Finding identity: Theory and data

MARY BUCHOLTZ and KIRA HALL

Abstract
This commentary responds to the papers in the special issue ‘Accomplishing
identity in bilingual interaction’ and particularly to the use of Bucholtz
and Hall’s (2004a, 2004b, 2005) framework for the linguistic analysis of
identities in interaction. The commentary focuses on the relationship be-
tween theory and empirical work, with attention to the role of ethnographic
context in the analysis of both microlevel interaction and macrolevel socio-
political and sociohistorical processes, the place of language ideologies in
the interactional construction of bilingual identities, and the necessity to
ground theoretical claims in rigorous empirical analysis.

Introduction
It is all too rare for researchers to have the opportunity to revisit their
past work and to see and respond directly to the ways that other scholars
have taken up their ideas, and so we are very grateful to the editors of
this special issue for their invitation to us to enter into dialog with the
contributors about the relationship between their work and our own. We
are still enjoying the novelty of having developed a framework that
others find useful enough to apply to their own data, and it is interesting,
often exciting, and sometimes surprising to see the varied ways that peo-
ple are engaging with the ideas we have put forward regarding the rela-
tionship between language and identity. In stimulating us to reflect on
what we had in mind when we developed the series of papers we have
published on this topic (Bucholtz and Hall 2004a, 2004b, 2005), the fore-
going articles have led us to think more deeply about a number of issues,
but perhaps most fundamentally about theorizing and its empirical un-
derpinnings, a relationship that lies at the heart of all our work on lan-
guage and identity. In our commentary, then, we focus not only on the
common themes that emerge from these papers but more generally on

Multilingua 27 (2008), 151⫺163 01678507/2008/027⫺0151

DOI 10.1515/MULTI.2008.008 쑕 Walter de Gruyter
152 Mary Bucholtz and Kira Hall

the issues they raise both individually and collectively regarding how to
find identity in interactional data and the role of theory in this process.
We are mindful of the many perils and challenges of conducting re-
search on identity, from treating it as an unanalyzed explanation for
social phenomena rather than an object of study in its own right, to
flattening the complexity of identity through ahistorical, decontextua-
lized, or overly deterministic analyses. We are therefore heartened to see
the authors of these articles taking seriously the many levels on which
identity operates: the interactional, the ethnographic, the historical, the
political.
We begin our discussion by highlighting the importance of simulta-
neously grounding the study of identity in the empirical evidence of in-
teraction and contextualizing the study of interaction through ethnogra-
phy. Such close attention to the workings of everyday life both in the
moment and over time is necessary in order to discover the specificity
and social meaning of local identity practices and categories as well as
the ways in which these are built up through the course of repeated
social actions and interactions. Conversely, this foundation in local prac-
tice allows researchers to understand how higher-level sociohistorical
and sociopolitical processes come to impinge on and be deployed by
social actors in the process of identity work. We demonstrate this point
by examining the central role of language ideologies of various kinds in
constructing identity in bilingual contexts, which emerges as a common
theme in all the articles. Finally, we reflect on the relationship between
theory and data in our own work and in this special issue. We argue for
the necessity of advancing theory within sociocultural linguistics while
keeping theories responsible to data, which must be the starting point
for any investigation of language in social life.

Identity in interaction
The contributors to this special issue share with us a belief that the social
world is constituted at the most basic level in and through interaction,
and their articles demonstrate a commitment to giving interaction the
serious analytic attention it requires in order to make empirically
grounded claims about precisely how this happens. Not surprisingly
given this emphasis, a number of the authors are inspired by conversa-
tion analysis and ethnomethodology, which provide rigorous tools for
analyzing the sequential structure of talk and the organization of mem-
bership categories. Moreover, as several of the articles demonstrate and
as we have argued in our own work, conversation analysis and other
forms of discourse analysis are often made more productive when com-
bined with ethnographic methodologies that allow researchers to make
 Finding identity: Theory and data 153

