, but he never gets a straight reply. At one point Esen reacts to his initiative, by playing with the word ‘Sverige’ (see Excerpt 5). At another point Ali discusses whether it is correct to write ‘Sweden’ on the cardboard. In Excerpt 1 we can observe how a theme unfolds. Note that Turkish is in italics, Danish in recte, other languages are underlined, [>] and [<] mark simultaneous speech, +/. interrupted speech, [/] self-interruption, and xxx unintelligible speech; see also MacWhinney, 1995. Excerpt 1 Erol: Eng: Selma: Eng: Esen: Eng: Selma:

Pamukkale’ye gittim ben I have been in Pamukkale Pamukkale mi in Pamukkale? ben çok gittim I have been there a lot ben < gitmedim > [>]

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 > [<] me too gitmedin mi have you not been there? nej biz hamamÏn ¸ok uzaktayÏz ondan gitmedik biz no we are very far from the Tyrkish bath, that is why we did not go biz stanbul’ dayÏz < [>] 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 > [<] Australia biz Ankara’dayÏz we are in Ankara ulan biz nerdeyiz man, where are we? biz de we too [>] dindi I lost my Turkey again <üç yerdeyiz biz zmir’de bak dört yerdeyiz > [<] zmir’de Ankara’da stanbul’da [<] [>] is it Turkey? xxx, look, look, look, look, Esen, Iraq, Iraq, Iraq, Iraq. [<] into my microphone that thing there of ours (blows into the microphone) vi er kun i [>1] og så on gün ferie ya gidiyoruz ya êØyle Mersin’e oluyor ya KuêadasÏ ‘na ya Pamukkale [>2] oluyor we are only in Ankara and then we go on ten days of vacation it will be Mersin or so, or KuêadasÏ, or Pamukkale, one of them [<1] [<2] Erol man, Erol biz Pamukkale’ye gittik ay ne güzeldi we went to Pamukkale, and was that pretty Marmara Deniz’e hepsine gittik ay to the Marmara Sea, we went everywhere, ey bir yere gittik böyle bedava otelin orda otel [>] we were at a place, a free place next to the hotel the hotel lots of rocks, oh man, so pretty [<] Iraq, there, Turkey

136

Bilingualism and Social Relations

In Excerpt 1 we see how the place-name Pamukkale triggers an exchange about vacations spent in Turkey. The most active participants are Erol and Esen, who both relate in detail where they usually spend their summer vacations with their families. Selma also contributes to the subject, although mostly by saying what she does not do. Then she turns her attention to the task (‘Australien’). Ali, however, after having given up telling where he goes on vacation (apparently he does not know where), shifts his attention to the task. He has lost what he calls ‘my Turkey’. In his next remark he tries to attract Esen’s attention, looking for Iraq. With the final utterance in the excerpt he seems to have found Iraq, and as a consequence its geographical neighbour, Turkey, on the map. There are several characteristics to notice in this excerpt. Firstly, it is noteworthy that not one of Ali’s utterances seems to be integrated into the flow of conversation between the others. All of his remarks are simultaneous with something another participant says, and both Erol and Esen blatantly ignore his contributions. It is not because Ali’s utterances are monologous remarks or independent statements – they all relate either to the vacation theme or to the task. But in this excerpt he does not get any attention from the others. Secondly, Ali is not the only one who experiences being ignored. Erol and Selma both react to Esen’s ‘ben çok gittim’, but Erol’s reaction is ignored by the two girls, and not until he persists with the details of his vacation does he get a reply from Esen (but Selma still ignores him and continues with the task). A certain hierarchy appears from this excerpt, with Esen as the strongest individual, the girls being stronger than the boys, and with Ali as the weakest individual. Thirdly, we observe how the subject of Pamukkale develops into the theme of vacation, and further into what is pretty or fun. Fourthly, we observe how the participants shift effortlessly from theme to theme and back again, even Ali. Although he does not receive any great amount of attention, he is still able to follow the conversation. His remarks are relevant enough, they just do not receive any reactions. This raises two issues. One issue has to do with the content of discussion: what do the participants talk about, and what do they say (see also Esdahl’s (this volume) concept of focus). The other issue has to do with the relations between the speakers: who has power (over whom), who makes the decisions, who teases whom, who supports whom? With respect to the content, the place-names are at the core of the conversation. Not only is the task about place names, but they also amount to a point of departure for other themes, and they constitute a frame of reference which the participants can always resort to at any point during the conversation (cf. first Selma’s and then Ali’s return to the task in Excerpt 1). A string of other place-names figure in the conversation and lead to sub-themes. The subject of travels also leads to discussions of airlines. At one point a jingle from an airline TV commercial plays a certain role (see Excerpt 3 below). Advertisements, pictures, and the different place-names trigger several introductions of short themes. There are also several instances of singing, humming, shouting and other ways of having fun which are not conversationally focussed. Most of the activities and themes grow out of the task’s place-names, one way or another. The activity, and the introduction of place-names is not equally frequent with the four speakers. Selma, Erol and Ali each mention about 20 different place-names (types) during the whole conversation. With Selma and Ali we find

