Z Erziehungswiss (2010) 13:549–572

DOI 10.1007/s11618-010-0159-y

Schwerpunkt

Towards a sociolinguistics of superdiversity

Angela Creese · Adrian Blackledge

Abstract:  The notion of ‘superdiversity’ has engaged scholars beyond the field of sociolinguis-
tics. In this paper we propose to shift the gaze to the linguistic, focusing on the ways in which the
new diversity becomes the site of negotiations over linguistic resources, and to widen the scope of
debate. The ways in which people negotiate access to resources in increasingly diverse societies
are changing. Looking at these phenomena through a sociolinguistic lens is key to a developed
understanding of superdiverse societies. García proposes the term ‘translanguaging’ to refer to
the multiple discursive practices in which multilingual speakers engage in order to make sense of
their worlds. Translanguaging goes beyond code-switching, but incorporates it. García points out
that multilinguals translanguage to include and facilitate communication with others, but also to
construct deeper understandings. Translanguaging includes but extends what others have called
language use and language contact among multilinguals. Rather than focusing on the language
itself, translanguaging makes it apparent that there are no clear-cut boundaries between the lan-
guages of bilinguals. This paper draws on sociolinguistic ethnographic research projects which
investigate the linguistic practices of children and young people in and around complementary
(community-language) schools, to argue that multilingual young people in English cities access
a wide range of semiotic resources in ways which are not bounded as ‘languages’. In developing
a sociolinguistics of superdiversity we should look closely at practices of translanguaging, and
consider the histories, geographies, and discourses which shape them.

Keywords:  Linguistic ethnography · Orders of indexicality · Sociolinguistics · Superdiversity ·

Translanguaging

Zu einer Soziolinguistik von Superdiversität

Zusammenfassung:  Der Begriff der Superdiversität beschäftigt auch Wissenschaftler außerhalb

des Feldes der Soziolinguistik. In diesem Beitrag schlagen wir vor, den Blick auf die Frage
der Sprache zu richten und sich auf die Wege zu konzentrieren, in denen die neue Diversität
zum Raum für Verhandlungen über sprachliche Ressourcen wird. Die Art und Weise, mit der
sich Menschen in zunehmend diversifizierten Gesellschaften Zugang zu Ressourcen verschaffen,

Published online: 15.12.2010

© VS Verlag für Sozialwissenschaften 2010
Prof. A. Creese, Ph.D. () · Prof. A. Blackledge, Ph.D. ()
School of Education, University of Birmingham,
B15 2TT, Birmingham, UK
e-mail: a.creese@bham.ac.uk
Prof. A. Blackledge, Ph.D.
e-mail: a.j.blackledge@bham.ac.uk
550 A. Creese and A. Blackledge

ändert sich. Ein Zugang zum tieferen Verständnis von superdiversen Gesellschaften eröffnet sich,
wenn man solche Phänomene aus soziolinguistischer Perspektive betrachtet. Garcia schlägt den
Begriff „Translanguaging“ vor, der auf die multiplen diskursiven Praxen verweist, in denen mehr-
sprachige Sprecher den Sinn ihrer Welten erfassen. Translanguaging (für das es keinen deutschen
Ausdruck gibt; gemeint ist: in Mehrsprachigkeit – sprachübergreifend – agieren) geht über Code
Switching im Sinne des Wechselns zwischen Sprachen hinaus, schließt dies aber mit ein. Garcia
weist darauf hin, dass Mehrsprachige in dieser Weise agieren, um Kommunikation mit anderen zu
erleichtern – aber auch um tieferes Verständnis zu erreichen. Translanguaging schließt ein, und
geht gleichzeitig über das hinaus, was als Sprachgebrauch und Sprachkontakt zwischen Mehr-
sprachigen bezeichnet wird. Statt auf Sprache selbst zu schauen, soll der Begriff verdeutlichen,
dass es keine klaren Abgrenzungen zwischen den Sprachen von zwei- oder mehrsprachigen Men-
schen gibt. – Unser Beitrag berichtet aus soziolinguistisch-ethnographischen Forschungsprojek-
ten, die die sprachlichen Praxen von Kindern und Jugendlichen innerhalb und im Umfeld von sog.
community-language schools untersuchen – Schulen also, die neben dem regulären Schulsystem
existieren, die oft von privaten Gemeinschaften getragen werden und in denen es um den Un-
terricht von Minderheitensprachen (community languages) geht. Wir legen dar, dass sich mehr-
sprachige Jugendliche in englischen Städten eine große Bandbreite von Ausdrucksmöglichkeiten
aneignen, die nicht in ‚einer Sprache‘ im traditionellen linguistischen Verständnis aufgehen. Eine
Soziolinguistik der Superdiversität sollte die Praxen des Translanguaging einbeziehen und die
Geschichten, Geographien und Diskurse berücksichtigen, die sie gestalten.

Schlüsselwörter:  Linguistische Ethnographie · Ordnungen der Indexikalität · Soziolinguisitik ·

Superdiversität · Mehrsprachige Praxis

1  Introduction

Western societies have become more diverse in recent times, as numbers and territorial
origins of migrants have expanded. This phenomenon has resulted in new demographic
patterns of migration and post-migration, termed ‘superdiversity’, and characterised by
“a dynamic interplay of variables among an increased number of new, small and scat-
tered, multiple origin, transnationally connected, socio-economically differentiated
and legally stratified immigrants who have arrived over the last decade” (Vertovec
2007a, p. 1024). The notion of ‘superdiversity’ has tended to attract the attention of
scholars in migration studies, ethnic and racial studies, urban studies, and sociology.
In this paper we propose to shift the gaze to the linguistic, focusing on the ways in
which the new diversity becomes the site of negotiations over linguistic resources.
We also propose to widen the scope of debate. The ways in which people negotiate
access to resources in increasingly diverse societies are changing in response to other
developments, and we argue that the new diversity is not limited to ‘new’ migrants who
arrived in the last decade, but includes changing practices and norms in established
migrant (and non-migrant) groups, as daughters and sons, grand-daughters and grand-
sons, great-grand-daughters and great-grandsons of immigrants (and non-migrants)
negotiate their place in their changing world. In this paper we propose that looking at
these phenomena through a sociolinguistic lens is key to a developed understanding of
superdiverse societies.
Towards a sociolinguistics of superdiversity 551

2  Superdiversity

Over the past decade or so, a range of immigration patterns and variables have altered
the composition, distribution and statuses of immigrant communities in Britain, Europe,
and elsewhere. The term ‘superdiversity’ has been coined to refer to the meshing and
interweaving of diversities, in which not only ‘ethnicity’, but other variables intersect and
influence the highly differential composition, social location and trajectories of various
immigrant groups in the twenty-first century (Vertovec 2007b). The new diversity incor-
porates the phenomenon of ‘transnationalism’, or, more specifically, ‘migrant transnation-
alism’, which refers to a range of practices and institutions linking migrants, people and
organisations in their homelands or elsewhere in a diaspora (Vertovec 2009). In recent
years transnationalism has become one of the key ways of understanding contemporary
migrant practices (Vertovec 2007b). The social relations of migrants and refugees are not
confined within nations but are ‘transnational’, hence a diaspora is a specific form of trans-
national community (Cox and Connell 2003, p. 330). Glick Schiller et. al. (1992) argued
that immigrants are increasingly becoming transnational, creating a single field for social
action that encompasses both the home and host societies. In recent years, the extent
of transnational engagement has intensified. The degrees to and ways in which today’s
migrants maintain identities, activities and connections linking them with communities
outside Britain are unprecedented (Vertovec 2007a). Levitt and Glick Schiller (2004)
argue that assimilation and enduring transnational ties are neither incompatible nor binary
opposites. Vertovec (2006) points to a ‘bi-focality’ of outlooks, a kind of “habitus of dual
orientation” (Vertovec 2009, p. 68) underpinning migrant lives lived here-and-there, and
suggests that such dual orientations have considerable influence on transnational family
life, and may continue to affect identities among subsequent post-migration generations.
Anderson (1998, p. 74) refers to “long-distance nationalism” to suggest that a strong alle-
giance binds members of an ethnic diaspora to their homeland. Glick Schiller and Fouron
(2002, p. 4) define long-distance nationalism as “a claim to membership in a political
community that stretches beyond the territorial borders of a homeland”. Vertovec (2009)
refers to ‘dimensions of transnational competence’, and suggests that within many fami-
lies whose lives are stretched across migrant sending and receiving contexts, transnational
patterns of everyday activity have become normative. Such patterns of everyday activity
include the purchase and use of cheap international phone cards, money transfer bureaux,
internet cafés, and so on (Blommaert 2010). Bolognani (2007) investigated the reproduc-
tion of ‘homeland’ attachments among second- and third-generation British Pakistanis in
UK. She found that young British Pakistanis retain a strong emotional link to the ‘home-
land’, and links are kept alive by trips to Pakistan, and by intercontinental marriage. At
the same time she found the notion of national belonging to the homeland to be “a com-
plex resource of negotiation of what the younger generation look for as effective, equal
and complete citizenship” (Bolognani 2007, p. 74). Vertovec (2009) acknowledges that
there appears to be a considerable variety of patterns and kinds of transnationalism among
different groups of ‘second-generation’ youth. He suggests these patterns are constituted
in an interplay of parents’ transnational habitus, an array of local conditioning factors,
and second generation youth’s own “hybrid or multicultural habitus” (2009, p. 76). This
may be a more linear account of the inter-generational re-orientation of habitus than we
552 A. Creese and A. Blackledge