sense of locally specific elements of interaction as well as to get at the

sociocultural context in which interaction unfolds and to which it con-
tributes. Not only does ethnography support and extend the conversa-
tion-analytic commitment to understanding interaction from the point
of view of those who participate in it, but it also ensures that researchers
view talk not as a chunk of text removed from any broader context but
as a dynamic interactional process embedded in and inseparable from
the social and cultural world from which it emerges.
With respect to bilingualism, the sociolinguistic phenomenon that is
the focus of this special issue, a commitment to interaction promotes an
interpretation of code choice that is firmly rooted in the endogenous
concerns of participants rather than in the theoretical goals of the ana-
lyst. Thus several of the articles offer support for the recent reconceptu-
alization of codeswitching as only those instances of language alterna-
tion that are demonstrably salient within the interactional context ⫺ a
functional rather than structural definition (e.g., Gafaranga and Torras
2002). As for the sociocultural phenomenon of interest in this special
issue, identity, we are also pleased to see that the contributors do not
limit themselves to investigating the well-worn categories of social iden-
tity that are deeply familiar to sociolinguists ⫺ race and ethnicity, gen-
der, age, class, and so on ⫺ important as these continue to be. Instead,
they often move to a finer level of analytic granularity to consider such
fleeting yet consequential subject positions as bully and bullied, advice
giver and advice rejecter, interpreter and recipient of interpretation, posi-
tions that in turn may (or may not) sediment into something connected
to those higher-order categories.
The question then arises as to whether categories at these different
levels are all the same kinds of identities, and if not, what sort of analytic
work is necessary to move from one level to the other. Our own answer
to this question is that the notion of identity, fundamental as it is to
sociolinguistic analysis, may not, on its own, be sufficient to the task of
explaining the social dynamics of interaction, particularly in its tradi-
tional conceptualization as a set of enduring social categories such as
gender, race, and class. A finer degree of analytic delicacy is required,
and this is provided by the concept of stance. Stance is fundamental to
identity construction; as Ochs (1992) has shown, direct indexicalities that
link linguistic structures to interactional stances in turn form the basis
for indirect indexicalities between stances and the social groups that (are
thought to) typically take them. Hence moving too quickly from linguis-
tic structures to identity categories without considering how interac-
tional stances mediate between the two short-circuits the analysis and
diminishes much of the utility of considering identity in the first place.
As researchers, we need to start with what speakers are accomplishing
154 Mary Bucholtz and Kira Hall

interactionally and then build upward to the identities that thereby

emerge. At the same time, in order to ensure that our analyses are cogni-
zant of the rich intertextual layers that resonate between these different
levels, we need to ground our interactional analyses both in the ethno-
graphic specificities that endow interactions with social meaning and in
the broader social, cultural, and political contexts in which social actors
are imbricated. In short, neither identity categories nor interactional
analyses alone are enough to account for how social positioning is ac-
complished through language; the two levels of analysis are most effec-
tive when they work in unison, and in conjunction with a focus on the
larger social, cultural, and political contexts in which identity work is
carried out.

Finding identities and ideologies in bilingual interaction

Identity work is a highly politicized process in which social actors claim,
contest, and negotiate power and authority. When identities are forged
in relation to language, they become bound up with language ideologies,
which, as historically rooted and publicly articulated statements of cul-
tural belief about language and its users, mediate between the interac-
tional moment and broader sociopolitical structures. By the same token,
ideology mediates between the interactional stances taken by speakers
and the indexical relationships to identity that are thereby produced.
The contributors illustrate through their data and analyses the varied
and intricate ways in which ideology becomes interactionally alive for
participants in bilingual talk.
Ideology plays multiple roles in Tim Greer’s examination of bilingual
teenagers’ translations between Japanese and English at an international
school in Japan, which skillfully uses close interactional analysis to dem-
onstrate that such translations are designed to include all participants.
Greer’s argument thus convincingly debunks the widespread and highly
paranoid ideology held by many monolinguals that bilinguals use an-
other language in order to exclude them. But we would be wrong to
assume that inclusive talk of this kind lacks a power dimension. To get
at this dimension requires the tools of ethnography, for while the great
strength of conversation analysis is its ability to illuminate the taken-for-
granted world of the participants, this commonsense understanding
often requires considerable knowledge of local ideologies and practices.
A speaker’s work in positioning herself as linguistically competent and
another as a linguistic novice does not end at the interactional level;
rather, it enters a chain of ideologization involving locally specific beliefs
about who is and is not able to speak a language fluently. In Greer’s
data, this ideology plays out in the dynamics of who is targeted as a recipi-
 Finding identity: Theory and data 155

