Paper presented to the ICLaVE I Conference, Barcelona, 29 June 2000
-- DRAFT VERSION --
Date of online publication: 3. August 2000
URL of this page: http://www.rzuser.uni-heidelberg.de/~iandrout/papers/iclavedraft.html
[c] 2000 by Jannis K. Androutsopoulos [email]

 

FROM THE STREETS TO THE SCREENS AND BACK AGAIN:
On the mediated diffusion of ethnolectal patterns in contemporary German*

Jannis Androutsopoulos
(Institut für deutsche Sprache Mannheim)
<androutsopoulos@ids-mannheim.de>


Introduction
This paper is concerned with a language trend which has been going on in the last few years in Germany: patterns of non-native German, often called "Türkendeutsch" ('Turk-German'), are reproduced in different kinds of mass media, and some media reproductions have triggered an imitation trend among native speakers of German. The aim of this paper is to describe the mediated reproduction of such linguistic patterns and their subsequent appropriation by native speakers. As the title indicates, my aim is to work out how these speech patterns move from their community of origin ("the streets") over mediated discourse ("the screens") to the face-to-face communication of native German speakers ("back again"). In this introduction, I will conceptualise my topic with regard to two areas of recent sociolinguistic research, i.e. new language varieties in Europe and language crossing[1]
.
The first development is the emergence of non-standard varieties of the majority language among young speakers of migrant descent. A well-known case is so-called Rinkeby-Swedish, a language variety that emerged in parts of Stockholm among 2nd generation migrants of different ethnic origin (Kotsinas 1989, 1992, 1998). It is characterized through phonetic-phonological, prosodic and grammatical deviations from standard Swedish and its vocabulary contains words and expressions from various home languages, e.g. 'girl' in Arabic and Greek, 'money' in Turkish, etc. Similar phenomena are reported from The Netherlands and France: Sansone (1995: 126) mentions the use of "community Surinamese-Dutch" among second/third generation Surinamese youth in Amsterdam. Dabène & Moore (1995: 38) note cases of "residual bilingualism" among young men of Algerian descent who use Arabic items with phatic and exclamatory functions in a French matrix, and Seux (1997, 94-5) describes a French school argot containing colloquial Arabic items. Language varieties of this kind will be referred to as "ethnolects" (or "ethnolectal varieties") in what follows. An ethnolect, then, is a variety of the majority language (or "host language"), which constitutes a vernacular for speakers of a particular ethnic descent and is marked by certain contact phenomena.[2]
.
The second tendency is language crossing, defined as the adoption of a language or language variety that belongs to another ethnic group (Rampton 1995). Early accounts of crossing involve Anglo adolescents using Afro-Caribbean Creole in the UK (Hewitt 1982). Rampton's own research is concerned with the multilingual practices of ethnically mixed peer-groups in England who use Creole, Stylised Asian English, and Punjabi.Other instances of crossing among young people in Europe involve both minority languages and emerging ethnolectal varieties: Features of Rinkeby-Swedish are spreading to Swedish adolescents (Kotsinas 1992, 1998), and native French speakers with Algerian friends use Arabic exclamations (Dabène & Moore 1995). In Germany, crossing into Turkish in Hamburg is reported by Auer & Dirim (in press), and crossing into non-native German is the topic of the present paper. On the other side of the Atlantic, instances of crossing include the use of Afro-American Vernacular English (Bucholtz 1999, Cutler 1999) and "Mock Spanish" (Hill 1995) among European American speakers. 
.
While literature on crossing is focused on face-to-face interaction, a closer examination reveals several references on media as well. Rampton (1995: 60) acknowledges that models for the language varieties used by his informants are widely available through mass media, though these models are quite different for each language variety concerned. Creole is especially present in music-related media. Through records, white youth come in contact with black speech "alone in their bedrooms", and this may increase their ambition to "act black" in real life (Rampton 1995: 238, 250). On the other hand, Stylised Asian English is reproduced in mainstream British press and broadcast, e.g. in soap operas, and these media stylisations "tied in with the stylised performance of Asian English that was common in local adolescent discourse." (Rampton 1995: 51-2). 
.
Even more references on media and crossing can be found in a recent volume on crossing and styling (Rampton 1999). In an examination of a white adolescent's use of AAVE, Cutler (1999) notes that his sources for language crossing included his rap CD collection as well as rap video clips broadcast on MTV. According to Hill, the use of (fragments of) AAVE into white speech "enjoys enormous circulation in mass-media" (Hill 1999: 549). Mock Spanish witnessed an "explosion" in the mass media in the 90s as well, and some of its media occurrences are reproduced in other media as well as in public and informal speech (Hill 1995). Hill makes a more general point by saying that the "sphere of the market" – a term including both cultural products and commodities – is "an important zone for mixing and crossing" (1999: 547). In summarizing the concerns of recent research on language crossing, Rampton points out "the dense interpenetration of local performance with styles of speech that are reflexively designed, produced and disseminated through mass-institutional and/or electronic communication systems" (Rampton 1999: 423). 
.
In sum, several researchers seem to agree that mass media can provide models that transgress local experiences of language crossing. However, neither these models nor the way particular crossing performances relate to them are examined in detail. While the expositions by Rampton (1995) and Cutler (1999) lead to the suspicion that at least some instances of crossing are verbatim or modified quotations of media phrases, these researchers do not clearly demarcate direct linguistic resources from mediated ones. As a result, the "browsing" of speech patterns between media and local interaction is still far from clear.
 