would propose ourselves, but it serves as a useful starting-point. At the very least, in
attempting to make sense of intergenerational post-migration experience, we should be
sensitised to situated dimensions of time and space, to stratified social systems, and to dif-
ferent patterns and nuanced practices of negotiation, as the sons and daughters, grandsons
and grand-daughters, and great-grandsons and great-grand-daughters of migrants shape
and re-shape their worlds.
We have said that superdiversity is characterised by a dynamic interplay of variables
in the migrant and post-migrant experience. Certainly it is no longer (if ever it was) suf-
ficient to view diversity simply in terms of ‘ethnicity’ or country of origin. Other factors
which come into play include, inter alia, differential immigration statuses, gender, age,
economic mobility, social class/caste, locality, and sexuality. Blommaert (2010) points
out that in increasingly diverse neighbourhoods extreme linguistic diversity may gener-
ate complex linguistic repertoires in which several (fragments of) ‘migrant’ languages
and lingua francas are combined. We suggest that the notion of superdiversity further
includes not only migrants over the last decade, but longer established and settled migrant
communities over several age cohorts, as their experiences shift within and across gen-
erations. Of course there have been periods in history when there has been a sudden and
rapid increase in the movement of peoples across territories. So this phenomenon is not
unique. But superdiversity, however we characterise it, serves to remind us to set aside
our assumptions about the national, regional, ethnic, cultural or linguistic characteristics
of particular ‘groups’ (Blommaert 2010). Vertovec (2007a) argues for more and better
qualitative studies of superdiversity, as neither social scientists nor policy-makers have
access to accounts of what meaningful interactions look like in superdiverse settings.
In such accounts of everyday practice the intermeshing and interweaving of the above
variables become clear. This call for detailed research into local, situated practice points
to that which is so far notable only by its absence from our discussion: a focus on the
sociolinguistics of superdiversity.

3  Languaging in superdiversity

Heller (2007, p. 1) argues for an approach to researching multilingualism which moves
away from a highly ideologized view of coexisting linguistic systems, to a more critical
approach which situates language practices in social and political contexts, and which
“privileges language as social practice, speakers as social actors and boundaries as prod-
ucts of social action”. Gal (2006) points out that in Europe a new elite of multilingual
speakers (of, e.g., French, German, and English) sustains a breadth of linguistic reper-
toires which transcends national boundaries. For such groups ethnolinguistic identity may
be only an occasional issue. For multilingual speakers of languages with lower status,
however, language issues may still be salient as people attempt to negotiate identities,
often from relatively powerless positions. Language ideologies are neither simple nor
monolithic, however. Notwithstanding the argument that minority language speakers are
subject to the symbolic violence of the dominant language ideology, some speakers who
(or whose families) may traditionally have been associated with minority ‘ethnic’ lan-
guages are using language and languages in new ways (Rampton 1995, 1999). While
Towards a sociolinguistics of superdiversity 553

some speakers are either unable to negotiate their identities from inextricably powerless
positions, and others in powerful positions have no need to do so, some speakers in mod-
ern nation-states are using their linguistic skills to negotiate new subject positions (Black-
ledge and Pavlenko 2001; Pavlenko and Blackledge 2004). In what Gal (2006, p. 27)
describes as ‘self-conscious, anti-standardizing moves’, such negotiations may include
linguistic practices which reframe previous standard varieties, incorporating, inter alia,
urban popular cultural forms, minority linguistic forms, hybridities and inventions. Here
language practices associated with immigrant groups no longer represent backward-look-
ing traditions, but may be linked to global youth culture and urban sophistication. Lan-
guages and language practices are not necessarily equated to national identity (but may
be so), and are not necessarily dominated by the standardized variety. Despite powerful
ideologies of homogeneity, populations in many countries—especially countries with
a history of recent immigration—continue to be heterogeneous in their practices. May
(2005, p. 337) proposes that linguistic identities need not be oppositional, and asks ‘what
exactly is wrong with linguistic complementarity?’. May calls for further ethnographic
studies which articulate and exemplify broad linguistic principles of language ideological
research in complex multilingual contexts. Heller and Duchêne (2007, p. 11) argue that
rather than accepting ideological positions in which there is competition over languages,
‘perhaps we should be asking instead who benefits and who loses from understanding
languages the way we do, what is at stake for whom, and how and why language serves
as a terrain for competition’.
Heller (2007) proposes four sets of concepts in the critical analysis of languages in
society. First, she argues that rather than treating notions of ‘community’, ‘identity’,
and ‘language’ as though they were natural phenomena, they should be understood as
social constructs. Specific or single categorisations therefore cannot be attached to an
individual based on their ‘ethnicity’, or ‘language’. Second, Heller refers to the work
of Giddens (1984) to consider language as a set of resources which are socially distrib-
uted, but not necessarily evenly. The third set of concepts holds that this uneven distri-
bution of resources is the product of political and economic processes, enabling us to
ask questions about what linguistic resources are assigned what value, and with what
consequences (Gumperz 1982). The final set of concepts considers the discourses which
inscribe value (or its lack) to particular linguistic forms and practices. In summary, Heller
views language(s) as:
sets of resources called into play by social actors, under social and historical con-
ditions which both constrain and make possible the social reproduction of exist-
ing conventions and relations, as well as the production of new ones (Heller 2007,
p. 15).
Linguistic practice can only be understood in relation to histories, power, and social
organisation. Conversely, structural analysis must include accounts of actual linguistic
practices, which at times may differ from those we might expect. Language practices are
ways of communicating that not only link members of a speech community in the present
to a (mythical or real) past and an imagined future, but also signal ideological positions
and boundary markers of group identity for inclusion and exclusion, and for constructing
and effacing another (García 2010, p. 523).
554 A. Creese and A. Blackledge

We agree with Makoni and Pennycook (2007) that the notion of languages as separate,
discrete entities, and ‘countable institutions’ (2007, p. 2) is a social construct. Makoni and
Pennycook argue for a critical historical account which demonstrates that, through the
process of classification and naming, languages were ‘invented’ (2007, p. 1). They add
that, in direct relation with the invention of languages, ‘an ideology of languages as sepa-
rate and enumerable categories was also created’ (2007, p. 2). Makoni and Pennycook
point in particular to the naming of languages such as ‘Bengali’ and ‘Assamese’ as the
construction of ‘new objects’ (2007, p. 10). Thus languages cannot be viewed as discrete,
bounded, impermeable, autonomous systems. Makoni and Pennycook propose that ‘local
knowledge’ is crucial to our understanding of language:
We are arguing for an understanding of the relationships between what people
believe about their language (or other people’s languages), the situated forms of
talk they deploy, and the material effects—social, economic, environmental—of
such views and use (Makoni and Pennycook 2007, p. 22)
This interrelationship between what people believe about language and languages, and
the way they access and make use of linguistic resources, provides a focus to our analysis
in this paper.
If languages are invented, and languages and identities are socially constructed, we
nevertheless need to account for the fact that at least some language users, at least some
of the time, hold passionate beliefs about the importance and significance of a particular
language to their sense of ‘identity’ (Blackledge and Creese 2010). It is now well estab-
lished in contemporary sociolinguistics (Harris 2006; Rampton 2006) that one ‘language’
does not straightforwardly index one subject position, and that speakers use linguistic
resources in complex ways to perform a range of subject positions, sometimes simultane-
ously. However, whilst accepting this, May (2001; 2005, p. 330) argues that ‘historically
associated languages continue often to hold considerable purchase for members of par-
ticular cultural or ethnic groups in their identity claims’. Rampton (1995) argues that we
need a better understanding of the linguistic formation and social identity of the bilingual
learner, and suggests a framework for viewing language education as a social activity in
which efforts are made to manage continuity, change and relationships between social
groups. He suggests that the terms expertise (proficiency in a language), affiliation (attach-
ment or identification felt to a language), and inheritance (ways in which individuals can
be born into a language tradition) are useful tools for describing the bilingual speaker
and bilingualism. Whilst it is certainly an oversimplification to treat certain languages
as ‘symbols’ or ‘carriers’ of ‘identity’, we are obliged to take account of what people
believe about their languages, to listen to how they make use of their available linguistic
resources, and to consider the effects of their language use—even where we believe these
‘languages’ to be inventions. Makoni and Pennycook (2007) propose the ‘disinvention’ of
languages, and a reinvention which acknowledges heterogeneity, arguing that languages
are discursive constructions which perpetuate social inequities. Makoni and Mashiri
(2007) suggest that rather than developing language policies which attempt at hermeti-
cally sealing languages, we should be describing the use of vernaculars which leak into
one another to understand the social realities of their users. They argue that it is neces-
sary to overcome existing ideas about language if we are to imagine alternative ways of
Towards a sociolinguistics of superdiversity 555