ent of translations (as it happens, the researcher himself is often positioned

in this way, on the basis of ideologies of language and nationality rather
than his actual fluency). In other words, although translation serves to in-
clude speakers in the interaction regardless of language ability, from an-
other perspective it serves to exclude or at least marginalize these same
speakers by singling them out as linguistically limited: adequation, the po-
sitioning of self and other as sufficiently similar, is often not far from dis-
tinction, or social differentiation. Thus accommodation through transla-
tion, while ensuring the inclusion of all participants, can also be under-
stood as partly hierarchical, while lack of accommodation may symboli-
cally signal inclusion by not foregrounding linguistic difference.
Another perspective on how translation ideologically creates interac-
tional inclusion through social distinction is found in Lisa Del Torto’s
investigation of intergenerational language brokering among Italian im-
migrant families in the United States and Canada. Del Torto rightly
points to the importance of temporality in identity production, noting
that identities do not emerge solely within a single interaction but rather
are forged through the accretion of interactional stances and position-
ings over time. She shows that in immigrant families, generational identi-
ties may also be linguistic identities, with members of the second genera-
tion inhabiting the role of language broker between the first and third
generations. Strikingly, this role, which has mostly been studied among
children in families that have recently immigrated, may endure through-
out the adult years, and it may be so integrated into the speaker’s habitus
that any indication of interactional trouble between the first and third
generations will trigger the second-generational speaker to step into the
brokering role. As Del Torto argues, this process is ideological as well
as interactional, for it is based on the second generation’s beliefs about
its own ‘in-between’ status as bilingual and bicultural relative to the
other two generations. The study offers a valuable reminder that while
the linguistic distinctiveness of immigrant generations may be an ideo-
logical construct as much as a sociolinguistic fact, it is one that allows
members of the second generation (and perhaps especially women) to
keep families unified across cultural and linguistic spaces.
The role of the language broker appears in very different form in
Holly Cashman’s sobering study of children’s teasing practices in a sec-
ond-grade bilingual classroom in the American Southwest, where it
shapes both the form and the content of the interaction. This dynamic
is seen most dramatically in the configuration of participation frame-
works ⫺ as when the English monolingual teacher, who inexplicably has
been charged with the education of a child who speaks no English, as-
signs two of his bilingual classmates to serve as linguistic mediators, and
in so doing enables them to torment him with impunity. It is therefore
156 Mary Bucholtz and Kira Hall

not surprising that the bilingual children conduct their attacks on the
new student entirely in Spanish ⫺ not only does their code choice ensure
that the target of their insults will understand them but it likewise en-
sures that the teacher will not. Further, the children’s grasp of exactly
what is at stake in this battle rises to the interactional surface when one
hurls the metalinguistic insult ‘You don’t even speak English.’ It is, in-
deed, precisely the new student’s Spanish monolingualism that makes
him such a vulnerable target. In addition, his Spanish too comes in for
negative evaluation, especially the use of güey, a highly ideologized in-
dexical feature of Mexican Spanish (cf. Bucholtz 2007). Through such
bullying practices it is paradoxically the targeted child who is positioned
as a ‘troublemaker’ by both peers and teacher alike. Cashman’s appeal
to the local ethnographic context of the community of practice as well
as the broader institutions and ideologies of language and nationality
that are implicated here starkly reveals how longstanding histories of
social inequality are reproduced through the cumulative force of social
interaction (cf. Goodwin 2002).
The ways that power and inequality manifest themselves in bilingual
interaction are also central to the article by Ashley Williams, whose
finely observed analysis of Cantonese speakers’ orientation to Mandarin
in San Francisco at a time when they have become outnumbered by
Mandarin speakers indicates that language ideologies are rarely mono-
tonic. In her data, speakers both valorize Mandarin as necessary and
easy to learn and condemn it as different and its speakers as annoying.
These language ideologies lead to remarkably complex stances of affilia-
tion and disaffiliation with Mandarin. On the one hand, ‘necessary’ and
‘easy’ do not inevitably signify entirely positive evaluations. Within the
United States, for example, there is an ideology that Spanish is both
easy to speak and increasingly necessary, as attested by Mock Spanish
on the one hand (Barrett 2006; Hill 1993), and on the other hand by the
existence of phrasebooks in Southern California to help well-to-do An-
glos talk to their household help (Schwartz 2006). In this situation, an
ideology of simplicity and necessity coexists harmoniously with more
explicitly negative and xenophobic attitudes. Likewise, ‘different’ does
not necessarily imply disaffiliation and may in some situations even be
a way of expressing admiration via exoticization, as with Americans’
evaluation of French as more cultured than English (Kinginger 2004;
Lippi-Green 1997) or of British English as superior to American English
(Jones 2001). Thus evaluations of Mandarin are tied to the interactional
goals speakers may have in expressing such views, and hence to stance
taking. As John W. Du Bois (2007) has discussed, a stance is not a dyadic
relationship between speaker and stance object (in this case, Mandarin)
but a triadic relationship through which a speaker simultaneously evalu-
 Finding identity: Theory and data 157