Languaging Among Fifth Graders

137

roughly 60 tokens; Selma has Australia 9 times, Turkey 8 times, and Japan 6 times; Ali has Paris 8 times. With Erol we find 80 tokens (Sweden 17 times, and Paris 9 times). Esen stands out with more then 40 different place-names on more than 100 tokens, Africa 8 times, Euro-Disney 8 times, Belgium 7 times, Turkey and France 6 times each. She may or may not know more place-names than the other speakers, but she definitely produces more names, and a much greater variety of places and names than the others do. The social relations among the four speakers are not simple. Madsen (2002) studies the power relations between the four speakers on the basis of four different quantitative criteria. She finds that Esen is by far the most powerful participant. The others are not very different from each other. Madsen posits Selma as the second most powerful speaker, and Erol as the least powerful speaker. Jacobsen (2002) agrees that Esen is indisputably the leader of the four. Jacobsen illustrates this with Esen’s use of Danish as a means of control. Jacobsen finds that Esen uses Danish more often than Turkish to get the others’ attention or to administer the conversation through reprimand and praise. The constant oscillation between languages which is characteristic of Conversation 501 (in comparison to earlier conversations) can be observed in Excerpt 2. Excerpt 2 Erol: Eng: Esen: Eng: Erol: Eng: Esen: Eng: Erol: Eng: Esen: Eng: Ali: Eng: Selma: Eng: Ali: Eng: Selma: Eng: Esen: Eng: Ali: Eng: Selma: Eng: Erol:

manyak fool ih Erol oh Erol (reprimanding) yapma valla küserim don’t, by God, I’ll be angry pis dirtbag terbiyesiz scum sen de you too <Erol ya> [>] Erol now <Erol xxx> [<] Erol xxx Esen valla sen deÈil misin Esen, by God, isn’t that you? åh ja Erol come on, Erol det er du selv you yourself are evde de böyle yapÏyorsan if you do like that at home too <SAS> [>] SAS (an airline) [<]

138

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

Bilingualism and Social Relations

what? hvad what? êimdi Zehra sildim Özkan sildim. now, Zehra, I have wiped it out, Özkan, I have wiped it out biz hiç ya+/. we have writ- (or perhaps don-) nothing åh ja. come on (complaining) åh ja <Erol.> [>] come on, Erol (complaining) <manyak.>[<] fool Esen dedi ki eh xxx’nÏn parasÏnÏbiz ödemiyoruz dedi manyak Dummkopf kart kart cik cik cik cik. Esen said that eh we do not pay xxx’s money she said, fool fool, cut cut no no no no det er løgn. that is a lie. yaz xxx as. write xxx ace. Erol üêüttün. Erol you have gone crazy.

It is not difficult to see that Danish is mainly used for quarrelling. However, Danish is not the only language used for this purpose – Turkish and German both provide explicit terms of negative evaluations of the other. The flow of the quarrelling is constant and fluent. The speakers use a variety of negative expressions, apparently as they come to mind. There is no apparent flagging or other marking of code-switches. Furthermore, the stream of invectives is briefly interrupted by a reference to the task (the remark about SAS). The speakers are obviously able to handle two themes at the same time without dividing the issues nicely between L1 and L2. The issue of the task appears intermittently throughout the conversation, and simultaneously other negotiations go on, in this case an elaborate exchange of negative evaluations – an exchange that fetches its material from both L1, L2, and an L3. The speakers seem to have no difficulty in handling several themes and several languages within the same conversation. In this excerpt there is no obvious significance to the involvement of the separate languages. The important thing is not how each of the languages is used differently from the other, but important is the very fact that more than one language is involved. This is not to say that the individual switches do not carry meaning, but if so, it is a very local meaning (cf. Hinnenkamp, this volume). The material gives us no reason to assume that the variation which characterises code-switching is disappearing. We do not here observe a fused lect in Auer’s (1999) terms, but we may witness an early stage in what could in theory become a development towards a fused lect. In social psychological terms, we may see this language use as a means to strengthen in-group ties, at the cost of (or with the extra bonus of) alienating