conceptualizing the role and status of individuals in the world, and that ‘a world in which
plurality is preferred over singularity requires rethinking concepts founded on notions of
uniformity over those predicated on diversity’ (2007, p. 27). Stressing the user rather than
the code or language is central to this argument. Rampton’s (1995; 2006) work on het-
eroglossia in urban contexts among adolescents shows how individuals may appropriate
and invent linguistic practices to negotiate their identities. Rampton (2006, p. 27) refers
to ‘stylisation’ as a particular kind of performance in which the speakers produce ‘an
artistic image of another’s language’ (Bakhtin 1981, p. 362). Hess-Lüttich (1978) speaks
of bilingualism as a ‘style resource’. Viewing language use as a styling process places the
social actor at the centre of analysis. Rampton argues (1998, p. 8):
instead of being the product of forces that actors neither control nor comprehend,
human reality is extensively reproduced and created anew in the socially and his-
torically specific activities of everyday life.
Such a social constructivist approach works with the agency of a situated speaker and
explains language use as contextually embedded. Viewing bilingualism as a ‘style
resource’ necessitates moving away from an emphasis on languages and their different
codes towards an account which describes the individual engaged in meaning making
and identity work. Whereas a typical linguistic analysis of code-switching might focus on
categorizing languages and describing the function these languages perform, a social con-
structivist approach problematizes the constructs of ‘language’ and ‘community’, resist-
ing classifications of languages or communities into bounded systems. The alternative to
a code-switching structuralist account seeks to show variety through heteroglossia and
stylisation.
García (2009, p. 45) proposes the term ‘translanguaging’ to refer to the multiple dis-
cursive practices in which multilingual speakers engage in order to make sense of their
worlds. Translanguaging goes beyond code-switching, but incorporates it. García points
out that multilinguals translanguage to include and facilitate communication with oth-
ers, but also to construct deeper understandings. Translanguaging includes but extends
what others have called language use and language contact among multilinguals. García
argues that rather than focusing on the language itself, translanguaging makes it apparent
that there are no clear-cut boundaries between the languages of bilinguals. Heller (2007)
suggests that if we tend to understand linguistic resources as whole, bounded systems
which we call ‘languages’, it is because nations and states have found it necessary to pro-
duce powerful discourses which constitute language ideologies in the process of national
belonging. Makoni and Pennycook (2007, p. 36) argue for language policy in education
which focuses on ‘translingual language practices rather than language entities’. García
(2007, p. xiii) points out that if language is an invention, we must observe closely how
people use language, and base pedagogical practice on that use, and not on what the
school system says are valuable practices. While largely accepting these arguments, we
also agree with Heller (2007, p. 342) that ‘bilingualism’ or ‘multilingualism’ allows us a
purchase on what count as relevant categories, and points us towards to a focus on how
boundaries happen, how people and practices are included and excluded, and what hap-
pens to them as a result (Blackledge and Creese 2010; Creese and Blackledge 2010). It
may be that some of the resources used in translanguaging repertoires are undeveloped or
556 A. Creese and A. Blackledge

‘unfinished’ as (what we traditionally call) ‘languages’ 2010). This is certainly an issue

where a speaker is required (institutionally or legally) to demonstrate proficiency in a
particular language, for example in the asylum process (Maryns 2006), or in citizenship
application (Blackledge 2009a, b, c). However, there is no evidence that translanguaging
practice is oppositional to the development of proficiency in standard or non-standard
varieties of individual ‘languages’. In practice translanguaging repertoires unproblemati-
cally incorporate linguistic items from a range of sources which do not require singular
proficiency.
Jørgensen (2010) points to the fluidity of late modern society, in which identities are
not necessarily imposed from above, but may be negotiable within certain social settings.
Jørgensen proposes that language users create, construct, and negotiate identities on the
basis of a range of resources which can be associated with meaning. To the extent that
such resources are part of language, identities are constructed and negotiated in linguistic
discourse. Identities are performed, constructed, enacted, produced, but only in interac-
tion with others. That is, ‘identities arise in interaction among people’ (Jørgensen 2010,
p. 4). As such, identities are to a large extent subject to negotiation. García (2010) refers
to the role of translanguaging in the negotiation and construction of identity, and suggests
that language choice involves negotiation in every interaction, as particular linguistic
resources may provide or prevent access to powerful social networks. That is, multilin-
gual speakers ‘decide who they want to be and choose their language practices accord-
ingly’ (García 2010, p. 524). However, not all linguistic resources are equally available to
all speakers at all times (Creese et al. 2006). Certain subject positions may either be non-
negotiable, or only partly negotiable, in particular places and at particular times, as social
contexts prevent individuals from accessing resources (Pavlenko and Blackledge 2004).
For these reasons, our analysis of language practices in superdiversity should attend to
spatio-temporal dimensions, and to relations of inequality and power.

4  Methodology and research design

The research reported in this paper investigated linguistic practices and identities in four
cities in England. The research project is a comparative sociolinguistic study of four
interlocking case studies with two researchers working in two complementary (‘herit-
age language’, ‘community language’, ‘supplementary’) schools in each community.
These are non-statutory schools, run by their local communities, which students attend
at the weekend or in the evening in order to learn the language normally associated with
their ethnic heritage. The case studies focused on Gujarati schools in Leicester, Turkish
schools in London, Cantonese and Mandarin schools in Manchester, and Bengali schools
in Birmingham. The project design is of four linking, ethnographically informed case
studies. Our research colleagues in the team were: Taşkin Baraç, Arvind Bhatt, Shahela
Hamid, Li Wei, Vally Lytra, Peter Martin, Chao-Jung Wu, and Dilek Yağcioğlu-Ali. Each
case study identified two complementary schools in which to observe, record, and inter-
view participants. We also collected key documentary evidence, and took photographs.
Two key participant children were identified in each school. These children were audio-
recorded during the classes observed, and also for 30 minutes before and after each class.
Towards a sociolinguistics of superdiversity 557

Key stakeholders in the schools were interviewed, including teachers and administrators,
and the key participant children and their parents. In all we collected 192 hours of audio-
recorded interactional data, wrote 168 sets of field notes, made 16 hours of video-record-
ings, and interviewed 66 key stakeholders.
The specific aims of the project were:
1. To explore the social, cultural and linguistic significance of complementary schools
both within their communities and in the wider society.
2. To investigate the range of linguistic practices used in different contexts in the com-
plementary schools.
3. To investigate how the linguistic practices of students and teachers in complemen-
tary schools are used to negotiate young people’s multilingual and multicultural
identities.
We have reported the findings of each separate case study elsewhere (Creese et al. 2007a,
b, c, d), and provided a detailed analysis and discussion of the research in a separate vol-
ume (Blackledge and Creese 2010). However, the present paper extends these discussions,
with specific reference to the emergent notions of superdiversity and translanguaging. In
the present discussion we focus exclusively on the two Bengali schools in Birmingham,
the second largest city in the UK after London, and the city with the highest proportion of
‘black and ethnic minority’ residents.