ates an object, positions herself, and aligns or disaligns with the ad-
dressee. The polyvalent stances the speakers take up are connected to
their alignments to other speakers and hence are less self-contradictory
than they may initially appear. In moving from interaction to broader
social processes, Williams’s use of the idea of ‘brought-along’ identities
is particularly helpful for thinking about the temporal and spatial dis-
placement of identity in immigrant communities, which helps shed light
on the ideological dynamics at work in this interaction.
Related to Williams’s concern with the ideologies that attach to spe-
cific languages, Helena Bani-Shoraka’s study of bilingual women’s al-
ternations between Persian and Azerbaijani in Tehran points to the ideo-
logical dimension of interaction, and how this is best uncovered by start-
ing sociolinguistic analysis not with identities but with interactional
moves, and particularly with the stances that speakers take in order to
accomplish their goals. As her data show, the meanings of Azerbaijani,
Persian, or Stylized Persian cannot be entirely given in advance but must
emerge within the discourse itself, informed both by the ethnographic
meanings of being an ethnically Azerbaijani bilingual in Iran and by each
speaker’s history of observing and using these codes. Such semiotically
complex linguistic phenomena do not bring the same indexical bundles
into every interaction; indeed, their value for speakers is their ability to
index different ideological configurations in different moments, thereby
requiring that the speaker’s projected meaning be arrived at through
negotiation among participants. Here ethnography can be put to use by
showing empirically what speakers know, and how they know, about the
ideologies underlying their bilingual practices.
The ideological ground on which bilingual interaction occurs is well
illustrated by Katherine Chen’s examination of ideologies about code-
switching in Hong Kong, which focuses on code choice not at the inter-
actional level but rather at the indexical level, as a semiotic marker of
social identity. Through a nuanced analysis of the distinctive language
alternation patterns of two different groups of Hong Kong young adults,
Chen shows that codeswitching is closely tied to ideologies regarding
English and how it is most appropriately integrated into Hong Kong
speech: those returnees who violate the local norm and use English more
extensively are viewed by locals as pretentious, an indication of the rela-
tive valorization of English vis-à-vis Cantonese. Chen’s findings regard-
ing the indexicality of language choice and its connection to social privi-
lege and hierarchy resonate with Hall’s current work on English as an
index of sexual modernity in India, where it is used to separate lesbian
elites from less highly educated transgender ‘boys’, who are bilingual but
prefer Hindi (Bucholtz and Hall under submission). It is certainly strik-
ing that, contrary to the usual expectation regarding linguistic accommo-
158 Mary Bucholtz and Kira Hall

dation, in Chen’s data it is the more privileged group ⫺ returnees, who

have had access to international schools and to overseas education ⫺
that must accommodate the less privileged group, the locals. This is not
a strategy of condescension (Bourdieu 1991) but a strategy of self-protec-
tion in order to avoid peer sanctioning. It is likely, then, that this indexi-
cal process shapes code alternation within interaction, whether through
metalinguistic commentary and explicit sanctioning of returnees’ code-
switching practices by locals or through less overt forms of orientation
to the markedness of extensive English use in the local context.
The theme of language ideology that runs through these articles gives
them added depth by reminding readers that identity does not operate
in isolation from other social dynamics. This focus on ideology also
helps move the analysis of identity work from the interactional level,
where it must begin, to the wider social world where it gains its meaning
and force.