Languaging Among Fifth Graders

139

several others at one time: the parent generation, the monolingual Danes, the teachers and school authorities. Such linguistic signalling of youth group membership is no different from the use of advanced vowel pronunciation in monolingual language use (cf. Eckert, 2000) or the use of a group specific argot (cf. Møller & Jørgensen, forthcoming). What is salient in the language use of the speakers in Excerpt 2 is not that they use more than one language. In a sense their language use works as if only one language was at play – there is nothing that distinguishes their use terribly much from the language use of monolinguals who use every available negative term to haul insults at each other. In fact, at least sometimes the attitudes of young bilinguals also point in the same direction, cf. Landsberg, 1997. This is not so everywhere and all the time. Cromdal (2000) (see also Cromdal, this volume) has found a discourse level division of labour between Danish and Turkish in Conversation 801 (eighth grade). In other moments during Conversation 501, there also appears to be a division of function between the languages involved, as can be observed in Excerpt 3. Excerpt 3 Ali: Eng: Esen: Eng: Ali: Eng: Erol: Eng: Selma: Eng: Esen: Eng: Ali: Eng: Erol: Eng: Sel: Eng: Ali: Eng: Esen: Eng: Selma: Eng: Erol: Ali: Eng: Selma: Erol:

hej benim kartÏma bakar mÏÏsÏnÏz hey will you look at my card o zaman êØyle kes ya det skal ikke [//] det fylder meget in that case cut it like this it is not [//] it is quite big Erol bak benim kartÏm gòzel deÈil mi Erol look isn’t my card nice bakayÏm # arkasÏna let me have a look # on the back kesti he has cut Italien Italy êurdan vardÏ bende koparttÏm gitti I had one of these I tore it <êeye sokarsÏn denersin olur # paran olur o zaman xxx>[>] you can put it into that one you can try # then you will have money xxx [< >] cut it off xxx [>] <no I am Danish>[<] reklâmda

140

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

Bilingualism and Social Relations

no I am Danish in the ad he yes no I am Finnish <Morocco>[>] >[<] hello English hello hello I would like a squash hello I would like a squash # I am Danish

In the beginning of Excerpt 3 Ali proposes a picture for inclusion in the task, and his proposal is initially accepted by Esen. However, she changes her position and rejects his proposal – switching to Danish. Ali then turns his attention to Erol, asking for support, but Selma agrees with Esen, and nothing comes of it. The discussion continues, mostly in Turkish, until Ali gets another idea. He finds another advertisement for the SAS airline, developed around the theme ‘Jackpot takes you there’ and he reads it out. This triggers a series of remarks: the girls immediately fall in, Esen even singing the jingle of the corresponding TV commercial, thereby reinforcing the turn of attention from the task to the advertisement theme. Erol develops this line further by contributing a line from another widely played TV-commercial: ‘Are you finished’. In this advertisement, a greengrocer asks an insisting customer – in English: ‘Are you finished?’, and the customer answers: ‘No, I am Danish’. Ali does not immediately realise that Erol has introduced a new ad, so he continues with the jackpot theme. But Selma has noticed, so her following remark is a reaction to Erol’s initiative. Simultaneously with her reaction Erol also continues his initiative – with the same words as Selma, but he adds in Turkish: ‘reklâmda’, thereby explaining the development to Ali. The four children now develop a series of puns on this finish-Finnish word pair. The original pun of the TV-commercial is picked up by Ali who turns it into the opposite: ‘no, I am Finnish’. Esen and Erol react simultaneously. Esen adds ‘Morocco’ – a double reference to the punning game and the map in front of them. Erol also double-comments, as he adds ‘English’ and thereby extends the ‘Finnish’ pun, but he does so in Turkish, and he follows up by reiterating the sketch from the TV-commercial from its beginning: ‘Hello’. Selma’s and Ali’s continuation into ‘I would like a squash’ is a further reference to the TV-commercial. With the change of language back and forth between English, Danish, and Turkish the bilingual adolescents not only play the original punning game, extending it further. They also build up an obviously shared attitude to the use of English. They make fun of the English which is extensively used in ads and slogans in Denmark. This works partly as sheer fun, partly as an alleviation of the disagreement which it follows. At a moment where they do not agree, and being in a situation which may not appeal too much to them, they use linguistic means to establish rapport – perhaps even as face-saving mechanisms. They select linguistic items freely from the languages they meet in their everyday. They choose these items and insert them into new combinations which relate new values and attitudes. The effect is that they all join in a mutual activity of poking fun at the world – through the simultaneous use of three languages. This unites