5  Negotiating superdiversity

5.1 Language as heritage

Our first example comes from an interview with the founder and administrator of one of
the Bengali schools. In the course of the interview the administrator made a forceful and
emotional statement following a question in which we queried the rationale for teaching
Bengali to children in Birmingham. He began his answer in Sylheti, and continued in
English:

Example 1 
ei bhaashar jonno 1952 te amaar theke dosh haath dure Barkat, Salam maara jaae 1952
te <because of this language in 1952 ten yards away from me Barkat and Salam were
killed in 1952> I was also a student in year 10. From Sylhet to Dhaka was 230 miles, we
marched there Sylhet to Dhaka 230 miles with slogans. We want our mother language, it
is a raashtro bhasha <state language> how will I forget about my mother language? my
brothers gave their life for this language. I will never forget it while I’m alive
(administrator interview, school A)

For the school administrator the ‘mother language’ was a vital symbol of the founding
of the Bangladeshi nation. More than fifty years earlier he had witnessed the incident in
which the ‘language martyrs’ were killed while demonstrating against the imposition of
Urdu as the national language by West Pakistan, and these events had informed his view
558 A. Creese and A. Blackledge

that British-born children of Bangladeshi heritage should learn and maintain the Ben-
gali language. The historic incident, which marks the Bangladeshi calendar as ‘Ekushey
February’, continues to be celebrated as a key moment in the collective memory of the
Bangladeshi nation, and in the Bangladeshi community in UK (Gard’ner 2004). It is
immediately clear that notwithstanding our argument that languages are socially con-
structed, for the school administrator the Bengali language holds considerable purchase
in terms of his identity and nationhood (May 2005). In our conversations with parents
and teachers in and around the Bengali schools in Birmingham we frequently heard the
view that the next generation should learn Bengali because ‘we are Bengali’. For many of
the participants passing on ‘the mother language’ was a symbolic means of transmitting
their heritage. It is clear that not only are spatio-temporal dimensions crucial to the mean-
ings evident here, but also the indexical nature of language (Blommaert 2010; Silverstein
2003), as one thing (here the Bengali language) points to another (Bengali/Bangladeshi
heritage/culture).
Of course the views of the school administrator, and of the teachers and parents, are
not necessarily shared by the students in the Bengali schools. The following example was
recorded on a hot Sunday afternoon in an overcrowded classroom in a borrowed space
(T is the teacher, and S1 and S2 students, both ten-year-old girls born in Birmingham to
parents who had migrated from Bangladesh):

Example 2 
T:  Bangla-e maato etaa Bangla class <speak in Bangla this is Bangla class> khaali
English maato to etaa Bangla class khene <if you speak in English only then why
is this the Bangla class?>
S1:  miss you can choose
S2:  I know English
S1:  why?
T:   because tumi Bangali <because you are Bengali>
S2:  my aunty chose it. she speaks English all the time.
(classroom audio-recording, school B)

In this interaction the teacher argues that the language of the classroom should be Ben-
gali. One of the students (S1) argues that it should be possible to choose which language
to speak in a particular context, and is backed up by her friend (S2). When S2 asks why
it is necessary to speak Bengali in class, the teacher says ‘because tumi Bangali’. In this
English and Bengali phrase the ideology of the school is summed up in the most succinct
terms. Bengali should be spoken, and should be learned, argues the teacher, because the
children are Bengali, as again the language is indexically linked to the heritage/culture
(here ‘Bangla’ refers to the Bengali language, while ‘Bangali’ refers to Bengali national
and/or ethnic belonging). S2 contests the teacher’s point, and in doing so (at this moment)
contests the ideology of the school. Reiterating her argument that it should be possible to
choose which language to speak, she cites her ‘aunty’, who has chosen to predominantly
speak English. The student’s ‘aunty’, herself of Bangladeshi heritage, is offered as an
example of someone who has resisted the notion of ‘one-language-equals-one-ethnic-
ity/culture’. For S2 language choice is flexible. For the school, in this example at least,
Towards a sociolinguistics of superdiversity 559

language learning is tied to ethnic and national belonging, and is inflexible. Having said
this, however, two ideologies of bilingualism are in play here. On the one hand the teacher
insists on an idealised construction of bilingualism which argues that the languages must
have separate functions, as Bengali is given the status of classroom language. On the
other hand another ideology and practice of bilingualism comes into play, as language is
used flexibly by the teacher (‘because tumi Bangali’) to interact with her pupils. In these
examples it is clear that linguistic ideology and practice are in complex relation, as certain
resources are held to be crucial to the heritage and identity of the group, but in practice
they constitute features of translanguaging repertoires.

5.2 Language as distinction

In conducting our fieldwork we found clear evidence that it is not sufficient to take one
‘ethnicity’ or ‘country of origin’ as our unit of analysis. That is, we found that our par-
ticipants made distinctions within the ‘ethnic group’, and these distinctions were often
represented as linguistic differences. What people believed about their language (or other
people’s languages), and the situated forms of talk they deployed, revealed divergent and
contested views about the value and status of particular linguistic resources. We saw that
attitudes to, and practices of, linguistic repertoires were tied in complex relation to dis-
tinctions made between speakers of those repertoires.
When we interviewed the administrators and teachers in the schools they spoke
emphatically about the need for children to learn Bengali, the standard, literate language
of Bangladesh. This was frequently held to be oppositional to Sylheti, which was the spo-
ken variety used by the families of students attending the Bengali schools. Whereas Ben-
gali was the language of the educated elite in Bangladesh, particularly Dhaka, Sylheti was
the language of rural Sylhet in north-east Bangladesh. One of the school administrators
(of a different school from that in Example 1) was emphatic that Bengali was not the same
as Sylheti, and that Sylheti should not be allowed to ‘contaminate’ the standard form. He
was concerned that Sylheti forms were beginning to appear in the spelling and gram-
mar of Bengali newspapers in UK, introducing ‘thousands of spelling mistakes–Bengali
newspapers I have seen in many places the spelling was wrong, sentence construction
was wrong’. For the administrator non-standard resources were ‘contaminating the lan-
guage’. He made this point about the necessity for children to learn standard Bengali:

Example 3 
I am always in favour of preserving languages and all these things. but it doesn’t mean
that this should contaminate other languages and give this more priority than the proper
one. we have to preserve the proper one first, and at the same time we have to encourage
them to you know, use their dialect. but we shouldn’t make any compromise between
these two
(administrator interview, school B)

This was a strongly articulated argument among people we met in our field work. The
administrator of the other school stated that:
560 A. Creese and A. Blackledge

Example 4 
bhasha to bolle Bangla bhasha bolte hobe Sylheti kono bhasha naa
<when you talk about language it means Bengali, Sylheti is not a language>
(administrator interview, school A)

For several respondents ‘Bengali’ constituted a more highly valued set of linguistic
resources than ‘Sylheti’, and was regarded as the ‘proper’ language. Pujolar (2007, p. 78),
referring to a different socio-historical context, makes the point that language policy may
operate to foster knowledge of some languages, ‘but delegitimise or ignore other lan-
guages and other forms of multilingual competence and performance’. Patrick (2007,
p. 127) similarly finds that in arguing in support of a particular language, ‘speakers can
be locked into fixed or essentialised notions of identity, ‘authenticity’ and place, which
provide no recognition of mobile, postcolonial speakers’. It was clear that for some of our
respondents not all linguistic resources were equally valued, and while some sets of lin-
guistic resources were considered to be ‘a language’, others were not. In this sense there
was a constant re-invention of ‘language’ on the part of some participants.
Those who spoke ‘Sylheti’ were often criticised by ‘more educated’ people who spoke
‘Bengali’. They were characterised by the administrator of one of the schools as members
of the ‘scheduled’, or ‘untouchable’ caste: people without rights or resources in the Indian
sub-continent:

Example 5 
publicraa ki dibe amar aapne especially bujhben amader desher je shob lok aashche ora
kon category lok aashchilo, mostly from scheduled caste, gorib, dukhi krishokra aash-
chilo. oder maa baba o lekha pora interested naa oder chele meye raa o pora lekha inter-
ested naa. oraa baidhitamolok schoole jete hoe primary schoole sholo bochor porjonto
jete hoe, ei jonne schoole jaai.
<what will the public contribute? you [the researcher, Shahela Hamid] especially will
understand what type of people came from our country. they belonged to the category of
scheduled caste, they are the poor, the deprived, farmers. their parents were not interested
in education nor are the children interested. they go to school because it’s compulsory>
(administrator interview, school B)

Here Sylheti speakers are referred to as the ‘scheduled caste’. Regarded as the least edu-
cated group in society, with no resources of any kind, they are considered to be the lowest
of the low (Borooah et al. 2007; Borooah 2005; Kijima 2006). Here linguistic features
were viewed as reflecting and expressing broader social images of people. Irvine and
Gal (2000, p. 37) suggest that people’s ideologies about language often “locate linguistic
phenomena as part of, and evidence for, what they believe to be systematic behavioural,
aesthetic, affective, and moral contrasts among the social groups indexed”. One of the
teachers argued that children should learn Bengali for ‘moral reasons’. Irvine and Gal
propose that a semiotic process of iconisation occurs, in which linguistic features that
index social groups appear to be iconic representations of them, as if a linguistic feature
depicted or displayed a social group’s inherent nature or essence. Bourdieu and Darbel
Towards a sociolinguistics of superdiversity 561

