Critical Perspectives on Linguistic Fixity and Fluidity: Languagised Lives 1138602973, 9781138602977

Citation preview

Critical Perspectives on Linguistic Fixity and Fluidity

This volume offers a critical perspective on current views on linguistic fixity and fluidity in sociolinguistics and highlights empirical accounts alternative to prevailing trends in the field. Featuring accounts from a broad range of regional contexts, the collection takes stock of such terms as “polylingualism”, “metrolingualism”, and “translanguaging” to question perceptions around multilingual and monolingual language use. The book critiques the status of fluid language use as a more “natural” language practice and in turn, its greater potential for corresponding social transformation, demonstrating the value of linguistic fixity and the continuous debate between fixity and fluidity in multilingual speakers’ lives. In providing these accounts, the book seeks not to advocate for linguistic fixity or fluidity, but to argue that sociolinguists pay close attention to the way both types of linguistic practice open up or close down avenues for social transformation. This collection is a key reading for graduate students and scholars in sociolinguistics, multilingualism, and linguistic anthropology. Jürgen Jaspers is Associate Professor of Dutch Linguistics at the Université Libre de Bruxelles (ULB), Belgium. Lian Malai Madsen is Associate Professor of the Psychology of Language at the University of Copenhagen, Denmark.

Routledge Critical Studies in Multilingualism Edited by Marilyn Martin-Jones MOSAIC Centre for Research on Multilingualism, University of Birmingham, UK Joan Pujolar Cos Universitat Oberta de Catalunya, Spain

Entangled Discourses South-North Orders of Visibility Edited by Caroline Kerfoot and Kenneth Hyltenstam Standardizing Minority Languages Competing Ideologies of Authority and Authenticity in the Global Periphery Edited by Pia Lane, James Costa, and Haley De Korne Multilingual Brazil Resources, Identities and Ideologies in a Globalized World Edited by Marilda C. Cavalcanti and Terezinha M. Maher Queer, Latinx, and Bilingual Narrative Resources in the Negotiation of Identities Holly R. Cashman Language and Culture on the Margins Global/Local Interactions Edited by Sjaak Kroon and Jos Swanenberg Agency in Language Policy and Planning Critical Inquiries Edited by Jeremie Bouchard and Gregory Paul Glasgow Researching Agency in Language Policy and Planning Edited by Gregory Paul Glasgow and Jeremie Bouchard Critical Perspectives on Linguistic Fixity and Fluidity Languagised Lives Edited by Jürgen Jaspers and Lian Malai Madsen For more information about this series, please visit: www.routledge.com/ Routledge-Critical-Studies-in-Multilingualism/book-series/RCSM09

Critical Perspectives on Linguistic Fixity and Fluidity Languagised Lives Edited by Jürgen Jaspers and Lian Malai Madsen

First published 2019 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2019 Taylor & Francis The right of the editors to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Jaspers, Jèurgen. | Madsen, Lian Malai. Title: Critical perspectives on linguistic fixity and fluidity : languagised lives / edited by Jèurgen Jaspers and Lian Malai Madsen. Description: New York ; London : Routledge, 2018. | Series: Routledge critical studies in multilingualism ; 19 | Includes bibliographical references and index. Identifiers: LCCN 2018051843 | ISBN 9781138602977 (hardback) Subjects: LCSH: Translanguaging (Linguistics)—Case studies. | Multilingualism—Case studies. | Sociolinguistics—Case studies. Classification: LCC P115.35 .C75 2018 | DDC 306.44—dc23 LC record available at https://lccn.loc.gov/2018051843 ISBN: 978-1-138-60297-7 (hbk) ISBN: 978-0-429-46931-2 (ebk) Typeset in Sabon by Apex CoVantage, LLC

Contents

List of Tables and Figuresvii   1 Fixity and Fluidity in Sociolinguistic Theory and Practice

1

JÜRGEN JASPERS AND LIAN MALAI MADSEN

PART I

Negotiating Fixity

27

  2 Recognizing Languages, Practicing Languaging

29

JANUS SPINDLER MØLLER

  3 Languagised Repertoires: How Fictional Languages Have Real Effects

53

KATHARINA RUUSKA

  4 Lingoing and Everyday Metrolingual Metalanguage

76

ALASTAIR PENNYCOOK AND EMI OTSUJI

PART II

Pursuing Real Languages

97

  5 Carving Out Breathing Spaces for Galician: New Speakers’ Investment in Monolingual Practices

99

BERNADETTE O’ROURKE

  6 Transformative Multilingualism? Class, Race and Linguistic Repertoires in Hong Kong KARA FLEMING

122

vi  Contents   7 The Symbolic Organization of Languages in a High-Prestige School

146

THOMAS RØRBECK NØRREBY AND LIAN MALAI MADSEN

PART III

The Meaning Potential of Fluid Language

165

  8 Inarticulate Voices: Translanguaging in an Ecology of Conflict 167 PANAYIOTA CHARALAMBOUS, CONSTADINA CHARALAMBOUS AND MICHALINOS ZEMBYLAS

  9 Speaker Perspectives, Linguistic Hybridity, and Language Learning

192

URSULA RITZAU AND LIAN MALAI MADSEN

PART IV

Dilemmas and Dialectics

215

10 The Deliberative Teacher: Wavering Between Linguistic Uniformity and Diversity

217

JÜRGEN JASPERS

11 Register Processes in Contemporary South African Schools: Dialectics of Fixity and Fluidity

241

JAMES COLLINS AND LARA-STEPHANIE KRAUSE

12 Discussion: The Elephant in Every Room

266

ROBERT MOORE

Contributors Index

285 290

Tables and Figures

Tables 7.1 8.1 10.1 12.1

Emphasis on French/Danish Divisions on School Webpage Pupils’ Ethnic and Linguistic Backgrounds Data from Four Linguistic-Ethnographic Studies Forms of Value and Scales of Comparison

150 177 224 271

Figures 6.1 Perceived Repertoires and Deficits 6.2 Expected Linguistic Trajectories

136 139

1 Fixity and Fluidity in Sociolinguistic Theory and Practice Jürgen Jaspers and Lian Malai Madsen

Introduction Multilingualism is an established topic in the study of language. Selfevident as this statement is, it is useful to remember that up until some 60  years ago it was something of a fringe topic, at best, in the broad field of linguistics, pursued by scholars who dared defy the motto that real linguists should investigate the systematicity of speakers’ cognitive abilities for producing language by relying on introspective intuitions of their own (and singular) native language. Studying multilingualism at the same time has long run up against popular assumptions that it is an exceptional, deviant phenomenon, or that it delays learning and causes linguistic deficiency—assumptions indebted to a predominant view, at least in the West, that communities are naturally monolingual, and to common sense metaphors of the mind as a limited container with room for only one complete language. That the first sentence on multilingualism on the Linguistic Society of America’s website, to take one professional organisation, still posits that ‘[c]ontrary to what is often believed, most of the world’s population is bilingual or multilingual’, illustrates the extent to which a scientific interest in the topic has been accompanied by normalising it as a social fact. While it would be rash to claim that these fears and misgivings are a thing of the past, it is difficult to ignore the currently more receptive societal climate for multilingualism. Just as diversity has become a corporate value, managerial and policy discourse (especially at EU level) is now rife with praise for multilingualism. In addition, the wider public has not been insensitive to its value, bearing in mind the growing interest in linguistic immersion or content and language integrated learning, attracted by promises of cognitive advantages or increased job opportunities. Multilingualism today matters, then, even if it matters mostly in economic rather than cultural terms, and even if, as several authors have pointed out, the type of multilingualism that is seen to matter in fact concerns a set of parallel monolingualisms, preferably of Western-European cut (see, for example, Heller 2007; Moore 2015).

2  Jürgen Jaspers and Lian Malai Madsen It is somewhat ironic, therefore, that multilingualism as a term, just when it has become a self-evident topic in the academy and beyond, is facing increasing dissatisfaction among sociolinguists who have been proposing a range of alternatives such as translanguaging, polylingualism and metrolingualism, among others. The main reason for proposing these and other terms is that they shift away from a focus on multiplying, switching or mixing distinct codes whenever speakers combine features that conventionally belong to separate ‘languages’, in favour of a focus on how speakers flexibly combine linguistic features of whatever pedigree, in line with local perceptions of language. Describing behaviour as multilingual may be correct if it corresponds with participants’ conscious alternation of what they see as several linguistic codes (‘French’, ‘Danish’). But on other occasions, such a description is jumping to conclusions, it is argued, since speakers may not regard what they say as ‘using multiple codes’ but as ‘the normal way of speaking’ or as ‘using one code’ if not a ‘continuously changing’ one. The prefix ‘multi’ thus merely pluralises language; it does not help with understanding its complexity (Otsuji & Pennycook 2010: 243). Sociolinguists moreover maintain that the occasions where multilingualism is inadequate for capturing language use are starkly increasing in (often urban) settings characterised by intense migration, mobility and communication technology, and that fluid language use is unnecessarily stigmatised or misrecognised, especially in education. The newly proposed terms for language thus strike several birds with one stone: they evoke a theoretical perspective that overcomes conventional ways of understanding language; they describe fluid linguistic behaviour that deviates from these conventional understandings; and they help draw attention to the normality and the ubiquity of linguistic practices that policymakers are hesitant to register on their discursive radar. This explains at least partly the current popularity of new terms for multilingualism. Translanguaging, polylanguaging and metrolingualism, to take three of the best-known ones, have each appeared in the title of highly cited publications, and translanguaging in particular has commanded a great deal of attention. There is little that seems to impede the eventual recognition of these terms as the label for a separate province of sociolinguistic research, or as the foundational precursor to a so-called ‘new sociolinguistics’ which more accurately addresses the complexity and unpredictability of contemporary communication. Pending this development, however, we believe that the production and uptake of these terms, and the interest in linguistic fluidity that they represent, merit closer consideration as a scholarly phenomenon, and that an analysis of this phenomenon can contribute to understanding its present salience in sociolinguistics, to examining its viability, and to gauging its potential impact on the world this discipline takes as its object (cf. Salö 2017).

Fixity and Fluidity in Sociolinguistics  3 In this introduction we draw attention to three reasons for doing so. A  first one is that these terms epitomise an anti-canonical standpoint, against linguistic fixity, the urgency of which appears to invite over-usage given that each term is recruited for descriptive, theoretical and ideological purposes at once. Rather than more, this leads to less precision: the fluidity these terms identify spills over into their flexible application. A  second reason is that scholars’ interest in highlighting linguistic fluidity contributes to viewing an investment in linguistic fixity as a sign of false consciousness or conservatism, at the same time as the idea of separate languages is difficult to avoid epistemically and ideologically. The third reason is the principal, but problematic, association of fluid language with ideas of natural language, liberation or transformation. In what follows we give a fuller account of each of these arguments, before explaining how the different chapters in this volume contribute to their elaboration.