Aims of this study 
The aim of this paper is to further our understanding of the relation between media and language crossing by examining a recent case of crossing in Germany: Young native speakers of German use patterns of non-native German which are often labelled "Türkendeutsch" or "Türkenslang". However, these labels are not always accurate, because some native German speakers actually imitate non-native German in general, irrespective of its particular ethnic origin. Therefore I will also use the term "ethnolectal German", meaning non-native German of the kind associated with migrant youths. In any case, this is a special case of crossing in that it does not necessarily involve any personal contact between crossers and source speakers. The linguistic material used by crossers comes not from direct interaction, but from different media genres. I will refer to this phenomenon as media-induced crossing
.
The mediated diffusion of "Türkendeutsch" will be reconstructed in this paper in terms of a "life-cycle". The life-cycle metaphor has been used in creole studies in order to model "separate developmental stages in the life of pidgins and creoles" (Mühlhäusler 1986: 135). The development to be discussed here is of a different kind, of course, as it involves the diffusion of socially restricted speech patterns over mass media to speakers of the ethnic majority. However, the life-cycle metaphor still remains useful, because it underlines certain connections between language in the media and in face-to-face interaction. It suggests looking at media language as a part of the global sociolinguistic condition of a speech community, which reflects existing variation, but also introduces new variation resources into the community's repertoire. What makes the German case particular valuable, is the fact that mediated representations of "Türkendeutsch" occurred as a vogue or trend within a couple of years, starting off with certain media products that received wide public attention. Hence, it gives us the chance to describe on-going processes of mediated linguistic diffusion, and to explore sociolinguistically important aspects of media discourse.
.
As this paper's title indicates, the life cycle of ethnolectal German is analytically divided in three distinct stages, which I will now outline together with the main empirical questions of this study. Stage I (the streets) includes sociolinguistic developments which are prior to media-induced crossing, i.e. the emergence of new migration-induced varieties of German, and instances of crossing that are based on direct contact between crossers and source speakers. In stage II (the screens), elements of ethnolectal varieties are represented in different kinds of mass media. The main linguistic question concerns similarities and differences among these media representations as well as their relation to real-life language variation. I will argue that media texts provide stylised versions of ethnolectal speech. Following Selting & Hinnenkamp (1989), stylisation is conceived of as the representation of socially typified meaning patterns in interaction: Styling means presenting oneself as a subsumed instance of a social type, using for this purpose linguistic as well as other semiotic means. What styling demands, is actively working out the typical or core features of this social type. It will be pointed out that various stylised forms of "Türkendeutsch" exist, depending on the particular communicators and genres involved. Finally, stage III (back to the streets) is the appropriation of ethnolectal elements by native German speakers. The notion of appropriation stresses the fact that recipients are not just imitating media fragments, but they may creatively modify them and use them for their own purposes. I will concentrate on the contexts of "Türkendeutsch" use in face-to-face communication and its symbolic associations from the viewpoint of native speakers. Relating to Hill's "dual indexicality" analysis of Mock Spanish (Hill 1995), I will argue that it does not capture all instances of crossing in hand. In some cases at least, the vogue status of "Türkendeutsch" and its media source seem to be more important to speakers' practices than the relation to its ethnic milieu of origin.[3]
.
The research reported in this paper is based on three data sets. The first part of the exposition draws on sociolinguistic interviews I conducted in 1997/1998 with a small group of undergraduate students in Heidelberg. The focus of these "first-stage interviews" was instances of crossing in multi-ethnic adolescent peer groups. The second part is based on the examination of media texts. Beginning in 1998, I kept a record as complete as possible of occurrences of Türkendeutsch, and of the metalinguistic discourse around it. Extracts of literature, comedy acts, rap lyrics, and Internet guest books have been subject to a closer linguistic analysis. The third part of the findings is based on interview discussions conducted with 15 speakers early this year.[4] These "third-stage interviews" do not include any actual, unobserved instances of crossing, but rather second-hand reports, whereby speakers narrate salient episodes of crossing into ethnolectal speech, and comment on different media forms presented to them during the interview.[5] I assume that when speakers report episodes from their own and other speakers' crossing practices, they select episodes that they most vividly recall, and would probably also narrate to other people. Although this data is by no means representative, I believe it is sufficient for a first, explorative presentation of the case in hand. 
.
More than tracing the route of certain speech patterns between media and real-life, this paper is an attempt to explain the remarkable popularity of ethnolectal speech in Germany. The question that asks for an explanation is, in my view, this: What does it take for an instance of "broken" German to be incorporated in the communicative repertoire of a native speaker with no personal contact to non-native speakers? I will argue that the recent popularity of ethnolectal speech in Germany is due to a complex motivation cluster. It involves the associations of the variety crossed into, the attractiveness of foreign linguistic material for young people, and the importance of using media quotations as a demonstration of actuality. Hence, the outlook of this paper will connect media-induced language crossing to a lesser-researched area in sociolinguistics, i.e. the effect of the mass media on people's speech.
 