(1991, p. 112) argue that some more powerful groups provide ‘an essentialist representa-
tion of the division of their society into barbarians and civilized people’. Here the fact of
speaking ‘Sylheti’, rather than ‘Bengali’, appeared to index the Sylheti group in particu-
larly negative terms, despite the relative similarities between the ‘Bengali’ and ‘Sylheti’
sets of linguistic resources.
Whilst some speakers in our study considered ‘Sylheti’ to be quite different from ‘Ben-
gali’, others regarded the two sets of resources as indistinguishable. As we have seen,
there were several instances of participants commenting on the differences between Syl-
heti and Bengali in terms of social status and value, but not everyone agreed about the
extent to which these sets of linguistic resources were distinct. While the administrator of
one of the schools argued that Bengali and Sylheti were ‘completely different’, a student’s
mother said they were ‘thoraa different’ <a little different>, while other parents also held
this view, saying they were ‘little bit different thaake’ <only> and even ‘the same’. Here
there was clear disagreement about the nature and extent of the differences between the
sets of linguistic resources used by the students’ parents at home, and the literate version
of the language taught in the complementary school classrooms. That is, there was disa-
greement about the permeability of the boundaries between languages. These differences
of perception were likely to be ideological. Those who argued that the ‘languages’ were
completely different from each other were speakers of the prestige language, unwilling
to allow the lower status language to contaminate their linguistic resources. Those who
argued that the ‘languages’ were almost the same as each other were speakers of Sylheti,
which was held to index the lower status, less educated group.
We asked one of the key participant children, eleven-year-old Masuda, and her cousin
Maria, about the languages they use at home. In this extract from field notes they argue
that they are not from Sylhet:

Example 6 
Masuda says ‘we speak the proper Bengali’, while Maria says ‘no, we speak a little bit
lower class, a little bit’. Masuda disagrees: ‘no, we do speak high class’. Masuda says
her father is from Dhaka. They say they are learning to join Bengali letters, and talk
me through an elaborate technical explanation of how letters are joined in Bengali, and
how this affects meaning and sound. They say they are ‘not bad, quite good’ at reading
Bengali, ‘but we were brought up to speak and read English, so we are a bit confused, a
little bit’
(Field notes, school A)

The children were not asked about associations between Bengali and social status, but
here they have a clear sense that it is possible to speak ‘high class’ or ‘low class’. They
also have a sense of their own achievement in learning Bengali.
On many occasions children demonstrated their awareness of differences between Syl-
heti and Bengali. The following example is from field notes taken during the first week of
observation in school B. The teacher (Al) is working with the older group of children:
562 A. Creese and A. Blackledge

Example 7 
Al administers to the older children what appears to be an ad hoc and informal test of
Bengali vocabulary. The children are enthusiastic about this, clearly enjoying their mul-
tilingualism. The group of girls is a little boisterous, but Al appears unworried by this.
When he asks them to compose a sentence in Bengali one of the girls, Tabeya, says in
English: “we don’t use this language”. Al is amused, and asks them to compose a sentence
in Sylheti. There is a discussion about linguistic differences between Sylheti and Bengali
in the phrase ‘I have a friend’. Tabeya says: “We say aamaar shoi aasoin (Sylheti), but
you say ‘aamaar ekti shoi aase’ (Bengali)”. She also gives an example in English, Sylheti
and Bengali: ‘aamaar ekti friend aase’. The notable thing here is that the Bengali phrase is
accompanied by much eye-rolling and eyebrow-raising from the speaker, and intonation
which indicates that speaking the phrase in Bengali is associated with a different social
class from the Sylheti phrase. That is, Bengali appears to be associated with putting on
airs, showing off, or sophistication.
(Field notes, school B)

In saying “we don’t use this language” the collective pronoun appears to refer to the
children and their families, distinguishing them from the teacher. Here the children, and
Tabeya in particular, show an understanding that Sylheti and Bengali are different from
each other, and also that they index different social groups and values. However, even in
this instance where linguistic differences are explicitly marked by children, and acknowl-
edged by teachers, the simple dichotomy between Sylheti and Bengali breaks down:
Tabeya borrows the Bengali word ‘shoi’ in her Sylheti sentence and uses the more polite
form of the Sylheti verb ‘aasoin’ in her sentence. Despite the clear ideological differences
between Sylheti and Bengali for many of our participants, the varieties were frequently
mixed in their linguistic practices.
There was an awareness of Bengali as the higher-status language in relation to Sylheti,
on the part of teachers (“I talk posh Bengali, and the children can’t understand me”),
children, administrators, and parents. The following extract is from audio recording of
another key participant child, Alamghir, at home. Here his mother asks him whether he
can speak ‘proper Bangla’:

Example 8 
Mother:    Banglae khoilaa naa. Don’t you speak Bangla proper Bangla.
Alamghir:  yeah. I can speak Dhaka language,
Sister:      go on then
Alamghir:  bala aachen <are you well?>
Sister:      that’s all you know
Mother:    baalaa aasen in Sylheti. In Dhakaaiyya you say aapni ki bhaalo aachen
Sister:      you know the lady
Alamghir: khub bhaalaa naam <very nice name> There you go you can’t speak English
(home audio-recording, school A)

Here Alamghir’s mother, aware of the digital recording device, jokingly asks her son
whether he can speak Bengali. He claims (in English) to be able to do so, and is teased
Towards a sociolinguistics of superdiversity 563

by his sister. Alamghir’s subsequent attempt to speak in the standard language is cor-
rected by his mother. Perhaps irritated by this, Alamghir retorts that his sister is not able
to speak English. In these examples Bengali was associated with being ‘high class’, and
was regarded as the ‘proper’ language. Here the question of speaking Bengali became an
opportunity for gentle teasing and mockery within the family.
On many occasions the research participants interactionally evidenced their aware-
ness of differences (perhaps mainly in status and value) between ‘Sylheti’ and ‘Bengali’.
The following example was recorded at the dinner table in the family home of one of the
students:

Example 9 
Mother:  khitaa hoise? Tamim, khaibaani saatni?
<what is the matter? Tamim, would you like some relish?>
Father:   aaro khoto din thaakbo
<how many more days is that [voice recorder] going be with you?>
Tamim:   aaro four weeks <four more weeks>
Father:   (XXXX)
Student:  no they said any. if you talk all English
Father:   ginni, oh ginni [calling his wife using a highly sylised Bengali term of
endearment]
Mother:   ji, hain go daakso kheno <yes, dear why are you calling me?>
tumaar baabaa shuddho bhasha bolen
<your father is speaking the standard language>
Father:   paan dibaa <can I have some paan> aapne aamaar biyaai kemne <how are you
my relation?>
(home audio-recording, Bengali case study)

Here the Sylheti-speaking parents play the roles of Bengali speakers, appropriating an
exaggereted, literary version of ‘high’ Bengali, and adopting the airs and graces which
they see as characteristic of the Bengali-speaking group. The terms of endearment used
here (‘ginni’, ‘hain go’) are forms of parody (Bakhtin 1973; 1984; 1986), exaggerations
beyond common usage, as speakers of Bengali are caricatured in mock-sophisticated dis-
course. This brief interaction is situated in a whole hinterland of language ideological
beliefs and practices, as the couple acknowledge differences between Bengali and Sylheti
as sets of linguistic resources, and the conditions which differentially provide and con-
strain access to linguistic resources. In parodic discourse the parents introduce into their
own voices the exaggerated voice of the Bengali speaker, and that voice clashes with its
host, as ‘discourse becomes an arena of battle between the two voices’ (Bakhtin 1994,
p. 106). The impromptu role-play light-heartedly, but not half-heartedly, ‘parodies anoth-
er’s socially typical…manner of seeing, thinking and speaking’ (Bakhtin1994, p. 106).
Here the repetition of a notional ‘high-class’ discourse is both creative (Pennycook 2010)
and evaluative (Vološinov 1973).
We heard further distinctions between certain sets of linguistic resources which to the
casual observer may have seemed to be identical, but which an emic perspective repre-
sented as having very important differences. In an interview with siblings Shazia and
564 A. Creese and A. Blackledge

Tamim, aged 10 and 11 years respectively, the children mentioned a drama activity, based
on a story of new arrivals from Bangladesh. In talking to the researcher (R) they described
this group as ‘freshies’ (cf. also Martin et al. 2004):

Example 10 
R:      what do you mean ‘freshie’, what does that mean?
Tamim:  freshie as in a newcomer
R:       is that bad to say to somebody?
Tamim: yea it’s kind of like a blaze but it’s also a word to describe a new person coming
from a different place
Shazia:  it’s not a good thing
Tamim: it’s kind of both. if you say it as in trying to tease somebody, ‘freshie’, and we
say it as in erm trying to say erm, as in they’re newcomers and they come from
a different country for the first time
R:       could you tell if someone was ‘freshie’?
Shazia:  well from Bangladesh it’s not always their skin colour, it’s sometimes how
they talk
R:     how do you talk ‘freshie’?
Tamim:  it’s kind of like they don’t know that much English
Shazia:  they might just show off in their language but if you ask them a question in
English they just
Tamim:  they’re like ‘what’, ‘what’, you know
Shazia:  they say strange words in their language and if you ask them a question in Eng-
lish they just say ‘what’ in their language
(student interview, school B)