Post-language theory Recent sociolinguistics has seen numerous new terms for characterising the flexible use of linguistic resources commonly associated with separate languages. In addition to polylingualism (Jørgensen 2008; Jørgensen & Møller 2014), translanguaging (García & Li 2014) and metrolingualism (Otsuji & Pennycook 2010), there is codemeshing (Canagarajah 2011), transidiomatic practices (Jacquemet 2005), truncated multilingualism (Blommaert et  al. 2005), flexible bilingualism (Creese  & ­Blackledge 2011), heteroglossia (Bailey 2007) and multilanguaging (Nguyen 2012). These terms add to a range of predecessors such as code-switching, code-mixing, crossing (Rampton 1995), fused lects (Auer 1999) and dual lingualism (Lincoln 1975), among others. If daunting at first sight, this terminological profusion usefully allows to distinguish, say, the idea of speakers who communicate each in a preferred, but mutually different, language (‘dual lingualism’) from the short-lived, exploratory experiments into a local out-group’s language (‘crossing’), and the unpredictable combination of linguistic features from diverse sources (‘polylingualism’). It is not unusual either to see several new terms highlight an important insight, and that from the total set of alternatives available at one point, only one or two withstand the test of academic criticism. The recent stream of terms epitomises an anti-canonical stance. Several sociolinguists have repeatedly insisted that the idea of separate, fixed languages must be seen as an invention and a socio-political tool of nationalist and colonialist projects that long went under the name of ‘modernity’ (Bauman  & Briggs 2003; Heller 2007; Makoni  & ­Pennycook 2007). They have argued in addition that these inventions, and the homogeneous speech communities they are commonly associated

4  Jürgen Jaspers and Lian Malai Madsen with, are empirically inadequate for addressing an era of unprecedented sociolinguistic complexity (Blommaert 2010) or the ‘complex linguistic realities of the 21st century’ (Li 2018: 14) which have invited ‘new ways of being in the world’ (García & Li 2014: 9). Because these inventions and concomitant assumptions ‘are so deeply embedded in predominant paradigms of language studies that they are rarely questioned’ (Otsuji & Pennycook 2010: 251), it is recommended to reconsider the existing disciplinary vocabulary: although notions like ‘native speaker’, ‘mother tongue’ and ‘ethnolinguistic group’ have considerable ideological force (and as such should certainly feature as objects of analysis), they should have no place in the sociolinguistic toolkit itself. (Blommaert & Rampton 2011: 4–5) Research on code-switching has been frequently lampooned in this regard as a domain that unnecessarily assumes, upon observing speakers’ use of linguistic features that are usually seen to belong to different ‘languages’, that such speakers combine autonomous linguistic entities. This is not only seen as empirically problematic, but also as reproducing the conditions for problematising speakers as deficient: it implies a view of the ‘perfect’ bilingual as a double monolingual, and of all other bilinguals as people who have difficulty keeping their two codes apart. Terms like metrolingualism or translanguaging are part of an attempt to go beyond these assumptions and undesirable consequences. Apart from spawning a flourishing of new terms, this anti-canonical stance has been productive in other ways. It has invited scholars to denaturalise common frameworks for language by revealing the continuous activity that is required to construct the idea of separate languages, and by exposing the impact of this activity on the organisation of unequal social relations; it has inspired scholars to focus on non-native, unusual, self-conscious language use in multi-ethnic, ephemeral, practice-related communities rather than on non-self-conscious, regular, native speech in established ethnic or social groups; and it has invited a focus on language as an emergent property of interaction, that is, on how the use of linguistic features co-occurs with other semiotic sign forms, and on how the recurrent production and evaluation of such conglomerates of verbal and non-verbal resources can over time lead to their recognition as a distinctive model for interaction (Agha 2007). Anti-canonical trends are not without risks, however. As Rob Moore (2007), an educational sociologist, argues, they are likely to invite schismatic discourse. Such discourse temporalises the social world and the research that addresses it by announcing a radical break with the past, both in the sense of postulating a new epoch that is fundamentally different from the previous one, and of carving up the scientific field in

Fixity and Fluidity in Sociolinguistics 5 traditional pre- and innovative post-approaches. Typical of such discourse ‘is its claim to originality—it is being thought for the first time’ (Moore 2007: 44). But underneath this rhetoric it has a paradigmatic form in that it ‘appears episodically across the intellectual field in its various disciplines and sub-disciplines’ (Moore 2007: 44) to ‘adjus[t] the time frames of intellectual production—rearranging the history of scholarship and the sense of continuities and discontinuities’ (2007: 40). Rather than minimising the value of incisive criticism or the possibility of societal change, Moore cautions that the logic of schism leans towards a simplified representation of both sides of the dichotomy, and to a relative lack of engagement with earlier work that offered similar insights or that addressed comparable circumstances. While Moore was mainly concerned with how postmodern approaches are distinguished from modern ones, the risks he points up may not be entirely hypothetical in this context (as is noted by Blommaert & Rampton 2011: 3; Makoni & Pennycook 2007: 3; Otsuji & Pennycook 2010: 245–246) when we observe, first, that the idea of languages as distinct codes has a long history of criticism (cf. Makoni & Pennycook 2007: 3). It will soon be half a century since Haugen claimed that the concept of a language as a rigid, monolithic structure is false, even if it has proved to be a useful fiction in the development of linguistics. It is the kind of simplification that is necessary at a certain stage of science, but which can now be replaced by more sophisticated models. (1972: 335) A decade later Strevens pointed out that A central problem of linguistic study is how to reconcile a convenient and necessary fiction with a great mass of inconvenient facts. The fiction is the notion of a ‘language’—English, Chinese, Navajo, Kashmiri. The facts reside in the mass of diversity exhibited in the actual performance of individuals when they use a given language. (1982: 23 in Kemp 2009: 16) Other authors have put forward similar ideas around that time (Ferguson 1982; Gumperz 1982; Harris 1981; Hymes 1973; Silverstein 1979) as well as a century earlier (cf. the references to Schuchardt 1884 and 1909 in Auer 2007 and Piller 2016). This suggests that the insight that the idea of autonomous languages is useful but reductive has long been a part of the sociolinguistic canon, if it has not been relatively mainstream, seeing as it has been formulated by some of the best and most widely acclaimed sociolinguists. Rather than incarnating a radical turn in the discipline, then, the current search for new terms to address the ‘inconvenient facts’ seems to stem from a long-recognised problem, one that probably haunts

6  Jürgen Jaspers and Lian Malai Madsen the entire discipline of linguistics—famously exemplified in Chomsky’s claim that performance features are irrelevant to analysing competence, whereby he recognised the reality of their inconvenience. Some authors have remarked, secondly, that the perception of substantial societal change, especially in relation to linguistic diversity, may have to be seen as a contemporary Western (European) impression of conditions that it would be difficult to characterise as radically different from what can be observed in contemporary non-Western settings and from historical linguistic practices across the globe (see, e.g., Lucassen & Lucassen 2013 and Mackey 2005, cited in in Wiese 2018; Pavlenko 2018; Piller 2016; also see Collins & Krause, this volume). The representation of extraordinary complexity today may thus depend on the relative oblivion of complexity elsewhere and before (cf. Makoni & Pennycook 2007). Thirdly, while research on code-switching often serves as a negative counter-example to a proper analysis of contemporary language use, this strand of research itself problematises the use of ‘a “language” [as] a prime of linguistic analysis’ (Auer 2007: 320; MacSwan 2017). To take one example, the key argument of Auer’s (1998) edited volume on conversational code-switching and code-mixing (which includes authors like Jørgensen and Li) is that verbal actions must be understood not by subsuming (‘coding’) them under pre-established external categories, but by explicating the systematic resources that members of a community, as participants in a conversation, have at their disposal in order to arrive at interpretations of ‘what is meant’ by a particular utterance in its context. (Auer 1998: 2) Especially with regard to code-switching, this approach is argued to have far-reaching consequences, since what linguists tend to take for granted as ‘codes’ (and hence classify as ‘code-switching’) may not be looked upon as ‘codes’ by members/participants [. . .] [I]t is not the existence of certain codes which takes priority, but the function of a certain transition in interaction. (Auer 1998: 2, 13, 15) The goal for linguists is then to find when participants meaningfully orient to a juxtaposition between sets of co-occurring linguistic features, which must be done by investigating the possible conversational function of such a contrast and the social indexicality of the features in question. In principle, therefore, there is no guarantee that what participants consider different codes on one occasion they will see as different codes on the next—especially not in communities where a mixed code, and eventually

Fixity and Fluidity in Sociolinguistics  7 a fused lect, is being developed on the basis of frequent, thus gradually less salient, alternations of linguistic features (Auer 1998: 16–21). Some of the chapters in the volume subsequently demonstrate that what a linguist may on structural grounds be tempted to define as codeswitching has to be understood as one ‘code-switched code’, a code in its own right that speakers meaningfully alternate with another inherently mixed one (Meeuwis & Blommaert 1998). Similarly, speech that appears to be the ‘use of one code’ is pointed out as a meaningful alternation of different, objectively closely related, codes (Alvarez-Cáccamo 1998). The point to draw from this is that the evaluation by a linguist that ‘a given arrangement of signs [is] a combination of elements of two systems is not only very difficult to make at times [. . .], it is also irrelevant’ (Auer 1998: 13; our emphasis). Other chapters in Auer’s volume underline that objectively hybrid linguistic practices cannot be taken as a straight sign that speakers have elaborate competence in the source varieties, and that such practices neither occur exclusively to project particular social identities. Indeed these practices may equally result from enduring restricted competence, from temporarily ‘crossing’ into linguistic varieties one is not usually seen to own (cf. Rampton 1998), and in many cases they serve the sequential organisation of talk (Auer 1984; Meeuwis & Blommaert 1998). The following terms for different types of linguistic fluidity result from this: code-switching (further differentiated according to its discursive and participant-related function), mixed code and crossing. Which is the appropriate term is presented as an empirical question based on participant understandings that are to be explored through a combination of conversation analysis and ethnographic accounts. Clearly, then, a critical concern with the idea of autonomous languages and a delicate approach to linguistic fluidity predate more recent calls that this is urgent. Let us now turn to newly proposed terms to see how they address a similar concern.

Terminological expanse Recent new terms are inspired by an ontology that understands language (use) as not driven by use of ‘a language’, but by people’s attempt to create meaning in context through deploying linguistic sign forms and accompanying non-linguistic signs (gestures, facial expressions and so on) which, as a result of their usage in earlier contexts, create expectations about the entities that can be referred to and about the type of social occasion people find themselves in (cf. Agha 2007; Silverstein 1985). In such a view people gradually become more familiar with the sets of linguistic resources and co-occurring non-linguistic signs that make sense in specific, and, in the course of their life trajectory, changing social settings.