The streets: ethnolectal varieties and crossing
The emergence of migration-induced bilingualism is a development common to most Western European countries in the second half of the 20th century. The linguistic repertoire of migrant communities can include different varieties of home and host language, and typically shows considerable intergenerational differences (cf. Dabène & Moore 1995, Lüdi 1996, Hinnenkamp 1998, Kotsinas 1998). In Germany, large numbers of migrant workers arrived from southern European countries and Turkey in the '50s and '60s. Similarly to the situation in other European countries, first-generation migrants spoke informally acquired learner varieties of German, which are generally known as Gastarbeiterdeutsch ('guest workers' German'). They display a number of simplification patterns and interferences from the respective home languages (cf. Dittmar 1997: 233-44, Barbour & Stevenson 1998: 214-23). In contrast, second/third generation migrant youths generally pass through the German schooling system. However, the acquisition of a native-like competence in the majority language may be impeded by certain sociodemographic factors, such as growing up in so-called "ghettos"[6] with relatively little contact to native Germans and high use of the home languages. Besides second/third generation youth, the migrant population in Germany includes recently arrived adolescents, i.e. from the Balkans, whose competence in German is developed less.Although the language of migrant youth in Germany has attracted some public and scholarly attention, the current state of research does not yield a fully coherent picture as far as new vernaculars and language crossing are concerned. The following discussion is based on recent findings by two different research projects and my own student interviews. 
.
Research on the in-group speech of German-Turkish peer-groups is currently carried out in Mannheim, a middle-sized industrial city in southern Germany. In discussing the sociolinguistic orientation of migrant youth, researcher Inken Keim proposes a distinction according to the youngsters' orientation "away from the ghetto" or "towards the ghetto" (Keim in press). While the way out of the ghetto is typically based on good grades and leads to a job in mainstream society, youths oriented towards the ghetto are "streetwise", i.e. subculturally oriented, and engage in minor criminal activities. According to Keim, it is in such ghetto-oriented groups that new ethnolectal vernaculars typically emerge.
.
The in-group speech of ghetto-oriented Turkish youths in Mannheim is called Mischsprache (i.e. 'mixed language'), Stadtteilsprache ('hood language') or Stadtteil-Slang ('hood slang'; Keim in press, Kallmeyer 2000). As the name indicates, this Mischsprache is a speech style with a high amount of German-Turkish code mixture, including various patterns of code-switching and transfer (Kallmeyer 2000: 264-5, Hinnenkamp 1998: 153-6). Its German parts display a number of non-native properties, whereby prosodic and phonetic features of Turkish are transferred into German. Prosodically, the continuous change between rising and falling intonation yields a "stomping" speech rhythm. Phonological features include the shortening of long vowels, aspiration of plosives, alveo-palatar instead of palatal voiceless fricatives, simplification of /ts/ to /s/. The example su weisch [su 'waiS] instead of zu weich [tsu 'waic] illustrates both the alveo-palatal fricative and /ts/ simplification. Grammatical features of the Mischsprache include the omission of articles and prepositions, e.g. isch geh [zum] bahnhof ('I go [to the] station') as well as errors in grammatical gender. Its lexicon includes group-specific words, phrases and communicative routines, e.g. the exclamation isch schwör 'I swear' (originally perhaps a calque from Turkish) and the term of address hey lan, whereby lan is equivalent to German Mann ('man'). The speech style of the Kanaken includes a range of ritual insults, such as the utterance siktir lan, isch schwör langer isch mach disch tot ('piss off, I swear, man, I'll make you dead'), the first clause of which is in Turkish.
.
Used by both male and female immigrant youth of migrant descent, this Mischsprache differs from learner varieties of German due to its stabilization (routinization) and socio-symbolic function as an in-group code that is opposed to other varieties of German and Turkish used by the same speakers. Although similar speech styles occur in other German cities as well, they are (still) very diffuse, covering a wide range of regional and social differences. 
.
Adolescent crossing into Turkish has recently been investigated by Auer & Dirim (in press). A major finding of this study is the widespread use of Turkish by non-Turks in multi-ethnic districts of Hamburg. Turkish is not only used when addressing (or in the presence of) Turks, but also among other youth of migrant descent and native Germans. Crossing into Turkish involves a variety of words and utterances, e.g. terms of address, discourse markers, recipient signals, formulaic utterances, greeting sequences, asking what time it is, etc. These findings are corroborated by my first-stage interviews that were conducted in Heidelberg, southern Germany. Informants reported that a number of Turkish words and phrases were common in ethnically mixed peer groups. The most salient ones were greetings, e.g. selan ('hi!'), merhaba ('hello!'), terms of address, e.g. móruk and lan ('man, mate'), "what's happening"-questions such as ne haber lan? (lit. 'what's new?'), terms of abuse and ritual insults, e.g. siktír lan ('get away, piss off') and amuná kojdúm (with a vernacular meaning similar to 'fuck you'). Still other items are used as expressive speech acts (i.e. the formula korrekt, lan!). There are also the social categorisations Lan and Moruk, which are at the same time used as terms of address. The fact that Auer & Dirim's findings suggest much larger amounts of Turkish spreading to non-Turks, could perhaps be attributed to the much larger size of migrant communities and "ghetto"-like districts in Hamburg.
.
From the viewpoint of the present study, more relevant than crossing into Turkish is crossing into "Türkendeutsch", because this variety is stylised in the media and subsequently appropriated by media recipients. According to my student interviews, bits and pieces of non-native German have been routinely used in ethnically mixed adolescent peer groups at least since the early 90s. Typical contexts include acknowledging ignorance, i.e. utterances of the 'I have no idea' type, and especially (playful) "attacking", i.e. utterances of the 'Do you have a problem?' or 'Are you looking at me?' type.[7] In discussing why non-native street-wise accent should appear in this context, my informants suggested that street-wise accent is "fearsome" (furchteinflößend) in the first place. It is specific to aggressive adolescent gangs, called die Schläger ('the bashers') or die Lans (Turkish for 'guys'), which also include recently arrived youths. Since Germans are afraid of such gangs, a German adolescent engaged in a fighting would have more chances of scaring his opponent and imposing respect by using this accent. Therefore, the use of non-native accent ny native speakers could evoke qualities of "dangerous", anti-social behaviour.[8] Judging from the third-stage interviews conducted in early 2000, there seems to be a widespread awareness of this speech among German adolescents. Informants reported that many of their age cohorts have had random, and often unpleasant encounters with migrant adolescent gangs, such as being threatened or even robbed out. Crossing into a sort of "broken" German may occur right after "having had stress with the Lans", as an informant put it, or while recalling such situations. For these speakers, mimicking non-native German could even be a kind of symbolic revenge for uncomfortable past situations. 
.
Both instances of crossing mentioned so far, i.e. into Turkish and "Türkendeutsch", are based on local experience. The social meaning of crossing resources is established within networks of the local community. It is exactly this precondition that dramatically changes at the next stage.
 