Shazia and Tamim negotiate their identity in opposition to that of the newly arrived chil-
dren, repeatedly referring to ‘their’ language, which they see as different from the language
they speak themselves. Here ‘what what’ is spoken with an intonation which appears to
represent some confusion on the part of the newly-arrived group. Here is another form
of parody, as the British children represent the new arrivals as inferior to themselves,
and make indexical links between linguistic resources and membership of the in-group.
Although the students speak the same ‘language’ as the new arrivals in daily interactions
with their parents, they nevertheless indicate that ‘how they talk’ is one of the defining
ways in which the ‘freshies’ are different from them. That is, the way the new arrivals
talk distinguishes them from the British-born children, and this distinction is more than
linguistic.
For some of our participants, some sets of linguistic resources were very considerably
privileged above other, similar sets of linguistic resources. While linguistic resources
which were described as ‘standard’, or ‘proper’, or ‘real’, or ‘book’ Bengali had come to
represent the ‘heritage’ of the Bangladeshi nation, sets of resources described as ‘Sylheti’
had come to be associated with the uneducated poor, who were held to be disinterested in
schooling, and unmotivated. Similarly, sets of linguistic resources associated with Bang-
ladesh were distinguished from very similar linguistic resources associated with Britain
(or Birmingham, or even a specific local neighbourhood in Birmingham). However, we
Towards a sociolinguistics of superdiversity 565

also saw that these distinctions were contested by others, who denied that clear differ-
ences existed, or made fun of the assumption that these differences were constitutive
of differences in social status. That is, our participants represented disagreements about
what constituted (a) language, and about the ideological links between speakers and the
sets of linguistic resources which they called into play.
What we have seen so far is that in the superdiverse city of Birmingham nuanced
linguistic differences, some of them agreed upon and others not, come to act as subtle
distinctions between groups and individuals. While some sets of resources are ideologi-
cally viewed as representing dimensions of nation, heritage, and culture, these same sets
of resources also come to represent their speakers as superior to speakers of other sets
of resources. Concomitantly, of course, speakers of the inferior variety are regarded as
inferior. Blommaert takes Foucault’s notion of ‘orders of discourse’, together with Sil-
verstein’s (2003) ‘order of indexicality’, to propose the term ‘orders of indexicality’, in
which ‘some forms of semiosis are systematically perceived as valuable, others as less
valuable and some are not taken into account at all, while all are subject to rules of access
and regulation as to circulation’ (Blommaert 2010, p. 38). Orders of indexicality are strati-
fied and systematically give preference to some modes of semiosis over others. It is such
systematic perceptions which determine that speakers of ‘Bengali’ are presupposed to be
of higher class or caste than speakers of (say) Sylheti or Hobiganji. These structures, con-
stituted in discourse, powerfully turn difference into inequality (Blommaert 2010, p. 41).

5.3 Translanguaging in superdiversity

We suggested earlier that a feature of superdiversity is that multilingual people do not

habitually make meaning through separate monolingualisms, but interweave and intermesh
a range of resources across borders and boundaries. This is where our study takes us for-
ward. Characteristic of the linguistic interactions of the students we audio-recorded in
and out of complementary school classrooms was a playfulness and creativity. Students
engaged with and accessed a broad range of linguistic resources. Bangladeshi-heritage
children watched Hindi films, and were familiar with Hindi songs. They sang along with
the songs, and were able to express their preferences and dislikes. In the following exam-
ple two sisters, Rumana (age 11) and Aleha (age 10) are watching a film just before going
to Bengali class:

Example 11 
[Rumana sings with the music on TV]
Rumana:  it’s a funny movie that. this one, Hera Pheri. really funny, I like this song
Aleha:  I like [to baby sister] talk, talk, say amaar naam Durdana say amaar naam
Durdana <say my name is Durdana>
Durdana:  one khe <who’s there?>
Rumana:  [singing along in Hindi and English] rock your body, rock your body, rock
your body, rock your body, tumhare bina <without you> chaenna aaye <there’s
no peace> rock your body
(home audio-recording, school B)
566 A. Creese and A. Blackledge

Here singing along with the Hindi film music (‘tumhare bina’ etc) seems to be a usual
feature of the children’s linguistic world, as they move in and out of English, Sylheti, and
Hindi while listening to, participating in, and enjoying the Hindi film. At the same time
they engage bilingually with their baby sister Durdana’s attempts to speak into the digital
recording device. Moments later the sisters are about to leave for their Bengali class:

Example 12 
Aleha:  Rumana, come on. I’m going amma, salam alaikum
 <mother, salam alaikum >
 salam alaikum abba, zaairam aami
 <salam alaikum father. I’m going>
(home audio-recording, school B)

Notable here is the unmarked and quite usual multilingualism of the interaction: English,
Sylheti, and an Arabic-derived phrase enjoy a flexible and non-conflictual co-existence.
We recorded many instances of flexible linguistic practice in the homes of students who
attended the Bengali schools. In the following example Tamim, a ten-year-old boy, is ask-
ing his mother whether he is allowed to go on the school camping trip:

Example 13 
amma aami camping-e zaaitaam. aafne last year here disoin aamaare disoinnaa. aami
camping zaaitaam aafne aamaare disenna
<mother, I want to go camping you allowed him last year but not me. I want to go
camping you didn’t allow me last year>
(home audio-recording, school B)

This is an unremarkable, quite usual example of flexible language practice in the students’
family settings, of the sort we heard on each occasion we audio-recorded the children and
young people at home. The Bengali suffix to the English word (‘camping-e’) was a rela-
tively common feature of translanguaging practice in the family and classroom settings.
Tanzim’s mother goes on to give him and his sister Tabeya a message for the complemen-
tary school teacher:

Example 14 
Mother:  Sir-re khoiyyo next week dibaa foisha.
  <Tell Sir, I’ll give the money next week>
  Seemare nitaanaa? e Tabeya doro
  <Aren’t you taking Seema with you? Tabeya, hold this>
  Seemare loiyya zaaito tor abbae
  <Take Seema, your father>
Tabeya:  Next week will be too late.
Mother:  Next week dimune. Tor abbaare dekhaaite hobe, foishao ghore naai.
  <I’ll give it next week. I have to show it to your father, there’s no money in the
house>
  Next week khoiyya dilaaitam
Towards a sociolinguistics of superdiversity 567

  <Next week I’ll tell him and pay the money>

  etaa roisenaa record hoise?
  < Has this part been left out or recorded?>
  Next Sunday, Saturday dimu, naa faaile naai
  <I’ll give it next Sunday, next Saturday if I don’t give it that’s it>
(home audio-recording, school B)

Here Tabeya’s utterance is entirely in English, while her mother uses Sylheti and English:
“next week”, “baby oh baby”, “record”, and “next Sunday, Saturday”. This is typical talk
in the domestic setting, with parents moving between Sylheti and English, and children
mainly using English. In Example 15 Tanzim and Saleha are discussing the forthcoming
school trips. Tanzim’s mother is present:

Example 15 
Saleha:   Do you have that channel? Do you have that channel?
Mother:  Saleha, khaiyya laao, beti
  <Saleha, have something to eat, girl>
Saleha:   Are you going to see the movie?
  Tabeya did you watch from the beginning? Hmn, why
Mother:   Tumaar chulto aasraaiyya zaao
  <Comb your hair before you go out>
Tanzim:   They are both the same
Saleha:   Can I stick it on this?
Tanzim:   She is sixteen, she is sixteen.
Saleha:   We’re gonna go on that one. Southend-on- Sea one, we might
Tanzim:   Tabeya I’m twice as lucky as you I’m gonna go to camping and  there are four
places left
Saleha:  Cadbury World khali
  <Cadbury World only>
  four places left. the paper or the fee. Taaraa khitaa likhoin.
  <what are they writing?>
  Sir khoisoin saartaa place roise, zigaaite sir-e
  <Sir says there are four places left, ask Sir>
(home audio-recording, school B)

Here Saleha moves between English, Sylheti and mixed English/Sylheti with ease. The
fact that Tanzim’s mother is present, and that she speaks Sylheti during this episode, may
influence the language choice during this interaction. However, it appears that Saleha is
speaking to the other children on each occasion, suggesting that her choices are not based
on different interlocutors, but on her ease with a multilingual domestic environment. In
the final example Tanzim’s mother has asked the teacher to go to her house to clarify the
cost of the forthcoming trip:
568 A. Creese and A. Blackledge

Example 16 
Tanzim:  Come in, sir
Teacher:  koi, koi, salaam alaikum
   <where, where, salam alaikum>
Mother:   booking form
Teacher:   salam alaikum, aapne bhaalo <salam alaikum, are you well?>
Mother:   ji baala <yes fine> shobe zaaite saae <all of them want to go>
Teacher: okay fruit picking, outdoor sport, boat trip-e three—you are not going
bowling
Tanzim:  yeah because my because my cousin we’re gonna go to Bournemouth.
Teacher:  achchaa koetaa seat aache dikhe
  <okay, let me see how many seats are available>
Mother:  koto seat aache aamraa zaanina <we don’t care how many seats are available>
era prothom students eder niye <they are your first students with them>

Here Tanzim’s mother is forthright in her negotiations with the teacher, who is the organ-
iser of the trips. She insists that her children have priority, as they have attended the
school longer than other pupils. This negotiation is conducted in a linguistic variety which
appropriates features of Bengali, Sylheti, English, and formal Arabic for the greeting, and
again an example of a Bengali suffix to an English word (‘boat trip-e’).
In summary, there were frequent examples of children and adults using more than
one language in an utterance. These were usually English/Sylheti examples, but there
were also examples of English/Bengali and Sylheti/Bengali. Here meaning-making was
conducted in linguistic repertoires which moved across and between ‘languages’, using
linguistic items across borders and boundaries. These examples of utterances seemed to
be an indication of the ease with which children, parents, and teachers adopted translan-
guaging repertoires in the multilingual settings of home and complementary school.

