286 III.

Basics III: Basic Sociolinguistic Concepts

39. Linguistic Repertoire

1. Preliminary Remarks conception of linguistic repertoire is not itself

2. Conceptual Foundations a focus of research but rather a taken-for-
3. Repertoires, Social Structure and Language granted starting point for ethnographic work
Change
on multilingualism and language use in con-
4. Social Meaning in Code-Switching
5. Literature (selected)
text. Also, current work in this tradition pro-
vides a socially-sensitive approach to dis-
course analysis. Perhaps the best way to expli-
1. Preliminary Remarks cate the repertoire concept is to discuss how
it provided a solution to certain conceptual
The totality of linguistic resources available difficulties in dialectology and structural lin-
to members of a community for socially sig- guistics, and then to outline briefly two lines
nificant interaction constitutes the linguistic of research that have developed around the
repertoire of that community. The linguistic idea of linguistic repertoire: one of these ex-
resources include all the different languages, plores the relationship between repertoires,
dialects, registers, styles and routines spoken social structure and linguistic change, the
by the group as well as communicative codes other investigates code-switching to discover
in other modalities such as writing. The no- how social meaning is created in interaction.
tion of repertoire takes an interacting group
as the point of departure and asks: how do 2. Conceptual Foundations
speakers allocate their linguistic means to the
situations, events, and social relationships Leading linguists in the early part of this
recognized in the community (cf. art. 32, 81)? century — Meillet, Saussure, Vendryes, Sapir
The linguistic varieties in a repertoire are and Bloomfield among others — devoted im-
defined not by their origins or structural unity portant passages or chapters in their major
but by their differing uses or functions in the works to the detailed presentation of linguis-
social life of a particular group. For instance, tic diversity. Dialectology provided a wealth
in many multilingual communities the choice of evidence (cf. art. 64). But in these works
of one language over another has the same there was no systematic linkage between the
social function or significance as the selection insightful observations of social, stylistic and
among lexical variants in monolingual soci- geographic variation on the one hand and
eties. the theoretical discussion of linguistic struc-
When the notion of repertoire was first ture on the other. Nor was there any system-
introduced in the early 1960s it marked a atic explication of the way different kinds of
radical change in perspective away from social variation — regional, class or occupa-
single languages or dialects as the fundamen- tional — might be related to each other. As
tal units of analysis in linguistics. It inaugur- Weinreich, Labov and Herzog (1968) have
ated a turn towards the study of groups of argued, from the Neogrammarians through
speakers who have alternative ways of ex- Saussure and the American structuralists and
pressing themselves which are deployed in generativists, structure was equated with lin-
particular social contexts and for culturally guistic homogeneity. What is more, structure
constructed purposes. The concept of reper- was expected to exist within a genetically
toire focussed attention on structured linguis- defined language while lack of structure or
tic heterogeneity and on the social meaning "freedom" was expected to reign outside it,
of linguistic choice; it contributed signifi- in language use (Hymes 1974, 90 — 92). For
cantly to studies of language change and to structural linguists of the 1950s the idiolect
understanding the relationship between lan- became the proper object of description.
guage and social structure. Hockett's influential text defined it as a single
In the emergence of sociolinguistics as a person's speech habit at a given time. Ideally,
research tradition or paradigm, the notion many idiolects would make a dialect, several
of linguistic repertoire played a central role. dialects a language. But this formulation only
Since then it has competed with similar for- heightened the difficulties for linguists work-
mulations that have somewhat different ing in the tradition of dialectology in complex
scope and emphasis: ways of speaking, lin- societies. The speech of social units such as
guistic variables and domains. At present the regions, villages, castes and occupations ra-
39. Linguistic Repertoire 287