Languaging Among Fifth Graders

141

them in the situation, and it contributes to bringing about and negotiating shared values which may reach beyond the situation. However, the languages are not used completely arbitrarily. Firstly, the use of English is primarily bound to the texts of the advertisements and TV-commercials, but the children do not restrict their use to verbatim quotations. They also elaborate on the lines of the ads, still using English. Secondly, Turkish seems to be used by Erol for side remarks, particularly to explain (to Ali) what is going on. In this excerpt as opposed to Excerpt 2, the languages do not seem to share their functions. Different codes may be used to separate conversational functions – but they do not have to. When different languages serve different functions, the division is not based entirely on norms of appropriateness (Boyd, 1985), not even on a distinction between we-code and they-code (Gumperz, 1982). Within the same conversation the codes may exchange their functions, for instance so that the other language may become the power language. In such circumstances, code switches do indeed carry local meaning, but it is the code-switch itself, not the direction of it, that matters (cf. Jørgensen, 1998). In Excerpt 4 we find the four speakers a bit later in the conversation than in Excerpt 3. Esen comments on their task (‘this is Greece’), and Selma adds ‘Sealand’ which she produces with an exaggerated intonation contour and an equivocal apposition. In Esen’s reply, the ‘Jackpot takes you there’ theme appears again, as it does more than once in the remainder of the conversation. The success of the jingle that we observed in Excerpt 3 has established it as a theme which the speakers can refer to within this conversation. It is part of the shared frame of reference which Conversation 501 becomes in the course of the conversation. But this is not the only instance of singing and humming. There is singing with Turkish sounds (both Erol and Esen), and there is Selma’s exaggerated intonation (while she speaks Danish) which borders on singing. All the three languages are thus employed within this aspect. Erol interrupts Esen’s singing in order to tease her. He comments on something she has pasted on the cardboard, unfortunately it is unclear exactly what or how: ‘She has pasted it on her (or his) ear’ in Turkish may refer to both Esen’s own ear or somebody else’s ear. Esen sees the comical side of pasting paper on one’s ear and giggles, while Erol follows up on his success. Selma’s remark adds a further layer on to this, as it may both mean ‘do not paste it on her (his) ear’, but also ‘do not paste it on your ear’. Again it appears as if Ali is not quite with them, as he refers to the fact that ‘they’ (probably those who are eventually going to listen to the recordings) can in fact hear what goes on. He may not have realised that the three others are creating yet another frame of reference – the pasting of cut-out paper on somebody’s ears. At least there is a short silence after his remark. Esen breaks the silence with a sentence that is both a reply to Ali’s comment, to Erol who fingers with his microphone, and a reference to her own singing of the Jackpot theme. If indeed Ali here shows that he has not understood the game the others were playing, Esen’s remark is a face-saver on his behalf. We may note that in Excerpt 3, Erol used Turkish in his face-saving efforts on behalf of Ali. From this we can see that a specific function may at one time be served by one language, at another time by another language – within the same conversation and within the same type of speech event.

142

Bilingualism and Social Relations

In the continuation, Erol concentrates on his microphone while Esen tells him there is something wrong. She does so in Danish, but the word for ‘wrong’ is pronounced with a distinctly exaggerated local vernacular pronunciation of the stød. Her remark comes across as playful and not necessarily informative. Her use of the local pronunciation for effect is not a privilege for bilinguals. As pointed out by Kristiansen (1990) and others, Denmark is characterised by a comparatively intolerant sociolinguistic atmosphere in which non-standard varieties are systematically stigmatised. The Sealand stød pronunciation is one stigmatised feature which is sometimes used for comical effects by non-Sealanders, and in such cases it is pronounced precisely the way that Esen uses here. This takes the power out of her somewhat critical remark to Erol. It is further interesting that these Turkish-Danes are Sealanders in their Danish, Esen less than the others. What Esen demonstrates here is her knowledge of the way in which standard speakers regularly exaggerate the Sealand stød as well as her capability and willingness to activate this feature as a mitigating factor in an utterance that is also a face-threat. The use of the Sealand stød becomes an act of solidarity in this context. This is only possible because the feature (i.e. the exaggerated Sealand stød) is out there in Danish society with the negative values that are attached to it among standard speakers. Esen is able to bring this feature into the conversation, turn the values around, and use the feature with positive values. Then Selma intervenes, as she attracts Esen’s attention. It appears that Selma shows Esen a picture, but Esen finds that it is useless for the purpose at hand. It is too pretty, and Esen finds that ‘In Africa it is not that green # and pretty # and beautiful…’. Esen’s first reaction is a sharp rejection, but it is mitigated through her building up of praise for the beauty of the picture that Selma has found. The word ‘beautiful’ amounts to a hyperbole, because of its face value, but first and foremost because it is in English, and they all react to this point with laughter. Her use of the English word which might be risky in other contexts, here helps to alleviate the face-threat involved in rejecting Selma’s proposal. We have already seen that face-saving is done by the participants in this conversation through intersentential code-switching. Erol switches into Turkish to explain the situation to Ali, and Esen switches into Danish under similar circumstances. These last bits have shown us that intrasentential code-switching can work the same way. Esen’s use of stigmatised local Danish in one utterance, and of English in another utterance, have exactly the same effect - they soften otherwise negative contents. Excerpt 4 Esen: Eng: Selma: Eng: Esen: Eng:

det er Grækenland og det # er øh this is Greece and this # is eh Sjælland<dos> [>] Sealand dos (with exaggerated intonation, ‘dos’ equivocal as ‘there are also’ or ‘fool’) [>] yes eh, it is here or there, make it there, Jackpot takes you there di di

Languaging Among Fifth Graders

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

143

[<] yapÏêtÏr kimse duymaz seni do not paste it on there, nobody can hear you daha iyi en +/. even better, the most (giggling) kulaÈÏna yapÏêtÏrmÏs manyak she has pasted it on her (his) ear, fool (they all laugh) Esen kulaÈina yapÏêtÏrmuê manyak. Esen, do not paste it on your (her, his) ear ama Erol duyuyorlar but Erol, they hear it # man kan høre jer ikke mig # (Erol fingers with his microphone) they can hear me, not you <ååååh.> [>] oh (protesting) < dududu> [<] bakÏn mikrofon dududududu. dududu (singing), look at the microphone, dudududu (singing) Erol den duer altså ikke derinde der er noget galt. Erol it does not work inside, there is something wrong (‘galt’ pronounced with an exaggerated Sealand stød) Esen er det ik [/] det er sgu ikke Afrika i Afrika er der ikke så grønt # og smukt # og beautiful eller [/] næh is that no [/] that is bloody not Africa, in Africa it is not that green # and pretty # and beautiful or [/] no (they all laugh) Afrika ne kadar çirkin add how Africa is ugly yerk se lige her bu Afrika mÏ. look here for a moment, is that Africa? åh ja come on (protesting) åååh oh (impressed) yes I am < xxx deÈÏl mi bu> [>] yes I am is that not xxx. [<] < kikicikcik Afrika deÈÏl.>[>] this is not Africa, kikicikcik (singing) this is not Africa [<] dan cici bororrom don cici kulaÈÏmdaki êeyleri duysun böcekleri dan cici bororrom don cici let him hear those sounds in my ear the insects

Reiff (2002) has noted that the punning game which we observed in Excerpt 3 may be described as an instance of performance. The fun that the speakers enjoy together has social psychological consequences – the young people realise that they have fun together, and therefore they are attracted to the in-group members with whom they have already had all this fun. This is not the same as saying that they play with language only or primarily to build the social relations with the others. Fun can be a purpose in itself. Nevertheless, fun can also be used as a tool,

144

Bilingualism and Social Relations

for instance as a power tool in conflicts or as a face-saving measure. And linguistically creative contributions to group conversations often enjoy the same positive reception as other expressive addresses in group interactions. There is value in terms of social accept, and perhaps even of enhanced status, of a positive group action to an inventive expression. The purpose of such performance can therefore also to a certain extent be understood as a social psychological phenomenon. Bauman defines performance as: a mode of communication, a way of speaking, the essence of which resides in the assumption of responsibility to an audience for a display of communicative skill, highlighting the way in which communication is carried out, above and beyond its referential content. From the point of view of the audience, the act of expression on the part of the performer is thus laid open to evaluation for the way it is done, for the relative skill and effectiveness of the performer’s display. Is is also offered for the enhancement of experience, through the present appreciation of the intrinsic qualities of the act of expression itself. Performance thus calls forth special attention to and heightened awareness of both the act of expression and the performer. (Bauman, 1986: 3) Performance as a relevant concept of the linguistic behaviour of grade school students has been demonstrated by Rampton (1999) in his description of Inner London school boys’ use of seemingly unrelated scraps of German picked up from German lessons. Firstly, the German material used by (and probably available to) these boys was indeed very limited. Secondly, the identity function of its use involved both a solidarity and a power dimension. Rampton characterises the use of German as ‘productively related to ritual, music and performance’ (1999: 496), although he also warns that performance as a concept may open a way to understanding routine linguistic production, but ‘it isn’t a free space where cultural materials and social identities are infinitely malleable’ (1999: 499). In Excerpt 5, Esen reacts to Erol’s repeated introduction of Sweden in his search. Esen runs through several different pronunciations of the word ‘Sweden’. Her basic pattern is a way of representing Swedish pronunciation – and especially the word ‘Sweden’ – which is common among Danes. The change of vowels in the three versions in the excerpt (there is more later in the conversation, but that is beyond the point here) bring into focus here Sweden and Swedes, but she also uses it as a point of departure for a fun ride through different pronunciations. The way she forms her utterances, at least the second and third ones, do not contribute content to the ongoing conversation, only entertainment. The entertainment does indeed rely on a Danish routine about Swedes, but it does more than that – it develops the theme and extends its phonetic shape. Thus it is a case of performance, supporting Reiff’s description of the Jackpot theme. Excerpt 5 Esen: Sverige Eng: Sweden (pronounced [sve:rige] with tongue tip r, consonant g, and imitated Swedish intonation) Erol: ha burda yazÏyor ah buraya yapÏêtrayÏm