6  Language and superdiversity

Research in the field of superdiversity has called for an anthropology of daily habits and
interactions, with a focus on the intermeshing of variables which weave together in newly
diverse communities (Vertovec 2007b). This research further calls for analysis of such
interactions which takes account of plurality of affiliation, the coexistence of cohesion
and separateness, and the fact that people cohere to different social worlds and com-
munities simultaneously. An anthropological lens must have a linguistic focus if it is to
understand what interactions and practices look like in newly diverse neighbourhoods.
The development of a theoretical framework for the study of language in superdiversity
develops, of course, from practices and interactions themselves, rather than from ‘big
theory’. However, ‘Big things matter if we want to understand the small things of dis-
course’ (Blommaert 2010, p. 41), and we can begin to piece together some of the compo-
nents of a theoretical perspective on linguistic approaches to superdiversity by reflecting
critically on the linguistic phenomena we encountered in and around the Bengali schools
in Birmingham.
Towards a sociolinguistics of superdiversity 569

In the stories of martyrdoms and heroic deeds which we often heard in and out of
the Bengali classrooms, it was clear that dimensions of time and space were crucial to
our understandings of beliefs about, and values attached to, certain linguistic features
and repertoires. But all of the voices we heard as we spent time in classrooms and fam-
ily homes were equally subject to analysis in terms of their location in and mobility
through time and space. Linguistic practices were always local, but at times they moved
between localities, were ‘translocal’ (Blommaert 2010), travelling (to put it crudely) in an
ideological triangle between inner-city Birmingham, a village in rural Sylhet, and Dhaka,
the capital city of Bangladesh. But locality was more than this, as digital communica-
tion made available resources which superseded territorial boundaries, offering linguistic
resources which resided in none of these localities, and may have been more at home on
the streets of New York City or Mumbai. One of the particularly striking aspects of the
voices we heard was the way in which sets of linguistic items were indexically linked
to their speakers. We encountered powerful discourses, ‘systemic patterns of indexical-
ity which were also systemic patterns of authority, of control and evaluation, and hence
of inclusion and exclusion’ (Blommaert 2010, p. 38). These patterns of indexicality had
become sedimented over time, over generations, in other times and other spaces, and
were now ‘relocalized’ (Pennycook 2010) in Birmingham as discriminatory discourses.
Once again it was necessary to look historically and geographically to make sense of
linguistic phenomena. More than this, local practice was frequently (and perhaps always)
subject to forces (ideologies) beyond the local. When a family sits at the dinner table, or a
student sits in a classroom, and makes fun of those very same discriminatory discourses,
they are not only entertaining themselves and others, but engaging with discourses which
have become sedimented through repeated acts of sameness (Pennycook 2010), many
times, in many places. That is, when Tabeya makes a light-hearted assertion about some
lexical difference between Bengali and Sylheti, that is all she is doing. But in our field
notes we can see that her statement is accompanied by a bodily demonstration of her keen
awareness of the ideological differences between those lexical items. Our audio-record-
ings, field-notes, interviews and other data, collected in and around Bengali schools in
Birmingham, suggest some dimensions of analysis to which we should be alert in our
ongoing research. These include the need to look historically and geographically, to situ-
ate discourse in terms of the patterns and beliefs evident in orders of indexicality.
None of this is meaningful, though, without reference to the notion of voice. If we are
to understand the meanings of discourses in contexts where a wide range of diversities
interweave and intermesh, we need to be alive to the idea that no ‘voice’ is single. That
is, voice is always plural, always multiple. This is not to say that every voice explicitly
speaks from more than one point of view. Not every speaker argues that ‘on the one
hand this, and on the other hand that’. But every voice bears the traces of other times
and other spaces, of words uttered before. Even simple repetition is never simple repeti-
tion: ‘repetition always entails difference, since no two moments, events, words can be
the same’ (Pennycook 2010, p. 43). Every utterance is therefore different, but bears the
marks of repetition. Every word ‘is entangled, shot through with shared thoughts, points
of view, alien value judgements and accents, weaves in and out of complex interrelation-
ships, merges with some, recoils from others…and having taken meaning and shape at
a particular historical moment in a socially specific environment, cannot fail to brush up
570 A. Creese and A. Blackledge

against thousands of living dialogic threads’ (Bakhtin 1981, p. 276). Here we move from
a view of transnationalism as ‘bifocality’ (Vertovec 2009) to a view of superdiversity as
multivocality. It is by looking closely, putting language under the microscope but also
in context, that we can identify at least some of Bakhtin’s thousands of living dialogic
threads. Analytical metaphors such as scale and indexicality are valuable points of entry
as we seek to understand the complexity of voice in superdiverse communities.
In this paper we have not engaged significantly with the educational implications of a
sociolinguistics of superdiversity. There is a good deal of work to do here. García (2009,
2010) has begun this work, arguing that in the education of multilingual students in the
twenty-first century we need to break visions of whole languages so that we can construct
new ones that are ‘more inclusive of differences, of the translanguaging that is so extensive
in bilingual communities and classrooms’ (García 2009, p. 387). The challenge, she suggests,
will be whether states can let go of their language-identity-allegiance associations, and allow
children to develop their multiple identity positions as they construct abilities to translanguage
and use multiple language in the flexible ways of the future. García (2010, p. 532) proposes
that we put at the centre people as actors who signify differently by performing different
language practices and ethnic practices. That is, language practices are most important for
the lives of people, and should form the centre of our analysis, not the periphery. The ability
to language, García reminds us, is ‘the most important signifying role of human beings—that
which gives life meaning’ (2007, p. 519). It is through languaging that people perform their
identifying. What we find as we look closely at language practices in superdiverse settings
is that new repertoires are emerging as people use linguistic items from a range of sources to
make meaning. Translanguaging practice puts alongside each other the local and the global,
offering languaging acts which are important for identity (García 2010). In developing a
sociolinguistics of superdiversity we should look closely at practices of translanguaging, and
consider the histories, geographies, and indexical orders which shape them.

References

Anderson, B. (1998). The Spectre of Comparisons: Nationalisms, Southeast Asia, and the World.
London: Verso.
Bakhtin, M. M. (1973). Problems of Dostoevsky’s poetics (Translated by R. W. Rotsel). Ann Arbor:
Ardis.
Bakhtin, M. M. (1981). The dialogic imagination. Four essays (Edited by Michael Holquist; trans-
lated by C. Emerson and M. Holquist). Austin: University of Texas Press.
Bakhtin, M. M. (1984). Problems of Dostoevsky’s poetics (Edited and translated by C. Emerson).
Manchester: Manchester University Press.
Bakhtin, M. M. (1986). Speech genres and other late essays (Edited by C. Emerson & M. Holquist).
Austin: University of Austin Press.
Bakhtin, M. M. (1994). Problems of Dostoevsky’s poetics. In P. Morris (ed.), The Bakhtin reader.
Selected writings of Bakhtin, Medvedev, Voloshinov (pp. 110–113), London: Arnold.
Blackledge, A. (2009a). ‘As a country we do expect’: The further extension of language testing
regimes in the UK. Language Assessment Quarterly, 6, 6–16.
Blackledge, A. (2009b). ‘Inventing English as convenient fiction: Language testing regimes in the
United Kingdom’. In G. Extra, M. Spotti, & P. VanAvermaet (eds.), Language testing, migra-
tion and citizenship: Cross-national perspectives (pp. 65–85). London: Continuum.
Towards a sociolinguistics of superdiversity 571