rely matched the set-inclusion relations hy- idea of repertoire extends structural princi-
pothesized of dialect and language (Fergu- ples to describe the patterned relationship
son/Gumperz 1960. Cf. art. 44). At the same between social context and language. Lin-
time, social and cultural anthropologists who guistic variation is not "free." Its pattern and
were looking to language differences as an structure can be discovered if the distribution
index of group boundaries also found that of linguistic variants is viewed in relation to
the linguist's definitions were unsafe guides, social groupings, social situations and com-
since the categories of structural linguistics municative purposes. "Within a socially de-
were not uniformly linked to social groups fined universe, forms are selected for study
(Hymes 1968). In the following decade the primarily in terms of who uses them and
emerging generativist revolution, while when, regardless of purely grammatical simi-
changing many of the theoretical commit- larities and differences." (Gumperz 1971,150)
ments of structuralism, left in tact the focus In studying repertoires rather than dialects
on homogenous language systems seen as or idiolects, one abstracts from speech and
autonomous with respect to social context. interaction along quite a different dimension
But the 1950s also saw important critiques and confronts a new set of research questions.
of the definition of language and dialect, that In short, the notion of linguistic repertoire
"most unsatisfactory terminological set" was proposed as a way of reconceptualizing
(Martinet 1954) and of dialectology itself the relationship between linguistic and social
(Weinreich 1954). These years witnessed a categories, between language and social life.
spurt of work on language variation and use In the study of linguistic repertoires it is
that is represented in an influential collection assumed that linguistic variation can occur
by Hymes (1964). Scholars were also dis- at any level of linguistic structure, involving
covering structure "between" systems, for in- choices among "referentially equivalent"
stance the work of Haugen and Weinreich on forms in phonology, syntax, morphology and
bilingualism. However, fundamentally new lexicon. In any social situation the choice of
ways of conceptualizing the relations between a variant in one part of an utterance limits
variation and context did not reach print the speaker in choices made later, in other
until the early 1960s. Linguistic repertoire parts. Such co-occurrence rules cut across the
was among these new models, all of which normal components of grammar. This is what
sought a unity between language and social makes it possible to divide the speech of even
life by focusing on social interaction and the monolingual speakers into varieties: co-oc-
use of language for communicative purposes. curring or co-varying sets of features or vari-
In a series of articles published throughout ants (Ervin-Tripp 1972). Notice that these co-
the 1960s, John J. Gumperz proposed, de- occurrence constraints become visible
fined and elaborated the concept of reper- through a methodology which insists on ethn-
toire. The influence of European and Indian ographic study of language use to ascertain
dialectology as well as the Prague school's what is situationally available to speakers.
interest in stylistics and standardization is The varieties are not defined by either the
evident in the development of the concept. linguist's or the speakers' named categories.
But equally important was Gumperz's close It becomes apparent that great linguistic dif-
scholarly and institutional link with anthro- ferences are sometimes socially trivial while
pology. Intellectually this included a continu- elsewhere trivial linguistic differences carry
ing concern with grand social theory as devel- great social consequences.
oped in anthropology as well as an ethno- A speaker's choice between varieties is also
graphic interest in interaction and cultural structured. It is systematically linked to social
knowledge. Developing in close conjunction relationships, events or situations. Again,
with the idea of repertoire were Hymes' no- ethnography is needed to identify the events
tion of 'ethnography of speaking' and the deemed separate and significant by the group
broader concept of 'communicative com- and the appropriate or expected choice of
petence' (cf. art. 34, 78) as well as Labov's a variety for the event. Repertoires may be
narrower concept of the 'linguistic variable' described for groups of varying sizes or even
(cf. art. 112). for individuals. Individuals in an interacting
Repertoire provided a new unit of analysis group need not share all varieties but the
that cross-cut the traditional units of lan- tacit conventions for using varieties and the
guage, dialect and idiolect by including social interpretive strategies associated with speak-
categories in a principled way. Indeed, the ing are expected to be known, if not necessa-
288 III. Basics III: Basic Sociolinguistic Concepts