Languaging Among Fifth Graders

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

145

ha it says it here, let me paste it on here Sverige Sweden (pronounced [sva:riga] with tongue tip r, consonant g, and imitated Swedish intonation) ya so ne Sverige‘si yazÏorsun. which Sweden are you writing valla valla by God küçücük yap Sverige make it very small, Sweden (pronounced [svæ:rigæ] with tongue tip r, consonant g, and imitated Swedish intonation) ne Sv [/] which Sw [/]

In conclusion we have found that the adolescents use the following languages and varieties: Standard Danish, exaggerated Sealand Danish, Turkish, English, stylised Swedish, German. At least, Danish, Turkish, and English are used for both exchange of information, practical task-solving, and administration of social relations in power struggles as well as in face-saving. In some situations, there is no clear difference in function between the languages, in other situations there is. The speakers do not rely on a concept of ‘appropriate’ language choice as defined by society outside their group. They may choose – and switch – between their languages independently of such norms, and they may achieve certain effects by their choices and switches. However, their language use is not completely separated from or ignorant of the norms of society. They, especially Esen, bring into the conversation signals of attitudes that are widely held in Danish society, and they use these values to jointly oppose such attitudes, to play with them, and to create a shared frame of reference which is obviously related to (commercialised) ‘youth culture’. There are many such references to ‘youth culture’, a few references to gender, and hardly any references to ethnicity. The language use that we have witnessed here, is primarily adolescent language. It violates, with premeditation, pleasure, virtuosity, skill and wonderful effects, all the three societal norms that bilingual children are confronted with in Denmark today. With respect to the third norm: the speakers in Conversation 501 curse and yell and scream like all other adolescents who are left alone. They fight, and they are nowhere near being soft in their sometimes playful (and sometimes not so playful) verbal fights. At other times, they can carefully calibrate a sentiment and avoid a face threat through punning, through poking their nose at the outside world, or in other ways. All in all: they speak precisely the kind of adolescent language that is downgraded by adults. With respect to the second norm: the speakers in Conversation 501 use whatever linguistic element or feature they find useful. There is not a single passage that even remotely looks like an attempt to speak one and only one language. This could be mistaken for a specifically ‘bilingual’ characteristic, but it is in fact also what adolescents do elsewhere when they are left alone (more about that below). The speakers violate with obvious pleasure and high sophistication the double