Interaction and sociocultural linguistics

When we articulated our understanding of the relationship between in-
teraction and identity in our 2005 article, we characterized our approach
as sociocultural linguistics (see also Bucholtz and Hall under submission).
Part of our goal in advocating a richly sociocultural approach to linguis-
tic phenomena is to promote an understanding of interactional moves
as the building blocks of social life. To continue this architectural meta-
phor, although we are certainly interested in what these blocks are made
of and how they are put together, the ultimate goal must always be to
get to the edifice that they construct. In other words, interaction is a
valuable starting point for the analysis of sociolinguistic practice, but it
cannot be the end point. Interaction gives us the empirical grounding we
need to move from the moment-to-moment unfolding of social activities
within language to the larger social, cultural, political, and historical
world that we seek to explicate as scholars of culture and society. So
whereas in our 2005 paper we urged sociocultural linguists to think
small, in the sense that we need to attend to the micro details of interac-
tional and linguistic structure that serve as the semiotic resources for
identity work, we also want to emphasize the importance of thinking
big, in the sense that as analysts of the sociocultural realm of language
use, we need to link individual linguistic forms to interactional moves
and then to broader identities, social structures, cultural processes, and
ideologies. This is only possible if we have a deep cultural understanding
of local meanings and practices, which usually means that we need a
significant ethnographic component in our work. Crucially, we need to
give equal weight to each of these contexts, from the interactional to the
 Finding identity: Theory and data 159

ethnographic to the larger sociocultural realm beyond the face-to-face,

real-time, everyday activities that constitute our recorded data. In so
doing, we need to resist the urge to approach our data from the top
down, setting out to find interactional manifestations of social structures
or power asymmetries or identity categories, or for that matter, theoreti-
cal concepts such as adequation, distinction, authentication, denatural-
ization, authorization, and illegitimation. The danger of such an ap-
proach is that we will force our data to fit into the structures, categories,
and concepts available to us rather than discovering the ways in which
they do not fit and thus developing new structures, categories, and con-
cepts and refining or revising the existing ones.
What, then, is the value of ethnography, of identity, and of theory in
sociocultural linguistics? To answer the first part of this question, we
briefly review the sorts of functions that the papers in this issue propose
regarding particular instances of codeswitching. For Cashman, code
choice serves to separate and differentiate the self from a negatively
evaluated other: a new kid, a bossy kid, a bully. Thus the kinds of people
that individual kids are seen to be within the classroom is crucial infor-
mation for the analyst, and this sort of information can only be gained
through ethnography. Chen demonstrates another dimension of code-
switching: the indexical meaning of a particular pattern of language al-
ternation. Her attention to the ethnographic categories of local and re-
turnee allow her to discern the powerful social meanings of structural
regularities, which might have been overlooked in an approach that
failed to recognize and investigate the important differences between
these two groups. Greer and Del Torto show in different ways that
codeswitching to provide translation can be a way of repartitioning the
participation framework to ensure the inclusion of (presumed) linguistic
novices, and hence paradoxically to overcome linguistic difference by
foregrounding linguistic difference. Ethnography complements conver-
sation analysis by highlighting how locally established practices between
Japanese and non-Japanese or between immigrant generations shape any
given interaction between these categories. Conversely, Williams ob-
serves that a speaker’s use of the same code can give rise to different
sorts of stances, even from one utterance to the next. Here ethnography
can help us understand the complexity of speakers’ language ideologies,
based on the need to reposition their ‘brought-along’ identities as they
negotiate, over the course of everyday interactions, a new ethnic configu-
ration in which they are now a linguistic minority within a linguistic
minority. Finally, Bani-Shoraka reminds us that not only is codeswitch-
ing multifunctional, it may also at times be nonfunctional, with little
interactional or social significance for the participants (although it may
be extremely significant to the analyst interested in establishing patterns
160 Mary Bucholtz and Kira Hall