rily used, by those who participate together language(s) of the subordinated, even in the
in significant interaction. ears of the subordinated themselves. Rather
Notice that the term repertoire, a metaphor than constructing typologies which are often
drawn from theater or performance, alludes implicitly based on evolutionary or moderni-
to a storehouse or treasury of possibilities. zation theory (e.g. Fishman 1978), much re-
The figure is significant. It constructed a dif- search in this vein relies on neo-marxist social
ferent view of speakers who were often so- theories and emphasizes historical changes in
cially subordinate and had been excluded or political economy within a single political or
marginalized by structuralist research inter- geographical region. It attempts to show the
ests: multilinguals and nonstandard speakers links between the historically changing distri-
of various kinds. In the progressive political bution of power and the evaluation and im-
climate of the late 1960s it provided an image position of linguistic forms. In a detailed and
of such speakers not as deviants but as people illuminating example G. Sankoff ([1976]
with positive, even remarkable skills. 1980) traced the change from egalitarian mul-
tilingualism in New Guinea to the use of
3. Repertoires, Social Structure and lingua francas and colonial languages. In the
Language Change pre-colonial era speakers preferred their own
local forms but local languages were not
In the many studies inspired by the repertoire ranked with respect to each other. With co-
concept, two directions emerge as notewor- lonial contact and urbanization speakers
thy. One starts with the correlation of types began to devalue their own language, accept-
of social structures with types of linguistic ing the authority of the lingua franca and the
repertoires. It is rooted in the important work colonial language. This both reflected and
of Ferguson (1959) defining diglossia (cf. art. spurred increasing social stratification.
33), and earlier research on dialects and stan- Another example is Hill's (1983) discussion
dardization. In three classic articles, Gum- of routes of language death among the Uto-
perz (1962, 1964, 1968, all reprinted in 1971) Aztecan languages. Hill argues that obsoles-
constructed a classification of speech com- cence occurred in contrasting ways depending
munities or sociolinguistic situations (cf. art. on how the particular group was linked to
32) based on the social function and diversity the world economic system.
of varieties in the community's repertoire. On a more microscopic social scale, the
This can be seen as fundamentally different internal structure of repertoires has also been
yet complementary to traditional linguistic of considerable research interest. In keeping
classification based on the genetic and typo- with previous work in dialectology, Gumperz
logical affiliations of languages. Gumperz distinguished between dialectal and super-
showed that differences in societal complexity posed varieties. Those linguistic forms linked
are reflected most directly in language atti- to particular social groups, often geographi-
tudes and in the shape and structure of lin- cally separate, are dialectal; those varieties
guistic repertoires. linked to activities or domains of use, such
However, a typology of speech communi- as the literary or ritual "high" forms in classic
ties can obscure the crucial historical pro- diglossia, are called superposed. In addition,
cesses linking social groups. Interesting cur- Gumperz suggested that two characteristics
rent work stresses the importance of seeing of repertoires are particularly sensitive to so-
language attitudes within a very broad polit- cial changes: the degree of compartmentaliza-
ical-economic framework. The commonest tion or separation of varieties and the linguis-
form of language change today is no longer tic distance between them.
the classic differentiation of languages as It is significant that in modern industrial
populations separate and disperse. Rather, in societies the dialectal/superposed distinction
our historical epoch colonial and capitalist typically collapses into a single dimension in
expansion, state formation and massive labor which the same linguistic features signal both
migration have brought populations together social status and the nature of the speech
into systems of political-economic domina- event (e.g. Labov 1972). In a parallel phe-
tion and cultural hegemony. The imposition nomenon students of bilingual communities
of standard and colonial languages is part of have observed that language shift occurs as
this process. Within a society the language the dominant group's language takes over
of the more powerful groups gains greater ever more of the functions of the indigenous
legitimacy, authority and prestige than the language. Often, new speech events and new
39. Linguistic Repertoire 289