146

Bilingualism and Social Relations

monolingualism norm. With respect to the first norm: The speakers in Conversation 501 show no attempt to accept the language hierarchy which is prevalent in Danish society in general. They distance themselves from values attached to English, exaggerated Sealand Danish, and stylised Swedish. Furthermore, they use at least Danish, English, and Turkish for the same conversational functions regardless of their expected prestige and value. A code-switch may be used for the same purpose regardless of its direction between the codes. Therefore the adolescents show no respect for the first norm of language behaviour either. There is nothing to indicate that the bilingual adolescents in any deeper sense are different from non-bilingual adolescents. They just happen to have a wider range of linguistic resources at their disposal. It would be next to impossible to maintain that there is something relevant of which ‘mono-lingual’ kids have one, and ‘bi-lingual’ kids have two. These ‘bi-lingual’ Turkish-Danes use several languages and varieties, and they cross into even more. Whether there are five, six, seven, or eight varieties is irrelevant. And if a ‘mono-lingual’ adolescent has six such varieties, and a ‘bi-lingual’ adolescent has eight, what is the point then in distinguishing between ‘mono-’ and ‘bi’-lingual? There is no point. We are all first and foremost ‘lingual’, we are languagers (a term proposed by Kanavilil Rajagopalan, personal communication) who possess the uniquely human quality of language, a species-specific phenomenon that we use to maintain mankind as a social species. Correspondence Any correspondence should be directed to Normann Jorgensen, Department of Nordic, University of Copenhagen, Njalsgade 80, DK-2300 Copenhagen S, Denmark ([email protected]). References Akkari, A., Perregaux, C. and Cataffi, F. (1998) Identities among youth in a multicultural neighbourhood. In Migracijske teme. Zagreb: Institut za migracije i narodnosti 14 (4), 259–274. Androutsopoulos, J.K. (1998) Deutsche Jugendsprache: Untersuchungen zu ihren Strukturen und Funktionen. Frankfurt am Main. Androutsopoulos, J. and Scholz, A. (eds) (1998) Jugendsprache, Langue des jeunes, Youth language: linguistische und soziolinguistische Perspektiven. Frankfurt am Main. Auer, P. (1999) From codeswitching via language mixing to fused lects: Toward a dynamic typology of bilingual speech. International Journal of Bilingualism 3 (4), 309–332. Auer, P. and Dirim, I. (2000) On the use of Turkish routines by adolescents of non-Turkish descent in Hamburg. In Holmen and Jørgensen Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe (pp. 157–194). Copenhagen Studies in Bilingualism, the Køge Series, vol. K7. Copenhagen: The Danish University of Education. Bauman, R. (1986) Story, Performance, and Event. Contextual Studies of Oral Narrative. Cambridge: Cambridge University Press. Boyd, S. (1985) Language Survival. A Study of Language Contact, Language Shift and Language Choice in Sweden. Göteborg: Department of Linguistics. Boyd, S., Holmen, A. and Jørgensen, J.N. (eds) (1994) Sprogbrug og sprogvalg blandt indvandrere i Norden, Vol. I–II. Københavnerstudier i tosprogethed bind 22–23. København: Danmarks Lærerhøjskole. Cromdal, J. (2000) Creating a monolingual story in a bilingual conversation. In Holmen and Jørgensen (2000), pp. 57–76).

Languaging Among Fifth Graders

147

Drange, E-M., Kotsinas, U-B. and Stenström, A-B. (eds) (forthcoming) Jallaspråk, slanguage og annet ungdomsspråk i Norden. København: Nord. Eckert, P. (2000) Linguistic Variation as Social Practice. The Linguistic Construction of Identity in Belten High. Oxford: Blackwell. Erikson, E.H. (1971) Identitet, ungdom og kriser. København: Hans Reitzel. Gumperz, J.J. (1982) Discourse Strategies. Cambridge: Cambrige University Press. Haberland, H. (2000) Kan dansk overleve som kultursprog? In A. Holmen and J.N. Jørgensen (eds) Sprogs Status i Danmark år 2011. Københavnerstudier i tosprogetlied, Vol. 32 (pp. 127–138). København: Danmark Pædagogiske Universitet. Haberland, H., Henriksen, C., Phillipson, R. and Skutnabb-Kangas, T. (1991) Tak for mad. Om sprogæderi med dansk som livret. In J.N. Jørgensen (ed.) Det danske sprogs status år 2001 - er dansk et truetsprog? Københavnerstudier i tosprogethed, Vol. 14 (pp. 111–138). København: Danmarks Lærerhojskule. Holmen, A. and Jørgensen, J.N. (eds) (2000) Det er Conversation 801, deÈil mi? Perspectives on the Bilingualism of Turkish Speaking Children and Adolescents in North Western Europe. Copenhagen Studies in Bilingualism, the Køge Series Vol. K7. Copenhagen: The Danish University of Education. Hvenekilde, A. (1994) Innledning. In A. Hvenekilde (ed.) Veier til kunnskap og deltakelse. Utviklingen af grunnskoletilbudet til elever fra språklige minoriteter (pp. 9–24). Oslo: Novus Forlag. Jacobsen, M.B. (2002) Magtens Kode. In J.N. Jørgensen (ed.) De unges sprog (pp. 146–171). København: Akademisk Forlag. Jørgensen, J.N. (1993) Children’s code switching in group conversations. In European Science Foundation Network on Code-Switching (pp. 165–181). Paris: European Science Foundation 1993. Jørgensen, J.N. (1997) Ungdomssprog og gruppeidentitet. Om herlovianersproget, en skoleargot. In Kotsinas et al. Ungdomsspråk i Norden (pp. 136–146). Stockholm. 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. 237–259). London: Routledge. Jørgensen, J.N. (2000) Language hierarchies, bilingualism and minority education in the Nordic countries. In E. Olshtain and G. Horenczyk (eds) Language, Identity and Immigration (pp. 219–237). Jerusalem: Magnes. Jørgensen, J.N. (ed.) (2001) En køn strid. Sprog, magt og køn hos tosprogede unge. Frederiksberg: Roskilde Universitetsforlag. Jørgensen, J.N. (ed.) (forthcoming) Studier i flersprogethed hos unge. København: Akademisk Forlag. Jørgensen, J.N. and Holmen, A. (eds) (1997) The development of successive bilingualism in school-age children. Copenhagen Studies in Bilingualism. 27. 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. Georgakopoulou (eds) Discourse Construction of Youth Identities. Kotsinas, U-B. (1985) Invandrarsvenska och språkförandringar. In Sture Allén (utg.) Svenskans beskrivning 15 (pp. 276–289). Göteborgs Universitet. Kotsinas, U-B. (1988a) Immigrant children’s Swedish – a new variety? In Journal of Multilingual and Multicultural Development 9 (1–2), 129–140. Kotsinas, U-B. (1988b) Rinkebysvenskan – en dialekt? In P. Linell (ed.) Svenskans beskrivning 16 (pp. 264–278). Universitetet i Linköping. Kotsinas, U-B. (1994) Ungdomsspråk. Uppsala: Hallgren and Fallgren. Kotsinas, U-B., Stenström, A-B. and Karlsson, A-M. (eds) (1997) Ungdomsspråk i Norden. Stockholm. Kotsinas, U-B. and Stenström, A-B. (eds) (1999) Ungdom, språk og identitet. København: Nord. Kristiansen, T. (1990) Udtalenormering i skolen. Skitse af en ideologisk bastion. København: Gyldendal.