of language use). She notes the danger of exoticizing codeswitching and

bilingualism, cautioning against the analytic proclivity to find meaning
in every change of code. Ethnography can demonstrate why such analy-
ses are off base by moving beyond single interactions to show that the
code choices of speakers at a particular moment are situated in relation
to their habitual practice.
In sum, ethnography provides the temporal depth we need to be sure
that what we as analysts see in the data is how speakers as cultural
members understand it. It allows us to understand the categories to
which speakers orient. It allows us to know if what someone does in a
particular interaction is what they habitually do, or a radical departure,
and it allows us to gauge such affective subtleties as whether people are
joking or serious, angry or engaged, friendly or mocking. It also allows
us to know who people are, who their friends and enemies are, what
they care about, what they do every day, what they believe and think
and feel. The sort of interaction we analyze as sociocultural linguists is
often a single instantiation of broader social patterns, frozen in time; it
is at best a slice of life, not life itself. Ethnography helps us to flesh out
the picture.
An ethnographic approach also requires more of us as researchers
than do other methodologies: it demands that we put ourselves into the
picture and examine our own role in the research process. Whether the
ethnographer initiates and directs the interaction (Williams) or simply
contributes to it as one participant among others (Del Torto); whether
she is a member of the community under study, as with Chen and Bani-
Shoraka; whether, like Greer, the researcher unexpectedly becomes ‘the
data’ (cf. Dumas 2005) or even becomes unintentionally complicit in
larger structures of power and inequality, as Cashman courageously ac-
knowledges ⫺ in every case, a complete and ethical analysis will consider
how we ourselves are inextricable from the research we do. Ethnography,
then, helps keep researchers honest as well as humble.
Turning to the second part of our question, what is identity good for,
both for sociocultural linguists and for the people we study? That is,
with what goal do we undertake the investigation of identity construc-
tion through language use? Undoubtedly every researcher will answer
this question somewhat differently, but here are some suggested possibil-
ities: (1) to describe an identity that has been unrecognized or misrecog-
nized by researchers or cultural members, as we ourselves have done
with nerds in the United States (e.g., Bucholtz 1999) and hijras in India
(e.g., Hall and O’Donovan 1996), respectively, in our own work; (2) to
demonstrate the importance of a particular interactional resource for
identity work that has previously been overlooked, as in Greer’s discus-
sion of the importance of eye gaze in determining recipiency and hence
 Finding identity: Theory and data 161

the interactional goal of translations into Japanese or English; (3) to add

greater nuance to the conceptualization of identity and its construction,
for example, by detailing the instability of identity categories from mo-
ment to moment, as Williams’s work suggests, or the interactional speci-
ficity of these categories, as Bani-Shoraka’s and Cashman’s studies show
in different ways; or (4) to show how identities are tied up with larger
sociopolitical processes, institutions, histories, and ideologies that are
consequential beyond the interaction itself, as in both Del Torto’s and
Chen’s situated investigations of widely shared language ideologies.
Given this range of possibilities, and others as well, it is always worth
stepping back to reflect on why we are interested in examining identity
and how it helps us better understand our data.
Our final question, and one that all researchers of language, culture,
and society must often ask ourselves, is what, if anything, is gained by
using theory in empirical analysis. While the two of us have often been
inspired by theory ourselves, both within linguistics and in other fields,
we also want to sound a note of caution: our framework, like all concep-
tual models, is only as useful as the insights it yields. It was developed
as a way to incorporate under a single rubric a wide range of empirically
documented linguistic patterns that were often discussed separately; our
goal was less to set an agenda for future research than to summarize
and bring together the findings of diverse lines of research in order to
provide a foundation for investigations yet to come. As gratifying as it
is to see our work cited and applied to new kinds of data, we hope that
the concepts we have developed will be used primarily as tools or vehi-
cles that facilitate new sociocultural linguistic work, and will not be
viewed as the ultimate goal of such work. Just as we would not feel that
we had learned anything by identifying all the nouns and verbs in our
data, or all the first pair parts and second pair parts, likewise, as socio-
cultural linguists, we can learn very little about language, culture, and
society, by merely identifying identities or even the tactics of intersub-
jectivity that produce them. Yet the analytic process of doing so is a
crucial step in understanding how people use language to perform social
actions. Thus just as interaction is an important starting point for socio-
cultural linguistic analysis, theory is best appealed to later in the analytic
process, for it is our data, not our theories, that are our best guide to
how language is used to create local social worlds.