8  Jürgen Jaspers and Lian Malai Madsen Complete knowledge of ‘a language’ is therefore impossible, if not difficult to imagine, since people only ever become familiar with, and to an even lesser extent competent in, a limited set of the ways in which linguistic and co-occurring sign forms can be used; they also lose familiarity with sign forms that are typically produced in social settings they have stopped participating in. One of the propositions in this regard is to describe communication through language as an act of ‘languaging’. This avoids the assumption of different languages in objective statements made by linguists, leaving it to the latter to find out how people perceive their languaging acts (e.g., as ‘switching codes’, ‘code mixing’, ‘pure language’, and so on) (Jørgensen 2003, 2008; García & Li 2014). Translanguaging, metrolingualism and polylanguaging can then be understood as labels that draw attention to specific types of languaging which defy, transcend or otherwise inflect established types of languaging (‘mono-languaging’), without multiplying languages (‘multilingualism’). Contrary to the Auer volume, however, these terms are not part of a single theoretical attempt to identify and explain different languaging types, but each function as standalone terms to highlight that a particular type of languaging cannot be confused with switching or mixing codes. These terms moreover do not allow to make further distinctions within non- or lesser-established types of languaging. So, while it can be useful to label speakers’ meaningful transgression of conventional models for languaging as an act of ‘translanguaging’, it is less useful that no other term is available to characterise those practices where speakers unintentionally transgress such models (Ritzau  & Madsen 2016), temporarily experiment with other people’s languaging models (as in ‘crossing’), meaningfully switch between languaging models or habitually adhere to a languaging model that only appears transgressive to those who have a different one. Metrolingualism as a term similarly and helpfully highlights the existence of fluid language practices, but although it is presented as ‘a broad, descriptive category for data analysis’ (Otsuji  & Pennycook 2010: 245) and as ‘a way of describing diverse grounded local practices’ (2010: 248), it does not offer further detail about how these practices may differ in form and function and by which other terms this can be described. Terms like polylanguaging, metrolingualism and translanguaging have furthermore acquired a range of additional meanings. Metrolingualism and translanguaging have both been presented as broad synonyms for ‘languaging’. Metrolingualism is suggested to ‘accommodate the complex ways in which fluid and fixed, as well as global and local, practices reconstitute language and identities’ (Otsuji  & Pennycook 2010: 244). Translanguaging scholars suggest that ‘[h]uman beings have a natural translanguaging instinct’ (García & Li 2014: 32) or see translanguaging as the unrestrained performance of ‘individual linguistic competences of speakers the world over, irrespective of whether we call them monolingual

Fixity and Fluidity in Sociolinguistics  9 or multilingual’ (Otheguy et al. 2015: 286). This raises the question what the prefixes ‘trans’ and ‘metro’ bring to bear. If metrolingualism refers to fluid as well as fixed practices, and moreover ‘as a practice is not confined to the city’ (Otsuji & Pennycook 2010: 245), the term may draw undue attention to the metropolitan melting pot. Likewise, if translanguaging is a natural capacity which logically also reproduces those types of languaging commonly known as monolingual language use, ‘languaging’ would avoid the assumption of a particular transgression where none is occurring in participants’ perception. Polylanguaging, in its turn, is not only used to refer to an unconventional type of languaging, but also lends its name to a ‘polylingualism norm’ (Jørgensen 2008; Jørgensen et  al. 2011; Møller 2016). This entails that ‘language users may know—and use—the fact that some of the [linguistic features] are perceived by some speakers as not belonging together’ (Jørgensen et al. 2011: 34), but also that speakers ‘employ linguistic features associated with different languages as a matter of habit’ (Jørgensen & Møller 2014: 73). More or less the same term is thus used to name speakers’ deliberate transgression of existing languaging models and to label a type of languaging where speakers habitually orient to a different, possibly local or less established, model for languaging. That this latter model is prefixed with ‘poly’ raises the question whether it is the observed speakers’ or the analysts’ idea of where linguistic features belong that inspires its name (cf. Jaspers & Madsen 2016). Some of these terms become passe-partout labels when it appears that they can name a particular theory of language as well as a languagepolitical project. So, apart from its descriptive function, metrolingualism also denominates an approach to language: Metrolingualism describes the ways in which people of different and mixed backgrounds use, play with and negotiate identities through language; it does not assume connections between language, culture, ethnicity, nationality or geography, but rather seeks to explore how such relations are produced, resisted, defied or rearranged; its focus is not on language systems but on languages as emergent from contexts of interaction’ (Otsuji & Pennycook 2010: 246) what [. . .] sets metrolingualism apart is its productive power to overcome common ways of framing language, its capacity to deal with contemporary language practices, and its ability to accommodate both fixity and fluidity in its approach to mobile language use. (Otsuji & Pennycook 2010: 252) Such quotes imply that metrolingualism is largely synonymous with (an undeniably interesting version of) ‘sociolinguistics’. But apart from the

10  Jürgen Jaspers and Lian Malai Madsen fact that this makes the object and the approach to that object go under the same name, with all the attendant risks of misunderstanding, metrolingualism is also understood in a more restrictive sense, as a political project against established ortholinguistic models for languaging: We are interested in the queering of ortholinguistic practices across time and space [. . .] We locate metrolingualism [. . .] as another practice of undoing, as both a rejection of ortholinguistic practices and a production of new possibilities. (Otsuji & Pennycook 2010: 246–247) Such a political project is legitimate, but its priorities may invite friction with the scientific approach that carries the same name, and that promises in principle also to describe and explain how ortholinguistic practices emerge from interaction and are adhered to by ‘straight’ speakers. Translanguaging similarly names an approach to language: [T]ranslanguaging is an approach to the use of language, bilingualism and the education of bilinguals that considers the language practices of bilinguals not as two autonomous language systems as has been traditionally the case, but as one linguistic repertoire with features that have been societally constructed as belonging to two separate languages. (García & Li 2014: 2) Note that here, in contrast with its definition as a universal instinct (including so-called monolingual speakers), translanguaging only refers to bilinguals’ use of language. More important is that this name for a broad approach to language and education is equally given to a specific educational philosophy, and to linguistic practices that demonstrate resistance to existing language models. For, as García & Li (2014: 74) argue, ‘[a] translanguaging theory in education views the incorporation of the students’ full linguistic repertoire as simply the only way to go about developing language practices in school, as well as to educate’. At the same time, translanguaging ‘refers to language actions that enact a political process of social and subjectivity formation which resists the asymmetries of power that language and other meaning-making codes, associated with one or another nationalist ideology, produce’ (García & Li 2014: 43). Consequently, translanguaging can denote a universal instinct, the fluid language of bilinguals, an approach to language, an educational philosophy, and resistant linguistic practices—not to mention that it retains its original pedagogical sense as the planned alternation of different languages in class. This chameleonic character makes translanguaging a term with a rich employability, but it makes it a challenge to know which of its meanings is in play (cf. Jaspers 2018a).

Fixity and Fluidity in Sociolinguistics  11 So, while there is no lack of new terms for describing fluid language, at least some of these are made to do far more work than their precursors in language contact research, functioning as mid-level concepts that seem more intended to sensitise us to particular types of languaging than to help us dive into the fray of their detailed interactional analysis. Such a strategy may be worthwhile in a world that is oblivious about, or hesitant to recognise, the diversity of languaging types that exist in it. The polysemy of scientific terms as such may not be a guaranteed problem either, as is demonstrated by the uptake and impact of terms like identity, ideology and discourse. Cameron (2001: 17) explains that discourse is several things at once. It is a method for doing social research; it is a body of empirical knowledge about how talk and text are organized; it is the home of various theories about the nature and workings of human communication, and also of theories about the construction and reproduction of social reality. It is both about language and about life. Such a polysemy is challenging but understandable in light of the term’s use in different domains across the humanities. Yet if ‘[a]ny term which tries to cover too much threatens to cancel all the way through and end up signifying nothing’, as Eagleton (1994: 11) warned in relation to ideology, it may be less than convenient, unless precision isn’t a priority, that new terms that have been proposed only quite recently attain a comparable ambiguity within a single discipline like sociolinguistics. We now turn to a second reason for critically reviewing the recent interest in fluid language, which is that it contributes to abnormalising an investment in its opposite, despite the reasonable epistemic and ideological grounds for doing so, and to downplaying the dilemmas that arise from living in a considerably ‘languagised’ world.

Languagised lives It would be an oversimplification to situate scholars of fluid language on one side of a fixity-fluidity dichotomy. Otsuji  & Pennycook (2010: 244), for example, argue that while one of their aims is ‘to demythologise notions of language mixing along the fault line of bilingualism, another is to demythologise hybridity as if cultural and linguistic fixity also were not part of its apparatus’. Hence, ‘[w]e cannot leap into an examination (or celebration) of hybridity as if fixed ascriptions of identity and their common mobilisation in daily interaction have ceased to exist’ (2010: 244), because what looks fluid may well have been inspired by mobilising fixed categories, while linguistic fixity may be informed by an awareness of fluidity. They conclude therefore that ‘it is important not to construe fixity and fluidity as dichotomous [. . .] but rather to view them as

12  Jürgen Jaspers and Lian Malai Madsen symbiotically (re)constituting each other’ (2010: 244; cf. also Jørgensen 2008; Møller 2016). We agree with this view. Yet it is the fluidity side of this symbiotic union that captures most of the current sociolinguistic interest, with fixity often relegated to antagonist status as ‘th[e] very language ideolog[y] that we need to supersede’ (Otsuji & Pennycook 2010: 251). An example of this is the influential revisiting of Gumperz’s (1964) notion of the linguistic repertoire (Blommaert & Backus 2011; cf. also Busch 2012). While Gumperz saw repertoires as speakers’ know-how of linguistic resources, following from their membership of relatively stable speech communities, Blommaert and Backus argue that this is hard to maintain in a world where intensified mobility and communication technologies make individuals experience a variety of more or less short-term and fluid types of groupness. Consequently, the linguistic knowledge individuals acquire must be related to their ‘active usage and passive exposure’ (Blommaert & Backus 2011: 5) to language in the inconstant collectivities that they move between. This makes variation and change ‘natural design features of language’ (Blommaert & Backus 2011: 6), and it results in speakers’ development of a dynamic ‘inventory of resources’ that includes linguistic elements besides ‘anything that speakers use to communicate meaning’ (Blommaert  & Backus 2011: 7). Far from offering a community-based snapshot of styles and registers, then, repertoire becomes in this view a linguistic track record of individual biographies, and this invites a challenge of the idea of complete knowledge of a language: [T]he ‘language’ we know is never finished [. . .] and learning language as a linguistic and a sociolinguistic system is not a cumulative process; it is rather a process of growth, of sequential learning of certain registers, styles, genres and linguistic varieties while shedding or altering previously existing ones. Consequently, there is no point in life in which anyone can claim to know all the resources of a language. (Blommaert & Backus 2011: 9) Thus Blommaert and Backus move away from a conception of multilingualism as complete knowledge of separate languages: they call for accounts that are ‘no longer trapped by a priori conceptions of language, knowledge (competence, cognition) and community’ (2011: 24) but that attend to how individuals display types of knowledge as different as speaking fluently, making oneself understood, or recognising a language one does not speak at all. Underlying their argument is a concern with justice: if no one is fully competent in a language, it is questionable to evaluate speakers according to benchmarks inspired by such competence (cf. Extra et al. 2009; Hogan-Brun et al. 2009). Hence, ‘a more practical (or polemical) motive’

Fixity and Fluidity in Sociolinguistics  13 of their work is to take issue with ‘dominant discourses [that] seem to increasingly turn to entirely obsolete and conclusively discredited models of language knowledge’ (Blommaert & Backus 2011: 4), such as the European Common Framework of Reference for Languages and its measuring bureaucracy. The spearhead of the attack is the precision of their approach: since this produces a more complex account, official instruments based on a view of separate languages must be considered ‘a form of science fiction’ since they ‘have only a tenuous connection with the real competences of people, the way they are organized in actual repertoires, and the real possibilities they offer for communication’ (Blommaert  & Backus 2011: 24). Despite this opposition to science fiction, it appears that it can be reasonable to recycle terms that reproduce it. For, when Blommaert and Backus demonstrate the complexity of the first author’s linguistic repertoire, they indicate that it counts 38 ‘languages’, and then break that number down into smaller figures for the ‘languages’ in which he has obtained maximal (two), partial (six), and minimal competence (eight), and for those he merely recognises (22). Aware of the fact that invoking ‘languages’ is problematic, they explain that the ‘reason for doing so is primarily didactic: we must start from a widely known vocabulary and attempt to refine it’ (2011: 2). The reductive terminology is thus a temporary, pragmatic solution: it makes the complexity of a repertoire more tractable than when each of the counted ‘languages’ would have had to be broken up in countless, nameless ways of speaking, writing, reading and recognising. Reverting to separate language terms, in other words, serves the epistemic purpose of idealisation, that is, ‘the intentional introduction of distortion into scientific theories’ (Weisberg 2007: 639; cited in MacSwan 2017: 175). The question is if this merely is a temporary move, since the point of evoking languages here does not seem to have been to provide a provisional description while we await a more sophisticated vocabulary, but to provide a minimalist idealisation, one that deliberately abstracts away from a more complex reality in order to clarify complexity through simplicity, that is, to elucidate with sufficient rather than maximal accuracy how layered and diverse actual repertoires can be. After all, ‘minimalist idealizers are not interested in generating the most truthful or accurate model. Rather, they are concerned with  [.  .  .] discovering the core factors responsible for the target phenomenon’ (Weisberg 2007: 655)—what they lose in completeness they gain in explanatory acuity. This suggests that sociolinguists may have more options than chucking the idea of a separate language out of their toolkit, since depending on their (pedagogical, explanatory, etc.) purpose and theoretical ideal (completeness, simplicity, predictive power, etc.), it can be epistemically defensible to ignore an excess of variation and to evoke the fiction of language—or, indeed, the fiction of style, register, genre, variety (MacSwan 2017: 175–177; cf. also Jørgensen 2010: 37).