148

Bilingualism and Social Relations

Kristiansen, T. (1993) Sidste nyt fra kampen mod de unge tunger. In K. Risager, A. Holmen and A. Trosborg Sproglig mangfoldighed – om sproglig viden og bevidsthed (pp. 88–97). Roskilde: ADLA. Kunøe, M. (1991) Ungdomssprog. In Sprogene i sproget. Modersmål-Selskabets Årbog 1991 (pp. 89–96). København: C.A. Reitzel. Landsberg, R.N. (1997) Bilingualism as one language: On the attitude of a group of adolescents to language in a multicultural environment in Sweden. In A. Sjögren (ed.) Language and Environment. A Cultural Approach to Education for Minority and Migrant Students (pp. 167–184). Borkyrka: Multicultural Center. Laursen, H.P. (2001) Magt over sproget. København: Akademisk Forlag. MacWhinney, B. (1995) The Childes Project. Tools for Analyzing Talk. Hillsdale, NJ: Lawrence Erlbaum. See also: http://childes.psy.cmu.edu/pdf/chat.pdf Madsen, L. (2002) De som har skal få. Københavnerstudier i tosprogethed. Køgeserien bind K12. København: DPU. Møller, J. (2001) Identitet og kodevekslen hos unge tosprogede med dansk-tyrkisk baggrund. Københavnerstudier i tosprogethed, Køgeserien bind K8. København: Danmarks Pædagogiske Universitet. Møller, J. and Jørgensen, J.N. (forthcoming) Sprogene på Herlufsholm. Sprog og holdninger i et sprogsamfund af unge. In Drange et al. Jallasprak, slanguage og annet ungdomsspråk i Norden. København: Nord. Nortier, J. (2000) Street language in the Netherlands. In Stenström et al. Ungdommers språkmøter (pp. 129–139). København: Nord. Preisler, B. (1999) Danskerne og det engelske sprog. Frederiksberg: Roskilde Universitetsforlag. Quist, P. (1998) Ind i gruppen, ind i sproget. Københavnerstudier i tosprogethed, Køgeserien bind K5. København: Danmarks Lærerhøjskole. Quist, P. (2000) Ny københavnsk ‘multietnolekt’. Om sprogbrug blandt unge i sprogligt og kulturelt heterogene miljøer. In Danske talesprog bind 1 (pp. 143–211). København: C.A. Reitzel. Rampton, B. (1995) Crossing. Language and Ethnicity Among Adolescents. London: Longman. Rampton, B. (1999) Deutsch in Inner London and the animation of an instructed foreign language. In Journal of Sociolinguistics 3 (4), 480–504. Reiff, K. (2002) Tosprogede dansk-tyrkiske unges brug af engelsk. In J.N. Jørgensen (ed.) De unges sprog (pp. 203–232). København: Akademisk Forlag. Sebba, M. (1993) London Jamaican: Language Systems in Interaction. London: Longman. S pr å kn yt t (2 0 00 ) 1. M e ldin gs b la d fo r No r s k Spr å k r å d. Os lo . (Als o o n http://www.sprakrad.no) Stenström, A-B., Kotsinas, U-B. and Drange, E-M. (eds) (2000) Ungdommers språkmøter. København: Nord. 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.