Conclusion
The articles in this issue clearly demonstrate that identity is an indispen-
sible concept for making sense of language as a fundamentally sociocul-
tural phenomenon, and that bilingual language use is an especially re-
162 Mary Bucholtz and Kira Hall

vealing and complex situation in which to investigate identity. Bringing

together the interactional, ethnographic, political, and historical contexts
in which bilingual speakers negotiate their own and others’ identities,
these authors offer a variety of ways to find identities in data, and to tie
these findings to larger theoretical questions of ideology, power, and
history. As sociocultural linguists continue to grapple with the question
of identity and when, how, and why it matters, the analytic toolkit will
be further expanded, providing new empirical bases for theory and new
directions for the study of language as a resource for the construction
of identity.
University of California, Santa Barbara
University of Colorado

References
Barrett, Rusty (2006). Language ideology and racial inequality: Competing functions
of Spanish in an Anglo-owned Mexican restaurant. Language in Society 35(2),
163⫺204.
Bourdieu, Pierre (1991). Language and Symbolic Power. Thompson, John B. (ed.),
Raymon, Gino and Matthew Adamson (trans.). Cambridge, MA: Harvard Uni-
versity Press.
Bucholtz, Mary (1999). ‘Why be normal?’: Language and identity practices in a com-
munity of nerd girls. Language in Society 28(2), 203⫺223.
⫺ (2007). Indexing gender and youth style through slang. In Gunnarsson, Britt-
Louise, Sonja Entzenberg, and Maria Ohlsson (eds.), Språk och kön i nutida och
historiskt perspektiv [‘Language and gender in present-day and historical perspec-
tive’]. Uppsala: University of Uppsala, 11⫺32.
Bucholtz, Mary and Kira Hall (under submission). All of the above: From paradigms
to coalitions in sociocultural linguistics. Journal of Sociolinguistics.
⫺ (2004a). Language and identity. In Duranti, Alessandro (ed.), A Companion to
Linguistic Anthropology. Oxford: Blackwell, 369⫺394.
⫺ (2004b). Theorizing identity in language and sexuality research. Language in Soci-
ety 33(4), 501⫺547.
⫺ (2005). Identity and interaction: A sociocultural linguistic approach. Discourse
Studies 7(4⫺5), 585⫺614.
Du Bois, John W. (2007). The stance triangle. In Englebretson, Robert (ed.), Stance
in Discourse: Subjectivity in Interaction. Amsterdam: John Benjamins.
Dumas, Nathaniel, (ed.) (2005). Plan B: When the Anthropologist Becomes Data. Spe-
cial issue of Kroeber Anthropological Society Papers 91.
Gafaranga, Joseph and Maria-Carme Torras (2002). Interactional otherness: Towards
a redefinition of code-switching. International Journal of Bilingualism 6(1), 1⫺22.
Goodwin, Marjorie Harness (2002). Exclusion in girls’ peer groups: Ethnographic
analysis of language practices on the playground. Human Development 45, 392⫺
415.
Hall, Kira and Veronica O’Donovan (1996). Shifting gender positions among Hindi-
speaking hijras. In Bergvall, Victoria L., Janet M. Bing, and Alice F. Freed (eds.),
Rethinking Language and Gender Research: Theory and Practice. London: Long-
man, 228⫺266.
 Finding identity: Theory and data 163
Hill, Jane H. (1993). Hasta la vista, baby: Anglo Spanish in the American Southwest.
Critique of Anthropology 13(2), 145⫺176.
Jones, Katharine W. (2001). Accent on Privilege: English Identities and Anglophilia in
the U.S. Philadelphia: Temple University Press.
Kinginger, Celeste (2004). Alice doesn’t live here anymore: Foreign language learning
and identity reconstruction. In Pavlenko, Aneta and Adrian Blackledge (eds.),
Negotiation of Identities in Multilingual Contexts. Clevedon: Multilingual Matters,
219⫺242.
Lippi-Green, Rosina (1997). English with an Accent: Language, Ideology, and Discrim-
ination in the United States. London/New York: Routledge.
Ochs, Elinor (1992). Indexing gender. In Duranti, Alessandro and Charles Goodwin
(eds.), Rethinking Context: Language as an Interactive Phenomenon. Cambridge:
Cambridge University Press, 335⫺358.
Schwartz, Adam (2006). The teaching and culture of household Spanish: Understand-
ing racist reproduction in ‘domestic’ discourse. Critical Discourse Studies 3(2),
107⫺121.