14  Jürgen Jaspers and Lian Malai Madsen For MacSwan, the analytic utility of such constructs trumps their political (ab)use. But on an epistemic and political level, positing the existence of ‘a language’ is probably neither entirely beyond the pale nor wholly unproblematic (cf. Makoni  & Pennycook 2007: 27): sociolinguists themselves have recruited analytic idealisations to dignify marginalised groups as systematic rather than disfluent speakers, with positive but also problematic outcomes (Jaspers 2016). Also, the inverse strategy—that is, positing that ‘languages do not exist’—is neither unreasonable nor trouble-free. Such a claim usefully denaturalises predominant views and opens up innovative ways of thinking about language, but we have just seen that to radically maintain it can be needlessly complicated for at least some epistemic purposes. Also, on a political level, it can be problematic radically to argue that precise knowledge of what language is disqualifies official models based on a different conception of language, or to advance a radical strategy of ‘disinvention’ to counter the ‘epistemic violence’ of invented languages many people have to endure (Makoni & Pennycook 2007: 16). For, while it is necessary to expose the unreliability of tests, the inaccuracy of the language model they are based on, and the undesirable effects that ensue from them, one could very well imagine the implication of highly precise language models and reliable tests in unjust conditions, or experiences of injustice when no test is applied and incompetence is no objection. Rather than on their precision alone, the fairness of such models would also seem to depend on why they are used, by and for whom, on which occasions, and with what kind of impact on those to whom they are applied. This also means that, if precision is not enough to decide on the political suitability of language models or tests, representing the issue as a scientific one threatens to depict political values as a matter of evidence, or to place the debate in a context of a widely shared value (the truth about what language is) rather than in a context of conflicting values about what type of language is desirable, where and when (cf. Pielke 2007: 43ff.). Such a strategy in fact follows a ‘linear model’ of science according to which ‘achieving agreement on scientific knowledge is a prerequisite for a political consensus to be reached and then policy action to occur’ (Pielke 2007: 13). From the perspective of this model, science is central to politics, and because scientific understandings are supposed to motivate political action, winning a scientific debate leads to a privileged position in political battle [. . .] Science thus becomes a convenient and necessary means for removing certain options from a debate without explicitly dealing with a dispute over values [. . .] For who can argue against truth? (2007: 124–125)

Fixity and Fluidity in Sociolinguistics  15 Typical of this model is to argue that a particular policy is based on junk science (or is obsolete, discredited) to imply that the ‘political agendas following from that science must be ill conceived and not deserving of support’ (2007: 126). But such a strategy does not suffice to determine whether to choose, say, for a more precise language test, for no test at all, or for a different course of action, since ‘what is factually true does not automatically render it desirable; we cannot simply induce what we should do by accurately observing what is currently being done, any more than we can induce the facts from the way we think things ought to be’ (Cameron 1995: 227). What ought to be done ultimately is a political, not a scientific, matter, although facts are important in debating the options. And to the extent that scholars represent the matter at hand as one of ‘what is’ rather than ‘what ought’ in order to disqualify particular options from the debate, they ultimately risk compromising the credibility of their insights if the public and decision makers start seeing these as an extension of scholars’ political stance. Because the odds for this are high when there is intense conflict over values, Pielke (2007: 135ff.) argues that a more sustainable role for science in such contexts resides in expanding decision makers’ set of alternatives, in sketching the relation between potential choices and outcomes and in proposing innovative options that may allow for compromise between conflicted parties. A radical insistence on the fictional nature of linguistic categories for political reasons, moreover, contributes to obscuring the value-based nature of these categories, their ‘social truth’ so to say, and the real dilemmas that they pose for ordinary speakers and sociolinguists alike. Makoni & Pennycook (2007: 3) concede in this regard that ‘while the entities around which battles are fought, tests are constructed and language policies are written are inventions, the effects are very real’. Yet the epistemological battle (2007: 21) they propose in response to this again reduces the scope of acceptable options and leaves little alternative than to identify as a symptom of false consciousness the fact that ‘languages are salient dimensions of [people’s] sense of self [. . .] some people’s “identity” is inexorably linked to their “language” ’ (Blackledge  & Creese 2010: 17) or that ‘particular languages clearly are for many people an important and constitutive factor of their individual, and at times, collective identities’ (May 2005: 330), and this may pave the way for various kinds of academically endorsed pressure and blame (cf. Jaspers 2018a, 2018b). When we see such involvements as reflections of a particular, predominant and debatable view of desirable language that competes with other views in an undeniably real political conflict, linguistic fixity and fluidity can be understood as contradictory social values, neither of which is by definition anathema to a justifiable political project, since this depends on considerations of desirability, feasibility and viability. Such a view, moreover, allows one to see fixity and

16  Jürgen Jaspers and Lian Malai Madsen fluidity as mutually presupposing values: it would not make sense to disprove or oppose linguistic fixity if it was not at the same time considered a socially relevant value; at the same time, the existence of linguistic fluidity is clearly presupposed, if not feared, by those who value linguistic fixity (for an analogous argument, see Billig et al. 1988: 125ff.). Without linguistic fixity, in other words, there is no fluidity, and vice versa (cf. McDermott & Varenne 1995; also see Busch 2012; Otsuji & Pennycook 2010: 244). The two are part of the same linguistic culture, where their mutual association with widely shared liberal values, such as communication, equality, inclusion and emancipation, invites dilemmas in everyday life and academia (cf. ­Billig et  al. 1988). We call this culture a ‘languagised’ one to underline the principal, enduring social value of separate languages in it, despite the insight that its proponents cannot legitimise their pursuit of this value as based on facts; calling it ‘languagised’ does not exclude the existence of linguistic fluidity in that culture, nor minimise the value it has for other parties. Indeed, it highlights that adhering to the latter value is fraught with difficulties and invites ambivalent responses. Such responses transpire when we see sociolinguists’ advocacy of linguistic fluidity coincide with their continuing professional investment in a register of English that eventually adheres to, and reproduces the conditions for recognising as a benchmark, the idea of a monolingual standard language. A similar response emerges when translanguaging scholars, in contrast to their claim that including students’ full linguistic repertoire ‘[i]s simply the only way [. . .] to educate’ (García & Li 2014: 74) propose that ‘students need practice and engagement in translanguaging, as much as they need practice of standard features used for academic purposes’ (García  & Li 2014: 71–72; also see Otheguy et  al. 2015: 283), which only seems possible by actively reducing, at least some of the time, students’ linguistic repertoires at school. Rather than being inconsistent, such responses demonstrate the complexity of dealing with a culture where linguistic fluidity and fixity presuppose each other and are both, albeit to a different extent, tied to widely shared ideals. So, instead of relegating fixity to the sociolinguistic scrapheap and viewing fluidity as the only ‘real’ target phenomenon, it may be more useful epistemically to explain how, why and with what outcomes laypeople and academics (dis)align themselves with both linguistic fixity and fluidity in the course of their social trajectories (cf. Pérez-Milans & Soto 2017 for a similar argument). On a political level, following Pielke (2007), we argue that in a context of enduring value-conflict over what language use is desirable, sociolinguists’ credibility and impact on policymaking diminishes if they represent the issue as a strictly scientific one, and that a more effective strategy resides in prefiguring the effects of several policy options and proposing innovative courses of action that may reconcile opposing parties.

Fixity and Fluidity in Sociolinguistics  17

Natural, transforming language use A third reason to look critically upon new terms for language (use) is that they are particularly associated with ideas of natural language, and with assumptions that recognising, or promoting, these natural practices will entail a form of liberation and social transformation. These assumptions also need closer consideration. Statements to the effect that natural, everyday language emerges in counterpoint to a notion of non-natural, ideologically constrained language can be found elsewhere in contemporary sociolinguistics. Shohamy, for example, argues that language is personal, open, free, dynamic, creative and constantly evolving. This concept of language does not have the boundaries of language x or language y, since it spreads beyond words and is manifested through a variety of multi-modal representations and different forms of ‘languaging’. Yet, in spite of such views, language is commonly perceived as closed, stagnated and rule-bound. Language is manipulated, as it is used as a symbol of unity, loyalty, patriotism, inclusion and legitimacy, especially by various collective groups. (2006: xvi) Such an argument evokes a Rousseaunian perspective on language use in which individuals’ authentic, in this case linguistic, nature suffers from the manipulative incursions of society. A similar conception emerges in research on translanguaging where a pre-social idiolect—‘the mental or psychological sense [which] encompasses the billions of individual linguistic competences of speakers the world over, irrespective of whether we call them monolingual or multilingual’ (Otheguy et al. 2015: 286)— is distinguished from ‘the social walls erected by the named languages’ (2015: 304) and where it is suggested that these named languages ‘have nothing to do [. . .] with the billions of the world’s idiolects, which exist in a separate, linguistically unnamed and socially undifferentiated mental realm’ (Otheguy et al. 2015: 293). In this ontology it becomes natural to see linguistic freedom as the unencumbered realisation of fluid language, and to see translanguaging as ‘the act of deploying all of the speakers’ lexical and structural resources freely’ (Otheguy et al. 2015: 297; emphasis in original), or ‘as the full use of idiolectal repertoires without regard for named-language boundaries’ (Otheguy et al. 2015: 304). Such a view, in effect, pulls up a wall between the mental and the social world to contest the existence of social walls for language. Linguistic anthropologists argue, however, that language must be seen as a phenomenon that is ‘irreducibly dialectic in nature [. . .] [as] an unstable mutual interaction of meaningful sign forms, contextualized to situations of interested human use and mediated by the fact of