functions are introduced, e.g., schooling, that highly "focused" linguistic norms de-
church, which can only be accomplished in velop under particular small-scale social con-
the dominant language, while events which ditions within widely differing societies.
are closely linked to the indigenous language The study of repertoire in contemporary
are forbidden or stigmatized (e.g. Eckert societies has revealed the importance of a
1980, Fishman 1964, Dressler/Wodak-Leo- "solidarity" dimension in the evaluation of
dolter 1977, Hill/Hill 1980). Then, if upward linguistic varieties, existing in addition to the
mobility is possible at all it involves rejection well-known "status" or prestige dimension.
of both the indigenous language and the "Languages of solidarity", be they working
events associated with it. Bourdieu (1977) class vernaculars or minority languages re-
sees the imposition of standard languages in duced in function, are often used by speakers
very much the same light, as a form of sym- to make economic claims on co-ethnics,
bolic domination exercised through control thereby forming part of a survival strategy
of the institutions of cultural reproduction, for beleaguered and powerless groups. At the
particularly the schools. Yet, even with re- same time, as Woolard (1985) pointed out,
stricted functions, stigmatized languages and "languages of solidarity" are best seen as
varieties persist even when mobility is pos- oppositional but contradictory products of a
sible only through the dominant language. capitalist era. They enact resistance to the
What is more, contrary to the implications dominant values and authority of the hege-
of Bourdieu's analysis, working class ver- monic public institutions in which the stan-
naculars have certainly not disappeared de- dard variety is required. But this opposition is
spite the enforcement of the standard in accomplished at the cost of creating linguistic
schools and public life (cf. art. 183). The forms which are then used by the larger
processes by which the "low" or stigmatized society to confirm the speakers' low status.
varieties of subordinated groups survive has Clearly this linguistic "resistance" should be
been a focus of debate. understood in light of broader theories of
By adopting the anthropological notion of cultural resistance and ethnic mobilization.
social network one can clarify the ways in Yet the societal emphasis on the binary
which different varieties are maintained or opposition between two codes is itself histor-
abandoned in the course of speakers' presen- ically situated. As an analytic construct it is
tation of their social identities in interaction. likely to be too simple a scheme to do justice
Gumperz (1964; 1982) noted that participa- to the many repertoires world-wide that in-
tion in varied networks gives speakers the clude three or more varieties. These codes are
knowledge, identity and right to switch be- often symbolically and structurally distinct
tween varieties in a single event. In a study but are not necessarily evaluated on a single
of an Austrian town, Gal (1979) described linear scale (see papers in Romaine 1982).
how language shift is occurring from Hun- Description of the full range of diversity in
garian to German by demonstrating that ma- repertoires, not simply different languages
crostructural changes such as industrializa- but the stylistic varieties within languages and
tion, urbanization and political centralization in monolingual repertoires, has also been
simultaneously changed the evaluation of the done under the rubric of ethnography of
local language and also disrupted social net- speaking (e.g. Bauman/Scherzer 1974. Cf.
works that had required members to show art. 78). Work in this direction includes the
solidarity through the use of the local variety. current concern with written as opposed to
The concept of social network helps in speci- spoken modalities and the range of genres
fying the process of language maintenance which link them (e. g. Tannen 1982).
and conceptually mediates between the In the studies mentioned above, the
individual level and macrosociological cat- emphasis has been more on language use and
egories such as class. Thus Milroy (1980) less on linguistic structure. But the locally
showed the mechanism by which working varying details of compartmentalization,
class speech is maintained in Belfast: closed, evaluation and social alternation among vari-
multiplex networks enforce the use of phonol- eties also affect their structure and thus the
ogical variants as well as constraining the language distance between them. This has
social meaning of those variants (see also become an important theme in the research
Labov 1973). Starting from the individual on language contact and change (Heath
speaker in a highly fluid society, LePage 1984). For example, Gumperz and Wilson
(1978) suggested the somewhat similar notion (1971) showed that the local versions of three
290 III. Basics III: Basic Sociolinguistic Concepts