The screens: mediated stylisations of ethnolectal speech 
 

Placing mediated stylisation after direct crossing is based both on conceptual decisions and on chronological evidence. Conceptually, I assume that the mediated occurrence of ethnolectal variation reflects existing, real-life sociolinguistic patterns (albeit in a stylised manner, as will be shown below). Chronologically, many (though not all) of my first-stage informants were familiar with ethnolectal German from the schoolyard, the local youth centre or their adolescent hangouts. While the mediated "Türkendeutsch" trend started off around late 1997, these speakers had experienced real-life crossing since the early 90s. In other words, crossing also occurred well before ethnolectal resources became widely available through media. 
.
Although it is strictly speaking not possible to list all media instances involving ethnolectal speech in the last three years in Germany, it is safe to say that they cover a wide range of mostly fictional, written and spoken genres. Spoken genres include comedy acts, radio shows, song lyrics, broadcast commercials, and films. Written genres include literature, comic strips, and instances of computer-mediated communication. Clearly, not all of these genres are equally suited for the dissemination of sociolinguistic patterns. Evidence suggests that audio-visual products with a nation-wide distribution, i.e. broadcast, films and audio CDs, have a potentially wider impact on recipients than written fiction or other written media. 
.
The earliest mediated reproduction of migrant speech in my data is Feridun Zaimoglu's book "Kanak Attak", published in 1995. However, the trend's starting point most probably was a film called "Knocking on heaven's door" which was screened in autumn of 1997, featuring young actor Moritz Bleibtreu in the role of Turkish gangster "Abdul". For many German adolescents, Abdul's broken German apparently was their first encounter with Türkendeutsch, and has been vividly mimicked in the months following the film.