18  Jürgen Jaspers and Lian Malai Madsen cultural ideology’ (Silverstein 1985: 220). In such a view the occurrence of linguistic features is inevitably tied to a specific context and will be somehow informed and understood by the ideas people have of culture, language and other people. On a developmental level, too, linguistic anthropologists insist on the socialising, and thus ‘manipulative’, nature of human interaction. How children learn to act and speak, as Kulick and Schieffelin (2004) point out, is profoundly influenced by parents’ ideological views on language, interaction, learning, culture and by what they assume that infants are. There is no escape in this perspective from the impact of ideology, no ‘unwalled’ language, although ideology does not totally determine speakers’ choices. Usage-based accounts similarly argue that children’s acquisition of linguistic competence results from their socialisation in specific localities: rather than instinctive or natural, therefore, ‘ways of using language across “language” boundaries are not “natural” but are learned in the same way other kinds of discourse patterns are learned’ (Ruuska, this volume). MacSwan (2017) claims that from a syntactic perspective, an explanation of the available bilingual data requires the postulation of at least some internal differentiation that is related to socially recognised ways of speaking (not necessarily ‘named languages’). Applying fluid language use is at the same time frequently considered to be liberating or transformative. Translanguaging research in particular argues that introducing fluid language at school is ‘transformative for the child, for the teacher and for education itself’ (García  & Li 2014: 68) because it enables pupils to construct and constantly modify their sociocultural identities and values, as they respond to their historical and present conditions critically and creatively. It enables students to contest the ‘one language only’ or ‘one language at a time’ ideologies of monolingual and traditional bilingual classrooms. (García & Li 2014: 67) Flores & García (2013: 246) likewise claim that allowing fluid language at school has the capacity to release buried histories, and encourages the emergence of ‘new subjectivities [.  .  .] that defy ethnolinguistic identities defined by a nation-state/colonial paradigm’, while García (2017: 24) maintains that translanguaging ‘gives agency to minoritized speakers, decolonizes linguistic knowledge, and engages all of us in the social transformations that the world so sorely needs today’. Frequently it is suggested that allowing translanguaging at school provides a ‘third space’ (Bhabha 1994) that generates new, now hybrid, identities, values and practices (Flores & García 2013; García & Li 2014; Li 2011). It is possible of course that allowing fluid language at school creates new opportunities for pupils and teachers, and that they experience this

Fixity and Fluidity in Sociolinguistics  19 as transformative. Yet expecting all of the above from fluid language ‘may be seen as a case of claiming quite a lot for a change in pedagogical practice’ (Block 2018; Jaspers 2018a). Linguistic anthropologists maintain as well that ‘all subjectivities [. . .] are negotiated and achieved, not given’ (Kulick  & Schieffelin 2004: 350). This entails, first, that subjectivities cannot simply be ‘set free’, but depend on their consistent cultivation— a process that requires distinction, and thus, new ‘walls’. Second, this cultivation of new, now fluid, linguistic subjectivities inevitably implies the simultaneous invention of new ‘bad subjects’ (2004: 354–360) who do not correspond to the valued subjectivity. New social walls can be expected to emerge as well through celebrating a ‘third’ space ‘which, if it is a space, must have boundaries of its own, and thus be based on oppositions to its own others’ (Friedman 1997: 79, in Allatson 2001: 199). Charalambous et al. (this volume) furthermore show that benevolently introducing fluid language at school can also fail to be seen as liberating, and may even lead to silence among pupils: in the Greek-Cypriot school they describe Turkish-speaking pupils as anxious to be seen as speakers of the aggressor’s language in that context. This illustrates that, however well-intentioned, pluralising language at school, or inviting pupils to use linguistic resources from the home in class, is always contextualised and thus inf(l)ected by immediate as well as by larger-scale conditions and expectations for language. Liberation, release or transformation are not dependent on the type of language itself, but on the relation between language, participants and setting, which in its turn needs to be understood against the background of wider-scale language ideologies and socio-­economic processes (Block 2018; Lorente  & Tupas 2014). Fluid language use therefore is, like hybridity, ‘neither intrinsically nor invariably a radical marker of subversiveness’ (Allatson 2001: 196). Its effects are a matter of empirical observation. Coupland (2001: 369; also see Canagarajah 2011) made a similar point when he argued that The social theoretic and discourse analytic wings of sociolinguistics are increasingly receptive to ideas of cultural hybridity [. . .] but perhaps sometimes uncritically. To recognize the view that social identities can be, or even necessarily are, multiple is usually argued to be a liberal and progressive antidote to assuming that each person inhabits a given and single social identity. This traditional and allegedly repressive ideology is called essentialism. The idea of authentic cultural experience is held to be essentializing and therefore suspect. But it is surely simplistic to posit that multiplicity and hybridity are necessarily good and that essence and authenticity are necessarily bad. There are challenging and intriguing qualifications and interaction effects. For example, the claim that social and cultural identities are necessarily hybrid in the late-modern world does undervalue many people’s

20  Jürgen Jaspers and Lian Malai Madsen experience of group membership. Welshness remains, in some ways and for some people, an essential and defining quality, not only an authentic experience but an authenticating one. We should not be too ready to dismiss cultural essentialism as if it were inevitably a naïve or pernicious assumption. Karim 1997, Said 1978, and others show that it can also be this, especially when a cultural uniformity is imputed by powerful outgroups with vested interests in containing and perhaps exoticizing cultural ‘Others’ [. . .] The conditions of late-modernity add layers of complexity and conditionality to many people’s senses of cultural essence, but they do not simply neutralize them. In this view it is conceivable that ‘obsolete’ models for language open up new horizons, serve to construct new subjectivities, or function as a linguistic compass in a confusing world. Rather than trying to swing the pendulum back towards monolingualism, though, this is to argue that sociolinguistics should concern itself with explaining the transformative potential of all language use, regardless of its fixed or fluid nature.

The chapters in this volume In this introduction we have taken a critical look at some new labels for language (use). We have addressed the conceptual imprecision entailed by their varied use, and the tendency to view opposition to fluid language practices as conservative or poorly informed, as well as the trend to imbue these practices with liberation or transformation. The rest of this volume brings together empirically based observations that in various ways address these assumptions. The chapters below illuminate how lives are, or become, ‘languagised’, how languages can be seen as valid and valuable in speakers’ lives, and how an aspiration for language(s) in the traditional sense does not solely amount to inequality or oppression. The first section of the book comprises chapters that each investigate how speakers’ lives in various places are ‘languagised’, that is, how speakers come to recognise established categories for language, how they reconcile them with their own practices and different purposes, and how they use these very categories, whether or not in line with their official conception, in their dealings with others. So, on the basis of a longitudinal study of speakers of Turkish minority background in a Danish suburb and on data collected in relation to a three-year long ethnographic study at a linguistically diverse public school in Denmark, Janus Møller describes how young Danes over time acquire an awareness of how their habitual linguistic practices (which combine resources from Turkish and Danish) relate to established models for ‘Turkish’ and ‘Danish’ and subsequently use this metalinguistic knowledge in interaction to (dis)align with these models. Katharina Ruuska draws on data from interviews with highly proficient adult

Fixity and Fluidity in Sociolinguistics  21 foreign language speakers of Finnish to investigate how their linguistic competence can be understood as fashioned in relation to the linguistic ideologies that surround them. She argues that an ability to keep ‘languages’ apart and to display this ability in relevant contexts may have to be seen as an intrinsic part of competent multilingual speakers in their particular, in this case European, context. The third chapter in this section, by Alastair Pennycook & Emi Otsuji, explores how workers in food markets and restaurant kitchens deploy language labels in diverse and flexible ways. This metrolingual metalanguage should not be taken at face value, they argue, but investigated for how Australians of different backgrounds constantly rearrange their relations, spaces, and speakers, and, in the process, reconfigure the meanings of language names. The second section of this volume concentrates on how conventional linguistic categories can, despite their seeming old-fashionedness, be valid in people’s life plans and conducive to the emergence of new subjectivities—as well as new ‘bad subjects’. In this regard, the chapter by Bernadette O’Rourke explores why and with which effects speakers invest in delineated minority languages rather than in a more fluid linguistic practice that combines resources from a minority and majority language. Drawing on interviews with speakers seeking to acquire Galician, a minoritised language in north-western Spain, she shows that investing in a ‘pure’ minority language competence is a valid concern for many speakers, which produces new subjectivities but also leads to new tensions and opportunities for sociolinguistic hierarchisation. Taking the case of Hong Kong, Kara Fleming in her chapter demonstrates how an official policy of ‘biliteracy and trilingualism’ can thwart social mobility when groups are differentiated as to the degree they meet the ‘ideal’ biliterate and trilingual. She describes how South Asians, in spite of their competence in English and Cantonese, continue to be constructed as disfluent, socially excluded and economically problematic and argues that multilingualism in this context acts as a smokescreen to hide race- and class-based stratification, even if South Asians invest in ­English as a way of transcending their predicament. In showing how types of multilingualism can be socially hierarchised, Fleming hints at how potential policies in favour of fluid language are not immune to being inflected by long-standing socio-economic inequalities. Thomas Rørbeck ­Nørreby & Lian Malai Madsen subsequently describe linguistic interaction in a French private school in Copenhagen, Denmark, that is, in an institution that epitomises the authority and success of ‘real’ language. On the basis of ethnographic fieldwork, Nørreby & Madsen explore how this investment in real language plays out interactionally in and outside of class, how it is legitimised through linking it up with a cosmopolitan ethos, and how pupils reconcile the linguistic regime at school with the wider sociolinguistic economy of Denmark.

22  Jürgen Jaspers and Lian Malai Madsen The third section in this book explores different meanings of fluid language use than the usually transgressive or liberating qualities with which it is often associated. Thus, the chapter by Panayiota Charalambous, Constadina Charalambous & Michalinos Zembylas explores how the well-meant introduction of Turkish in a Greek-Cypriot primary school class was resented by ­Turkish-speaking pupils as an unwelcome opportunity for others to perceive their ‘speaking Turkish’ as a sign of ‘being Turkish’, which is (still) not held in high regard in Greek Cyprus despite the peace-building efforts. Rather than liberation, translanguaging in this context contributed to awkward silence and feelings of insecurity. Ursula Ritzau & Lian Malai Madsen’s chapter illustrates that linguistic hybridity needs to be examined in close connection to its local and conversational settings. Consequently, what looks like ‘polylanguaging’ to the sociolinguist may need to be explained as oriented to an idea of ‘pure’ language. They argue this on the basis of data collected among Swiss-German university learners of Danish as a foreign language, showing how these students sometimes indeed knowingly align (and amuse) themselves with ideas of hybrid language, while in other cases their hybrid language is undeniably oriented to learning Danish. The fourth and final section of this book contains chapters that demonstrate the simultaneous push and pull of monolingual vs. plurilingual and standardising vs. vernacularising tendencies, showing that this often produces dilemmas and makeshift strategies rather than clear-cut choices. Jürgen Jaspers starts from the fact that Flemish teachers in Belgium are consistently found to report negative attitudes towards their pupils’ use of non-Dutch home languages, and that these attitudes are seen to be detrimental to pupils’ well-being and learning outcomes, and reflect teachers’ monolingual habitus. On the basis of ethnographic research at four secondary schools, he argues, however, that these teachers frequently, in contrast to their rigid attitudes, vacillate between linguistic uniformity and diversity because language-political concerns are cross-cut by pedagogical and organisational ones. Instead of unthinking agents of monolingualism, teachers must thus be approached as deliberative thinkers, in this and many other contexts. James Collins & Lara-Stephanie Krause in  their chapter examines how teachers and multilingual student bodies in two South African public schools in the metropolitan Cape Town region co-construct linguistic and social hierarchies. Although different from each other in terms of school language and pupils’ linguistic background, Collins & Krause show on the basis of classroom interaction, as well as school staff and out-of-school commentary, that in each school, teacher and pupils negotiated tensions between, and hierarchically organised, notions of standard vs. vernacular varieties of Xhosa, Afrikaans and English; in so doing they took up differing, ambiguous positions vis-à-vis

Fixity and Fluidity in Sociolinguistics  23 these varieties as they grow up in a society where identity politics and social mobility aspirations take shape against the backdrop of the history of apartheid. The volume is concluded with a discussion by Robert Moore.