genetically unrelated languages in northern 4. Social Meaning in Code-Switching

India converged grammatically while retain- (cf. 129)
ing separate lexicons. The results were at-
tributed to the daily alternation between these A second line of research which emerged
languages within a single repertoire over from the study of repertoires deals with the
many generation. Note that the languages way in which choice among linguistic vari-
were equal in status but strict separation of eties conveys social meanings. Code-switch-
the groups was required by the caste system. ing is at the center of this set of problems
The common assumption that lexicon is al- and has inspired an enormous literature (see
ways most susceptible to external influence review in Heller, in press). Two types of code-
while grammar is least so is clearly in need switching are generally distinguished. In
of rethinking (see also Scollon/Scollon 1979). "situational switching" societal norms of
It seems that the part of language that is choice link linguistic varieties to settings,
likely to change at contact is itself socially interactants, topics and other situational fac-
conditioned and not determined by linguistic tors. Alternation between varieties (also
structure alone. called codes) is predictable from independent
A related but conceptually separate issue information about the culturally defined
is the change that can occur in linguistic vari- components of the speech event. But, al-
eties when their everyday social uses are re- though some groups limit their switching to
stricted in specific ways. Again, the particular this sort, many other groups engage in the
local conventions for the allocation and eval- much more problematic "conversational" or
uation of the repertoire can be linked to struc- "metaphorical" code switching (Blom/Gum-
tural changes in the constituent varieties, but perz 1972, Gumperz 1982 a). In this type,
segments from both codes occur within a
this time the focus is not on co-existent lin-
single exchange in a single speech event, often
guistic systems but on the communicative re-
in a single sentence. This is not random mix-
quirements of speech events. Notice that two
ing, nor due to ignorance of the languages.
levels of social analysis are involved and
Rather it follows specifiable linguistic rules
linked, and both affect language: sociohistor-
and is done by those who are most fluent in
ical changes in the speech community, and both languages (Poplack 1980). Most impor-
the constraints of face-to-face interaction tantly, it is meaningful. Some claim that rapid
(Sankoff 1980, xix). For example, some vari- intrasentential switching is a separate mode
eties become socially restricted during lan- or variety in the repertoire and characterizes
guage shift since they are used for ever fewer particular groups. But most studies assume
communicative functions. Such cases provide that individual switches are themselves sig-
illuminating contrasts to pidginization. Re- nificant. While one cannot predict the occur-
cent work on socially restricted varieties has rence of conversational switches one can
explored a number of the resulting simplifica- learn to interpret them as the listeners do.
tions in morphology, phonology, subordina- The goal then is to explicate the cognitive
tion devices and stylistic markers (e.g., Do- and social processes by which speakers are
rian 1981; Gal 1984; Hill 1983; Dressier 1972; able to derive conventional meanings and
Trudgill 1976/1977). It is essential to consider social consequences from such switching.
the implications of these results for broader
grammatical theories. Given the failings of Two principles are crucial in this endeavor.
variable rules (Kay 1978) and the insistence in First, the sociolinguistic rules which link lin-
the generative paradigm for an autonomous guistic varieties and events are different from
grammar, the integration of these sociolin- ordinary linguistic rules which sort sentences
guistic insights with those of the generative into grammatical and not grammatical. When
approach remains problematic. Various dis- sociolinguistic rules are violated the result is
course perspectives on grammar seem more not simply an unacceptable or odd utterance.
promising (papers in Givon 1979). Clearly we Rather, as in the case of Grice's maxims,
need to understand considerably more about floutings of unmarked or expected language
the relationship between the social-historical choice are assumed to be intended and in
response to them listeners produce new inter-
and the social-interactional pressures on lan-
pretations of the speaker's meaning. This
guage as these interact with cognitive and
process of interpretation is at the center of
purely grammatical constraints (Slobin 1977).
the study of code-switching.
39. Linguistic Repertoire 291