Blackledge, A. (2009c). Being English, speaking English: Extension to English language testing
legislation and the future of multicultural Britain. In G. Hogan-Brun, C. Mar-Molinero, & P.
Stevenson (eds.), Discourses on language and integration: Critical perspectives on language
testing regimes in Europe (pp. 83–108). Amsterdam: John Benjamins.
Blackledge, A., & Creese, A. (2010). Multilingualism. A critical perspective. London: Continuum.
Blackledge, A., & Pavlenko, A. (2001). Negotiation of identities in multilingual contexts. Interna-
tional Journal of Bilingualism, 5(3), 243–259.
Blommaert, J. (2010). The sociolinguistics of globalization. Cambridge: Cambridge University Press.
Bolognani, M. (2007). The myth of return: Dismissal, survival or revival? A Bradford example of tran-
snationalism as a political instrument. Journal of Ethnic andMigration Studies, 33(1), 59–76.
Borooah, V. K. (2005). ‘Caste, inequality, and poverty in India’. Review of Development Econom-
ics, 9(3), 399–414.
Borooah, V. K., Dubey, A., & Iyer, S. (2007). The effectiveness of jobs reservation: Caste, religion
and economic status in India. Development and Change, 38(3), 423–445.
Bourdieu, P., & Darbel, A. (1991). The Love of Art. European museums and their public. London:
Polity Press.
Cox, J., & Connell, J. (2003). Place, exile and identity. The contemporary experience of Palestin-
ians in Sydney’. Australian Geographer, 34(3), 329–343.
Creese, A., & Blackledge, A. (2010). Translanguaging in the bilingual classroom: A pedagogy for
learning and teaching. Modern Language Journal, 94(2), 103–115.
Creese, A., Bhatt, A., Bhojani, N., & Martin, P. (2006). Multicultural, heritage and learner identities
in complementary schools. Language and Education, 20(1), 23–44.
Creese, A., Bhatt, A., & Martin, P. (2007a). Investigating multilingualism in Gujarati complemen-
tary schools in Leicester. Birmingham: University of Birmingham.
Creese, A., Blackledge, A., & Hamid, S. (2007b). Investigating multilingualism in Bengali comple-
mentary schools in Birmingham. Birmingham: University of Birmingham.
Creese, A., Lytra, V., Baraç, T., & Yağcıoğlu-Ali, D. (2007c). Investigating multilingualism in Turk-
ish complementary schools in London. Birmingham: University of Birmingham.
Creese, A., Wu, C. J. & Wei, L. (2007d). Investigating multilingualism in Chinese complementary
schools in Manchester. Birmingham: University of Birmingham.
Gal, S. (2006). Migration, minorities and multilingualism: Language ideologies in Europe. In C.
Mar-Molinero & P. Stevenson (eds.), Language ideologies, policies and practices. Language
and the future of Europe (pp. 13–27). Basingstoke: Palgrave Macmillan.
García, O. (2007). Foreword. In S. Makoni & A. Pennycook (eds.), Disinventing and reconstituting
languages (pp. xi–xv). Clevedon: Multilingual Matters.
García O. (2009). Bilingual education in the 21st Century: A global perspective. Oxford: Wiley.
García, O. (2010). Languaging and ethnifying. In J. Fishman & O. García (eds.), Handbook of
language and ethnicity: Disciplinary and regional perspectives (2nd ed., vol. 1, pp. 519–535).
New York: Oxford University Press.
Gard’ner, J. M. (2004). Heritage protection and social inclusion: A case study from the Bangladeshi
community of East London. International Journal of Heritage Studies, 10(1), 75–92.
Giddens, A. (1984). The constitution of society. Berkeley: University of California Press.
Glick Schiller, N., & Fouron, G. (2002). Georges woke up laughing: Long distance nationalism and
the search for home. New York: Duke University Press.
Glick Schiller, N., Basch, L., & Blanc-Szanton, C. (1992). Towards a transnational perspective on
migration: Race, class, ethnicity and nationalism reconsidered. Annals of New York Academy
of Science, Vol. 645. New York: New York Academy of Science.
Gumperz, J. (1982). Discourse strategies. Cambridge: Cambridge University Press.
Harris, R. (2006). New ethnicities and language use. Basingstoke: Palgrave Macmillan.
Heller, M. (2007). Bilingualism as ideology and practice. In M. Heller (ed.), Bilingualism: A social
approach (pp. 1–24). Basingstoke: Palgrave.
572 A. Creese and A. Blackledge

Heller, M., & Duchêne, A. (2007). Discourses of endangerment: Sociolinguistics, globalization,

and social order. In A. Duchêne & M. Heller (eds.), Discourses of endangerment. Ideology and
interests in the defence of languages (pp. 1–13). London: Continuum.
Hess-Lüttich, E. (1978). Bilingualismus als Stilmittel: Drogenkultur und Sprache. In G. Nickel
(ed.), Rhetoric and Stylistics (pp. 69–82). Stuttgart: HochschulVerlag.
Irvine, J., & Gal, S. (2000). Language ideology and linguistic differentiation. In P. Kroskrity (ed.),
Regimes of language. Ideologies, polities and identities (pp. 35–84). Santa Fe: School of
American Research Press.
Jørgensen, J. N. (2010). The sociolinguistic study of youth language and youth identities. In J. N.
Jørgensen (ed.), Love ya hate ya: The sociolinguistic study of youth language and youth identi-
ties (pp. 1–15). Cambridge: Cambridge University Press.
Kijima, Y. (2006). Caste and tribe inequality: Evidence from India, 1983–1999. Economic Develop-
ment and Cultural Change, 54, 369–404.
Levitt, P., & Glick Schiller, N. (2004). Conceptualising simultaneity. A transnational social field
perspective on society. International Migration Review, 38(3), 1002–1039.
Makoni, S., & Mashiri, P. (2007). Critical historiography: Does language planning in Africa need a
construct of language as part of its theoretical apparatus? In S. Makoni & A. Pennycook (eds.),
Disinventing and reconstituting languages (pp. 62–89). Clevedon: Multilingual Matters.
Makoni, S., & A. Pennycook (2007). Disinventing and reconstituting languages. In S. Makoni &
A. Pennycook (eds.), Disinventing and reconstituting languages (pp. 1–41). Clevedon: Multi-
lingual Matters.
Martin, P., Creese, A., Bhatt, A., & Bhojani, N. (2004). Final report on complementary schools and
their communities in Leicester. Leicester: University of Leicester & University of Birmingham.
Maryns, K. (2006). The asylum speaker. Language in the Belgian asylum procedure. London: St Jerome.
May, S. (2001). Language and minority rights: Ethnicity, nationalism, and the politics of language.
London: Longman.
May, S. (2005). Language rights: Moving the debate forward. Journal of Sociolinguistics, 9(3), 319–347.
Patrick, D. (2007). Language endangerment, language rights and indigeneity. In M. Heller (ed.),
Bilingualism: A social approach (pp. 111–136). Basingstoke: Palgrave.
Pavlenko, A., & Blackledge, A. (2004). New theoretical approaches to the study of negotiation of
identities in multilingual contexts. In A. Pavlenko & A. Blackledge (eds.), Negotiation of iden-
tities in multilingual contexts (pp. 1–33). Clevedon: Multilingual Matters.
Pennycook, A. (2010). Language as a local practice. London: Routledge.
Pujolar, J. (2007). Bilingualism and the nation-state in the post-national era. In M. Heller (ed.),
Bilingualism: A social approach (pp. 71–95). Basingstoke: Palgrave.
Rampton, B. (1995). Crossing: Language and ethnicity among adolescents. London: Longman.
Rampton, B. (1998). Speech community. In J. Verschueren, J. Ostman, J. Blommaert, & C. Bulcan
(eds.), Handbook of pragmatics (pp. 1–34). Amsterdam: John Benjamins.
Rampton, B. (1999). Styling the other: Introduction. Journal of Sociolinguistics, 3(4), 421–427.
Rampton, B. (2006). Language in late modernity. Cambridge: Cambridge University Press.
Silverstein, M. (2003). Indexical order and the dialectics of sociolinguistic life. Language and
Communication, 23, 193–229.
Vertovec, S. (2006). The emergence of super-diversity in Britain (Centre on Migration, Policy and
Society, Working Paper No. 25). Oxford: University of Oxford.
Vertovec, S. (2007a). Super-diversity and its implications. Ethnic and Racial Studies, 30(6),
1024–1054.
Vertovec, S. (2007b). Introduction: New directions in the anthropology of migration and multicul-
turalism. Ethnic and Racial Studies, 30(6), 961–978.
Vertovec, S. (2009). Transnationalism. London: Routledge.
Vološinov, V. N. (1973). Marxism and the philosophy of language (Translation. L. Matejka & I. R.
Titunik, first published 1929). London: Seminar Press.