References Agha, A. (2007). Language and social relations. Cambridge: Cambridge University Press. Allatson, P. (2001). Beyond the hybrid. Genre, 22, 191–207. Alvarez-Cáccamo, C. (1998). From ‘switching code’ to ‘code-switching’. In P. Auer (Ed.), Code-switching in conversation (pp. 29–48). London: Routledge. Auer, P. (1984). On the meaning of conversational code-switching. In P. Auer (Ed.), Interpretive sociolinguistics (pp. 87–112). Tübingen: Narr. Auer, P. (Ed.) (1998). Code-switching in conversation. London: Routledge. Auer, P. (1999). From codeswitching via language mixing to fused lects. International Journal of Bilingualism, 3, 309–332. Auer, P. (2007). The monolingual bias in bilingualism research, or: Why bilingual talk is (still) a challenge for linguistics. In M. Heller (Ed.), Bilingualism: A social approach (pp. 319–339). New York: Palgrave Macmillan. Bailey, B. (2007). Heteroglossia and boundaries. In M. Heller (Ed.), Bilingualism: A social approach (pp. 257–274). New York: Palgrave Macmillan. Bauman, R., & Briggs, C. (2003). Voices of modernity. Cambridge: Cambridge University Press. Bhabha, H. (1994). The location of culture. London: Routledge. Billig, M., Condor, S., Edwards, D., Gane, M., Middleton, D.,  & Radley, A. (1988). Ideological dilemmas. London: Sage. Blackledge, A.,  & Creese, A. (2010). Multilingualism: A critical perspective. ­London: Continuum. Block, D. (2018). The political economy of language education research (or the lack thereof): Nancy Fraser and the case of translanguaging. Critical Inquiry in Language Studies (in press) Blommaert, J. (2010). The sociolinguistics of globalization. Cambridge: ­Cambridge University Press. Blommaert, J., & Backus, A. (2011). Repertoires revisited: ‘knowing language’ in superdiversity. Working Papers in Urban Language and Literacies, 67. Blommaert, J., & Rampton, B. (2011). Language and superdiversity. Diversities, 13, 1–21. Blommaert, J., Collins, J., & Slembrouck, S. (2005). Spaces of multilingualism. Language & Communication, 25, 197–216. Busch, B. (2012). The linguistic repertoire revisited. Applied Linguistics, 33, 503–523. Cameron, D. (1995). Verbal hygiene. London: Routledge. Cameron, D. (2001). Working with spoken discourse. London: Sage. Canagarajah, S. (2011). Codemeshing in academic writing. The Modern Language Journal, 95, 401–417. Coupland, N. (2001). Dialect stylisation in radio talk. Language in Society, 30, 345–375.

24  Jürgen Jaspers and Lian Malai Madsen Creese, A., & Blackledge, A. (2011). Separate and flexible bilingualism in complementary schools. Journal of Pragmatics, 43, 1196–1208. Eagleton, T. (1994). Ideology. London: Routledge. Extra, G., Spotti, M., & Van Avermaet, P. (2009). Language testing, migration and citizenship. London: Bloomsbury. Ferguson, C. (1982). Foreword. In B. Kachru (Ed.), The other tongue (pp. vii–xi). Oxford: Pergamon. Flores, N., & García, O. (2013). Linguistic third spaces in education. In D. Little, C. Leung & P. Van Avermaet (Eds.), Managing diversity in education (pp. 243– 256). Bristol: Multilingual Matters. Friedman, J. (1997). Global crises, the struggle for cultural identity and intellectual porkbarreling. In P. Werbner & P. Modood (Eds.), Debating cultural hybridity (pp. 70–89). London: Zed. García, O. (2017). Problematizing linguistic integration of migrants. In J.-C. Beacco, H. Krumm, D. Little & P. Thagott (Eds.), The linguistic integration of adult migrants (pp. 11–26). Berlin: De Gruyter Mouton. García, O., & Li, W. (2014). Translanguaging. Basingstoke: Palgrave Macmillan. Gumperz, J. J. (1964). Linguistic and social interaction in two communities. American Anthropologist, 66, 137–153. Gumperz, J. J. (1982). Discourse strategies. Cambridge: Cambridge University Press. Harris, R. (1981). The language myth. London: Duckworth. Haugen, E. (1972). The ecology of language. Stanford: Stanford University Press. Heller, M. (2007). Bilingualism as ideology and practice. In M. Heller (Ed.), Bilingualism. A social approach (pp. 1–22). Basingstoke : Palgrave Macmillan. Hogan-Brun, G., Mar-Molinero, C., & Stevenson, P. (Eds.) (2009). Discourses on language and integration. Amsterdam-Philadelphia: John Benjamins. Hymes, D. (1973). Speech and language. Daedalus, 102, 59–86. Jacquemet, M. (2005). Transidiomatic practices. Language and power in the age of globalization. Language & Communication, 25, 257–277. Jaspers, J. (2016). (Dis)fluency. Annual Review of Anthropology, 45, 147–162. Jaspers, J. (2018a). The transformative limits of translanguaging. Language & Communication, 58, 1–10. Jaspers, J. (2018b). Language education policy and sociolinguistics. In J. Tollefson & M. Pérez-Milans (Eds.), The Oxford handbook on language policy and planning (pp. 704–723). Oxford: Oxford University Press. Jaspers, J.,  & Madsen, L. M. (2016). Sociolinguistics in a languagised world. Introduction. Applied Linguistics Review, 7, 235–258. Jørgensen, J. N. (2003). Languaging among fifth graders. Journal of Multilingual and Multicultural Development, 24, 126–148. Jørgensen, J. N. (2008). Polylingual languaging around and among adolescents. International Journal of Multilingualism, 5, 161–176. Jørgensen, J. N. (2010). Languaging. Nine years of poly-lingual development of young Turkish-Danish grade school students, vol. I–II. Copenhagen: University of Copenhagen. Jørgensen, J. N., Karrebæk, M. S., Madsen, L. M., & Møller, J. S. (2011). Polylanguaging in superdiversity. Diversities, 13, 23–38. Jørgensen, J. N.,  & Møller, J. S. (2014). Polylingualism and languaging. In C. Leung  & B. Street (Eds.), The Routledge companion to English studies (pp. 67–83). London: Routledge.

Fixity and Fluidity in Sociolinguistics  25 Karim, H. K. (1997). The historical resilience of primary stereotypes. In S. H. Riggins (Ed.), The language and politics of exclusion (pp. 153–183). Thousand Oaks, CA: Sage. Kemp, C. (2009). Defining multilingualism. In L. Aronin & Britta Hufeisen (Eds.), The exploration of multilingualism (pp.  11–26). Amsterdam-­Philadelphia: John Benjamins. Kulick, D., & Schieffelin, B. (2004). Language socialisation. In A. Duranti (Ed.), A companion to linguistic anthropology (pp. 349–368). Malden, MA: Blackwell. Li, W. (2011). Moment analysis and translanguaging space. Journal of Pragmatics, 43, 1222–1235. Li, W. (2018). Translanguaging as a practical theory of language. Applied Linguistics, 39, 9–30. Lincoln, P. C. (1975). Acknowledging dual-lingualism. Hawaï: Dept of Linguistics, University of Hawaï. Lorente, B. P.,  & Tupas, T. R. F. (2014). (Un)emancipatory hybridity: Selling ­English in an unequal world. In R. Rubdy & L. Alsagoff (Eds.), The globallocal interface and hybridity (pp. 66–82). Bristol: Multilingual Matters. Lucassen, J., & Lucassen, L. (2013). European migration history. In S. J. Gold & S. J. Nawyn (Eds.), Routledge international handbook of migration studies (pp. 52–63). London-New York: Routledge. Mackey, W. F. (2005). Multilingual cities. In U. Ammon, N. Dittmar, K. J. Mattheier & P. Trudgill (Eds.), Sociolinguistics (pp. 1304–1312). Berlin: Walter de Gruyter. MacSwan, J. (2017). A  multilingual perspective on translanguaging. American Educational Research Journal, 54, 167–201. Makoni, S., & Pennycook, A. (Eds.) (2007). Disinventing and reconstituting languages. In S. Makoni & A. Pennycook (Eds.), Disinventing and reconstituting languages (pp. 1–41). Clevedon: Multilingual Matters. May, S. (2005). Language rights. Journal of Sociolinguistics, 9, 319–347. McDermott, R., & Varenne, H. (1995). Culture as (dis)ability. Anthropology & Education Quarterly, 26, 324–348. Meeuwis, M., & Blommaert, J. (1998). A monolectal view of code-switching. In P. Auer (Ed.), Code-switching in conversation (pp. 76–98). London: Routledge. Møller, J. S. (2016). Learning to live with ‘languages’. Applied Linguistics Review, 7, 279–303. Moore, R. (2007). Sociology of knowledge and education. London: Continuum. Moore, R. (2015). From revolutionary monolingualism to reactionary multilingualism. Language & Communication, 44, 19–30. Nguyen, H. H. (2012). The multilinguaging of a Vietnamese American in South Philadelphia. Working Papers in Educational Linguistics, 27, 65–85. Otheguy, R., García, O.,  & Reid, W. (2015). Clarifying translanguaging and deconstructing named languages. Applied Linguistics Review, 6, 281–307. Otsuji, E.,  & Pennycook, A. (2010). Metrolingualism. Fixity, fluidity and language in flux. International Journal of Multilingualism, 7, 240–254. Pavlenko, A. (2018). Superdiversity and why it isn’t. In S. Breidbach, L. Kuster, B.  Schmenk (Eds.), Sloganizations in language education discourse. Bristol: Multilingual Matters (forthcoming). Pérez-Milans, M.,  & Soto, C. (2017). Light communities? Implications for research on diversity and activism. Working Papers in Urban Language & ­Literacies, 224.

26  Jürgen Jaspers and Lian Malai Madsen Pielke, R. A. (2007). The honest broker. New York: Cambridge University Press. Piller, I. (2016). Monolingual ways of seeing multilingualism. Journal of Multicultural Discourses, 11, 25–33. 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 (pp. 290–317). London: Routledge. Ritzau, U.,  & Madsen, L. M. (2016). Language learning, polylanguaging and speaker perspectives. Applied Linguistics Review, 7, 305–326. Said, E. (1978). Orientalism. London: Routledge & Kegan Paul. Salö, L. (2017). Sociolinguistics and epistemic reflexivity. Working Papers in Urban Language and Literacies, 206. Shohamy, E. (2006). Language policy. London: Routledge. Silverstein, M. (1979). Language structure and linguistic ideology. In P. R. Cline, W. F. Hanks & C. L. Hofbauer (Eds.), The elements (pp. 193–247). Chicago: Chicago Linguistic Society. Silverstein, M. (1985). Language and the culture of gender. In E. Mertz & R. J. Parmentier (Eds.), Semiotic mediation (pp.  219–259). New York: Academic Press. Strevens, P. (1982). The localized forms of English. In B. B. Kachru (Ed.), The other tongue (pp. 23–30). Urban IL: University of Illinois Press. Weisberg, M. (2007). Three kinds of idealization. The Journal of Philosophy, 104, 639–659. Wiese, H. (2018). Urban contact dialects. In S. Mufwene  & A. M. Escobar (Eds.), The Cambridge handbook of language contact (to appear). Cambridge: Cambridge University Press.

Part I

Negotiating Fixity

2 Recognizing Languages, Practicing Languaging Janus Spindler Møller

Introduction You do that thing: You say three words in Turkish, then a word in Danish just comes in. It’s like that. You just make a mix. Just like when you are hungry and eat shawarma, and you mix kebab and chicken. —(Mehmet, 7th grade, interview, May 2018, my translation from Danish)

This description of language use occurred during a conversation about language and daily life between four boys from Copenhagen and two fieldworkers. The boy Mehmet (with Turkish-Danish minority background) describes a way of speaking where language resources associated with more than one language are juxtaposed. This type of practice is well known to sociolinguists and has been documented in a range of different settings worldwide (e.g., Lehtonen 2016; Sultana et  al. 2015). On a basic level, Mehmet’s description involves the linguistic categories of ‘Turkish’ and ‘Danish’ and the agentive ‘mixing’. Beyond this, he underlines the ordinariness of this practice by comparing it to eating shawarma when hungry. Mehmet’s description thereby points to two interrelated aspects of his awareness about communication in daily life. The first is the almost trivial point that certain linguistic resources are thought to belong to certain languages. The second aspect concerns the knowledge that linguistic features associated with different languages may still be combined in interaction, no matter whether people think they belong together or not. On a more general level, Mehmet shows that he is not only aware of the fact that he is living in a languagized world; he also plays a role in languagizing it. For what enables Mehmet to characterize the practice as a mix is precisely the very idea of different languages. In the description of ‘mixing’, the recognizable entities of Danish and Turkish play central roles. In this way, fixity and fluidity become mutually constitutive in the description (cf. Otsuji & Pennycook 2014). Thus, Mehmet’s account does not imply a deconstruction of the idea of languages as coherent packages,

30  Janus Spindler Møller but rather a recognition of separable languages with boundaries between them, and an awareness of how linguistic resources associated with different languages are routinely juxtaposed in everyday interaction. As the example indicates, the idea of languages as bounded entities, and the knowledge that languages are frequently combined in daily interaction among peers, co-exist. In this chapter, I will address ways in which this knowledge concerning languages materializes in interaction. I  will look at instances where speakers demand certain ways of speaking in relation to languages from their interlocutors. I will analyze discussions between the participants of what (combinations of) languages to use in specific settings. Moreover, I  will address instances like the opening example by Mehmet where the participants describe practices and norms in relation to languages. This approach provides knowledge of the role the concept of languages play in the participants’ metapragmatic awareness (Silverstein 2001). The different types of metalinguistic accounts will demonstrate how the recognition and construction of ways of speaking comes to serve as a resource for identifying and dis-identifying, for policing each other’s behavior, for signaling group membership, and for local positioning. In other words, I will focus on how ways of speaking that in one way or the other relate to the phenomenon of languages become models of behavior. I will draw on two data sets which have in common that the use of linguistic resources that are associated with different languages has been observed as a mundane everyday activity among the participants ­(Jørgensen 2010; Madsen et al. 2016). In the end, I will argue that what may be described as monolingual and polylingual behavior simply characterizes two normative understandings of languaging. Neither is more ‘natural’ than the other. Depending on a number of factors, the participants opt for the norm that makes situational sense, and this goes for monolingual as well as for polylingual behavior. Before I  turn to the examples, I will briefly describe how I understand the concepts of languages and languaging and the interrelations between them. Languages From a sociolinguistic perspective, there can be little doubt that the notion of languages should be viewed as a construction rather than an objectively observable coherent entity. There is no apparent method to determine where one language begins and another ends. In addition, there is no general, objective way to determine when human beings ‘speak a language’. However, this does not prevent language users from treating languages as separable entities. On a large scale, speakers worldwide are socialized into viewing language through the lens of languages. A typical timetable for the senior classes in the Danish school system, for example, will consist of labels for lessons such as Danish, English, German,

Recognizing Languages, Practicing Languaging  31 French, Spanish, etc. Furthermore, on a societal level, languages are often habitually treated as emblematic components in the construction and maintenance of nations and play a central role in people’s sense of national belonging. Human social life is, for better or worse, languagized (Jaspers & Madsen, this volume). When children learn to communicate through language they gain a repertoire of linguistic forms of expression to communicate needs, thoughts, moods, and so on. During this process, they also gain knowledge of the fact that some of the linguistic resources they encounter are thought to belong together in ‘languages’ and in a range of other enregistered ways of speaking (Agha 2007), and that these are distinct from other languages or ways of speaking. Learning about such organization of language includes the fact that languages are given names (such as Danish, Norwegian, Turkish, etc.), and learning the norms for use associated with different ways of speaking is a part of acquiring communicative competence (cf. Hymes 1972). In this sense, languages can be described as sociocultural constructs (Blommaert & Rampton 2011; Heller 2007; Makoni  & Pennycook 2007). This sociocultural construction may be observed on the level of everyday interaction when behavior is ordered along the idea of languages as well as in century-long standardization processes of ‘national languages’ (Agha 2007; Gramling 2016). Yet, it is important to remember that speaking in one language at a time is no more natural than juxtaposing features associated with different languages in the same interaction, as sociolinguists have amply documented. Furthermore, it follows from viewing languages as sociocultural constructions that perceptions of what linguistic resources belong where may differ from speaker to speaker and from situation to situation. In this sense, ‘speaking a language’ can be considered as just one type of languaging based on specific conceptions of linguistic features’ belonging and expectations of their use. Languaging After a decade of studying the linguistic practices of speakers with Turkish-Danish backgrounds, Jens Normann Jørgensen and his research team decided to adopt the concept of languaging (Jørgensen 2010; Jørgensen et al. 2011; Madsen et al. 2016). Languaging refers to ‘the use of language by human beings, directed with an intention to other human beings’ (Jørgensen & Møller 2014: 67). With this formulation the team avoids presupposing that the default linguistic behavior is the use of resources that are associated with one language at a time (or, for that matter, any other labeled ways of speaking). This attempt is in line with a more general trend in contemporary sociolinguistics, which discards the view of languages as coherent packages. A driving force behind this development is that other sociolinguistic terminology, based on a view of

32  Janus Spindler Møller ways of speaking as separable, coherent entities (such as ‘bilingualism’, ‘native speaker’, etc.), has proven insufficient to describe the language use of contemporary urban youth (e.g. Jørgensen & Møller 2014; Madsen et  al. 2016; Møller  & Jørgensen 2013; Rampton 1995, 2006). It has proven more fruitful to view interaction from a languaging perspective, where language users ‘employ whatever linguistic features are at their disposal to achieve their communicative aims as best they can’ (Jørgensen et al. 2011). In accordance with Jørgensen (2010), I view languaging as an umbrella term for speaking, writing, signing or any other way in which speakers try to influence each other by means of linguistic resources. In other words, languaging is a social, and socially meaningful, activity. The description of languaging as intentional underlines the empirical interest in how linguistic resources are used to engage in social life. The claim is that people use language intentionally in interactions to grasp, influence, and shape the world, and the empirical task is to analyze how. A central assumption is that languaging is always designed in one way or another to influence the interlocutors just as participants in interaction are more or less constantly occupied with making sense of each other’s interactional contributions (cf. Rampton 2006). Another assumption is that languaging is reflexive in the sense that language users over time construct knowledge about how combinations of linguistic resources are reacted to in certain types of situations. This includes strategic as well as habitual use of recognizable ways of speaking (Agha 2007). The ways in which such knowledge interrelates with what goes on in interaction are of course complex and the assumption is not that it is always possible analytically to establish the intention ‘behind’ language use or that communicative aims are always clear and unambiguous for language users and their interlocutors. However, instances of meta-communicative activity may provide insights into how participants grasp and organize different types of languaging and the data examples in this chapter are explicit accounts of such activity. The Role of Languages when Studying Languaging Viewing languages as sociocultural constructions is not new. In fact, as early as 1589, George Puttenham, an English literary critic, had already highlighted the processes involved when ways of speaking become identified as languages: ‘After a speech is fully fashioned to the common understanding, and accepted by consent of a whole country and nation, it is called a language’ (cited in Gramling 2016). Here languages are described as depending on ‘common understanding’ and ‘accept[ance] by consent’. In this way the quote is a reminder that clusters of linguistic features become ‘languages’ because they get recognized as such by a population, and, perhaps more importantly, that the idea of the division of language

Recognizing Languages, Practicing Languaging  33 into languages has had more than 400 years to gain strength (and has done so in close interplay with the ideological constructions of ‘nations’). In present times, the concept of languages is for many people (not least in Western societies) perceived as an unquestionable and naturally given way of organizing language. This leaves contemporary sociolinguistics with a paradoxical challenge. Sociolinguistic work has powerfully demonstrated that what some consider naturally given entities of language are in fact ideological constructions. This destabilization of linguistic fixity has been a driving force in the development of new terms such as translanguaging (García & Li 2014), metrolingualism (Pennycook & Otsuji 2015) and polylingual languaging (Jørgensen et  al. 2011, also discussed below), terms which to a large degree aim to problematize conventionalized understandings of languages. However, this deconstruction of languages is at odds with the widespread idea of languages as recognizable, countable, and bound up with personal histories and identities. While sociolinguists have come to understand the idea of languages as a social construct, and language use as a dynamic, fluid process, they must equally recognize that the participants in their studies to a large degree organize language in languages. The question is then what terminology can be used to describe the types of practices described by Mehmet in the opening example where linguistic resources are organized according to an idea of separable languages. While I  acknowledge the insights gained from studies using a lens of code-switching (e.g., Gumperz 1982) or code-alternation (e.g., Auer 1995), I prefer not to use these terms because the acts of ‘switching’ or ‘alternating’ render monolingual behavior as the unmarked way of speaking. No matter how influential the idea of languages as entities has become, this does not qualify so-called monolingual behavior to automatically constitute the point of departure in sociolinguistic studies of languaging. As stated by Otsuji & Pennycook: It is not so much that languages, cultures, the local, or the global exist in isolation only to become hybrid when they come into contact under particular circumstances, but rather that their prior separation was always a strange artifact of particular ways of thinking. (Otsuji & Pennycook 2014: 84) This means that in order to understand diversity in language use, sociolinguistic studies must address the ways of thinking that lead to viewing certain forms as hybrid under certain conditions. In order to incorporate these reflections of the ontological status of languages, my theoretical point of departure is to distinguish between different types of languaging. Without giving any of them a special status compared to the other, I will distinguish between monolingual and polylingual languaging as outlined in Jørgensen et  al. (2011). Monolingual

34  Janus Spindler Møller languaging may be defined as using linguistic resources associated with one language at a time. Polylingual languaging may be defined as the use of linguistic resources which participants in a communicative encounter perceive as belonging to different languages. Note here that it is not straightforward to determine to what extent participants perceive linguistic features as belonging to different languages. I  will address this challenge by basing myself in this chapter on exchanges between the participants where they explicitly address the situated use of one or more languages. The guiding principle behind this is that the distinction between monolingual and polylingual languaging is a relevant analytical tool only as far as the distinction can be documented as relevant to the participants in the study, and this relevance needs to be argued for empirically. From this perspective, it matters less whether the empirical documentation occurs during mutual interaction between the participants, during retrospective interviews with fieldworkers, or in other data sources, as long as the limitations of the different data types are taken into consideration in the analysis. Another important thing to remember when applying the notions of mono- and polylingual languaging analytically is that languages are sociocultural constructions. Recognition of linguistic features as belonging to one or the other language presupposes ideological processes where groups of people recognize clusters of language resources as particular ways of speaking. Agha’s (2005, 2007) notion of enregisterment theorizes such processes. Enregisterment denotes the social processes that shape a way of speaking into a recognizable register. The formation of registers (i.e., recognizable ways of speaking) links the use of specific (combinations of) linguistic resources to interpersonal conduct, norms of situational use, metapragmatic classifications, and evaluations of speech behavior. As a result, using a register has the potential to evoke a schema of behavior, or a behavior model—and this is evoked regardless of whether the enregistered way of speaking is produced strategically or habitually. Situated language use is thereby understood as a social, collaborative activity pointing to roles, personas, and expectations. One advantage of the notion of enregisterment is that it accounts for the sociocultural construction of ‘national languages’ as well as other ways of speaking, including conceptualizations of polylingual practices. Previously I mentioned how the notion of polylingual languaging has much in common with other recently developed terms that aim to nuance widespread understandings of linguistic diversity (see also Jaspers  & Madsen, this volume). I will, however, mention one important way where the idea of polylingual languaging in my view distinguishes itself from the notion of translanguaging. In a description of translanguaging, Li Wei provides the following argument for the term: What I like about William’s and Baker’s idea of Translanguaging is that it is not conceived as an object or a linguistic structural phenomenon

Recognizing Languages, Practicing Languaging  35 to describe and analyse but a practice and a process—[. . .] a process of knowledge construction that goes beyond language(s). (Li 2018: 15; emphasis in original) This description allows me to state that the notion of polylingual languaging does not go ‘beyond language(s)’. In fact, it does quite the opposite. It helps to describe exactly the juxtaposed use of linguistic resources associated with different languages (or other enregistered ways of speaking). I am not claiming this is an easy task. The use of features associated with different languages may be enregistered as a new way of speaking in itself, and it may prove difficult to establish from a situated participant perspective what resources are associated with what languages. However, the notion of polylingual languaging aims at describing a certain practice that distinguishes it from another (namely that of monolingual languaging). All the examples I address below have in common that the participants explicitly address the use of languages. As mentioned above, this may involve descriptions of practices as well as expectations to situated language use. When it comes to expectations, it turns out that the practices of monolingual and polylingual languaging are also relevant on the level of norms (cf. Jørgensen 2010). In some of the examples, we shall observe how the participants specifically or indirectly refer to monolingual and polylingual norms for behavior. In the analysis of these instances, Agha’s (2007) schematic overview of three general types of normativity in relation to language is useful: (a) A norm of behavior: An externally observable pattern of behavior, e.g., a statistical norm or frequency distribution on some order of behavior (b) A normalized model of behavior: A reflexive model of behavior, recognized as ‘normal’ or ‘typical’ by (at least some) actors, i.e., is a norm for them (c) A normative standard: A  normative model, linked to standards whose breach results in sanctions; a norm codified as a standard. (Agha 2007: 126; emphasis in original) In Agha’s gradation, type (a) denotes behavior that is statistically observable and distinct from other types of behavior, but which is not necessarily recognized or reflected upon by its speakers. Type (b) introduces reflexivity concerning the norm among the users. A  group of speakers may recognize a certain behavior as their typical behavior, which also enables processes of identification and dis-identification. With type (c), an element of institutionalization is introduced in which recognizable ways of speaking are presented as the only ‘right’ or ‘standard’ way of speaking. In the data analysis below, we shall observe how the participants organize observable patterns of monolingual and polylingual behavior into

36  Janus Spindler Møller ‘normalized’ as well as ‘normative’ models of behavior, but first I  will introduce the two studies providing data to this chapter in more detail.

Data and Earlier Project Findings The two projects have in common that they study language use among school children over time, and that they to a high degree involve participants with linguistic minority backgrounds. The so-called Køge project was initiated in the late 1980s, and the last data so far were collected in 2007 (see Jørgensen 2010 and Møller 2009 for overviews). The original research interests were educational as well as sociolinguistic, with a particular focus on multilingualism—many of the participants were children of immigrants from Turkey. Data were collected from the same pupils in two parallel classes throughout their time in the public school system (1989–1998) and again in 2006–2007. During the ten years of school, a team of project workers visited the classes yearly and collected a range of data. The data included interviews conducted in Danish as well as Turkish and group conversations without the presence of adults. In these group conversations, the participants were given a task (for example, building a sculpture in Lego, making a cartoon on cardboard, etc.) and were recorded while doing it. One of the groups recorded each year consisted exclusively of speakers with a Turkish-Danish minority background. In 2006–2007 a new round of data were collected among the same participants. Part of the data this time included interviews about language use and recordings of conversations during restaurant visits (see also Schøning & Møller 2009). The yearly group conversations with the group of exclusively TurkishDanish minority background participants turned out to be a particularly interesting set of data in relation to studies of polylingual languaging. All in all, this part of the data involves 43 recorded conversations (Jørgensen 2010: 315) lasting about half an hour each. When these participants started their school careers in first grade in 1989, their mutual conversations were predominantly carried out in linguistic resources associated with Turkish. During the years in school, the project documents how all the participants become highly proficient in Danish. Alongside this development they also developed the ability to use linguistic resources associated with Danish and Turkish (and for that matter a number of other ways of speaking) juxtaposed in a way that in fact corresponded very well to what Mehmet describes in the opening example of this chapter. From the fifth grade onwards, most of the conversations between participants with a Turkish-Danish background were characterized by frequent juxtaposition of language resources associated with Danish and Turkish and other languages (see Jørgensen 2010; Møller 2016). This made it problematic (if not meaningless) to claim that the conversations were Turkish-dominant or Danish-dominant, and more importantly it

Recognizing Languages, Practicing Languaging  37 became problematic to claim that the participants were ‘switching’ or ‘alternating’. As discussed above, these terms precondition a state where people do not ‘switch’ or ‘alternate’ and thereby nominate the activity of ‘speaking a language’ (or any other established variety) as the analytical point of departure. What we found at this point of the data collection challenged this ontological understanding. When situations acquired ‘monolingual’ behavior, the participants seemingly adjusted to this with ease (we have plenty of evidence of this ability), but when interacting with peers with similar linguistic backgrounds they just as habitually juxtaposed linguistic resources associated with different languages. This type of observations led to the development of the Copenhagen version of the theory of languaging (as a replacement for ‘speaking a language’, ‘bilingual interaction’, etc.—see, e.g., Jørgensen & Møller 2014). Worth noting is that none of the 43 group conversations with the exclusively Turkish-Danish background participants qualifies as strictly monolingual in the sense that each conversation would contain linguistic resources associated with more than one language. This observation further underlines that for the participants, ‘monolingual conversations’, whether in Danish or Turkish, are not the ‘natural’ or the ‘default’ way of speaking. For this group of participants, interactions, which include the use of resources associated with more than one language, are not an exceptional but a quite ordinary activity (see Jørgensen 2010; Møller 2016). The other project providing data to this chapter is called the Everyday Languaging project (Madsen et al. 2016). This project also involves school classes, and the part of the project I  will draw on here follows pupils in two parallel forms in two years from the middle of the seventh grade until the end of the ninth. Where the Køge project mainly involved participants whose parents migrated from Turkey, the participants in this project represented a range of different linguistic backgrounds. In the Everyday Languaging project, ethnographic observation played a central role. A  team of fieldworkers followed the participants in the project period during school and leisure time, on social media, and with their families (Ag 2018). The main research interest was to get an in-depth account of how the participants used and organized language across a number of social activities in their daily life. Data include observations, recordings, social media activities, interviews, and written essays about language use. A central theoretical interest was to investigate processes of the enregisterment of language resources, including how and with what effect these resources were employed in daily interaction. In the project’s early stages, the team came across a linguistic phenomenon the participants themselves labeled as ‘street language’, ‘ghetto’, ‘slang language’, or ‘perker language’ (perker being a term that is associated with immigrants, particularly from Middle-Eastern countries). In theoretical terms, this way of speaking can be characterized as

38  Janus Spindler Møller a contemporary urban vernacular (Rampton 2011: 19–20; cf. HyttelSørensen 2017), in terms of being a conceptualized register (in Agha’s understanding) that is associated with contemporary urban environments and enregistering activity that typifies it and that sets it apart from, for example, standard varieties, learner varieties, and so on. What makes this way of speaking relevant for the focus in this chapter is an understanding among the participants that part of what characterizes it is the juxtaposition of language resources associated with a range of different languages. The participants in our project were aware of this way of speaking to a degree where they could imitate it, classify typical users and situations of use, and furthermore describe its place in a hierarchy of ways of speaking in their everyday. For example, they conceptualized the opposite of speaking street language as speaking ‘integrated’ (associated with polite academic behavior) (Madsen 2013; Møller & Jørgensen 2013). An important difference between the Køge project and the Everyday Languaging project lies in the methods of data collection. Where the Køge data constitute yearly pinpricks over a long time span, the part of the Everyday Languaging project I  deal with here consisted of intense ethnographic observation over a shorter time. Furthermore, the original focus in the Køge project was not on enregisterment and metalinguistic awareness, but as we shall see the project provided data for analyzing this anyway. On the other hand, the focus on enregisterment was very central in the Everyday Languaging project. One should thus be careful to compare the studies in terms of which group is the more reflexive one about language use. The participants in the Everyday Languaging project had many more opportunities to reflect about language use, and not surprisingly, their descriptions come out more detailed at an earlier life stage than what we find in the Køge data. However, the two projects have an important aspect in common: in both sets of data, we can observe how the participants recognize certain linguistic behaviors, how they describe them, and how they organize them in systems of normativity and social belonging.

Data Analysis In the following presentation of examples, the notion of a language emerges in one way or the other. Sometimes the participants reflect on languaging practices, sometimes they discuss or impose norms on these practices. The overall aim in this section is to investigate how the participants employ the idea of languages in social activities and how this interplays with models of behavior. I start with four examples from the Køge material and proceed with three examples from the Everyday Languaging project.

Recognizing Languages, Practicing Languaging  39 ‘Now We Speak Danish’ The first example comes from a group conversation between two girls and two boys in Køge, recorded during the sixth grade (where the participants are around the age of 12). Prior to the extract, a fieldworker had instructed the participants that this year they were going to build a Lego sculpture and that the whole group was to cooperate in doing so. Then the fieldworker left the room. The transcribed exchange occurs after a few minutes of working on the task. As suggested above, the group conversations between participants with TurkishDanish backgrounds are at this age level characterized by the frequent juxtaposition of language resources associated with both Danish and Turkish. This is also the case in the example, but analytically I  will focus on a metalinguistic activity concerning the language choice that takes place:

Example 1. Køge. Group Conversation. Sixth Grade. Participants: Canan, Hüseyin, Merva, Achmet. Task: Make a Lego Sculpture   1. Hüseyin: [>] [xxx cannot close]   2. Merva: [] vi ikke hvad [we speak Danish don’t you think] 5. Murat: [] [okay shall we leave now, then] 5. Ceyhun: [] [it is like mixed] [