A second and related principle is drawn ways of speaking: culturally preferred ways
in part from the interpretivist tradition in of managing interaction, presenting social
sociology. Rather than being determined by identity and attitudes, ordering and packag-
the social context in which it occurs, the ing information in conversation (e. g. papers
chosen forms of speech themselves may de- in Gumperz 1982 b). It may be, however, a
fine the speech event or social relationship. strategy which is unique in its use of macro-
For example, the switch from a formal pro- political divisions to make interactional
noun of address to an informal one is not meaning. It is separate from, but involves an
simply a reflection of a changed social organization of grammatical knowledge. In
relationship but is actually among the means the acquisition of code-switching skills, as in
by which that relationship has been changed. the acquisition of communicative competence
Similarly, if a linguistic routine or variety is generally, children learn not only how to talk
strongly associated with an activity, then to appropriately but also beliefs about what can
start the routine is to signal the start of the be said and known as well as categories and
activity. This is what Gumperz (1982 a) has evaluations of people and events. In short,
called contextualization cues. In this view so- they learn the culture inscribed in interaction
cial identities are in part enacted in speech (Ochs/Schieffelin 1984).
and thus cannot always be separately meas- It is noteworthy that the analysis of code
ured and later correlated with independent switching draws on a cognitive vocabulary
measures of speech. The same can be said derived from ethnoscience and generative
for situations since speech styles can define grammar and uses the critical notion of infer-
events (Coupland 1980). ence which it shares with philosophical ap-
Many functions of conversational code proaches to pragmatics. What is more, the
switching (CCS) have been identified. To general goal is to understand how an utter-
interpret CCS one must first discover the ance gains the meaning it does in particular
local cultural opposition between values, eth- social and cultural contexts, a goal it shares
nic identities and activities which is symbol- with many versions of discourse analysis, for
ized by the two codes. Such cultural under- which it offers particularly illuminating and
standings are ordinarily shared only as a re- complicating evidence.
sult of frequent interaction and shared group
membership, so CCS is unlikely to occur
among strangers. Armed with background 5. L i t e r a t u r e (selected)
knowledge, interactants may juxtapose the Bauman, Richard/Scherzer, Joel, eds., (1974) Ex-
codes to accomplish metaphorical and there- plorations in the ethnography of speaking, New
fore indirect, ambiguous or "off-record" ex- York.
pressions of a variety of conversational acts:
Blom, Jan-Petter/Gumperz, John J., (1972) "Social
mitigating or aggravating requests, denials, meaning in linguistic structures: Code-switching
topic shifts, elaborations, validations or clari- in Norway", in: Gumperz, J. J./Hymes, D., eds.,
fications. These discourse or stylistic effects Directions in sociolinguistics, New York.
depend on the listener's inference that the Bourdieu, Pierre (1977) "The economics of linguis-
speaker is serious, angry, deferent, distant, tic exchanges," in: Social Science Information 16
solidary, etc. And these inferences in turn (6), 6 4 5 - 6 8 .
depend on participants' knowledge of the de-
Coupland, Nikolas (1980) "Style shifting in a Car-
tails of the social context, particularly what diff work setting," in: Language in Society 9 , 1 —
choices are marked, or unmarked, in particu- 12.
lar social relationships. Thus code choices
Dorian, Nancy (1981) Language death, Philadel-
simultaneously manage (or change) the con-
phia.
versational task and the social relationship
(Scotton 1976, Heller in press) as well as the Dressier, Wolfgang/Wodak-Leodolter, Ruth, eds.,
(1977) "Language death," in: Linguistics 191.
speaker's claims to identity.
What is important here is not so much the Dressier, Wolfgang (1972) "On the phonology of
many functions that CCS has been shown to language death," in: Chicago Linguistics Society 8,
448-57.
accomplish but rather the generalization that
it always involves culturally specific expecta- Eckert, Penelope (1980) "Diglossia: Separate and
tions about speaking and about norms (or unequal," in: Linguistics 18, 1053 — 1064.
competing norms) of interpretation. Finally, Ervin-Tripp, Susan (1972) "On sociolinguistic ru-
it is but one of many discourse strategies or les: Alternation and co-occurrence," in: Gumperz,
292 III. Basics III: Basic Sociolinguistic Concepts

J. J./Hymes, D., eds., Directions in sociolinguistics, Kloss, Heinz (1967) Grundfragen der Ethnopolitik
New York, 2 1 3 - 2 5 0 . im 20. Jahrhundert, Wien.
Ferguson, Charles (1959) "Diglossia," in: Word 15, Labov, William (1972) Sociolinguistic patterns,
325-40. Philadelphia.
Ferguson, Charles/Gumperz, John J. (1960) "In- Labov, William (1973) "On the linguistic conse-
troduction," in: Ferguson, Ch./Gumperz, J. J., quences of being a lame," in: Language in Society
eds., Linguistic Diversity in South Asia, Bloom- 2, 8 1 - 1 1 5 .
ington. LePage, Robert (1978) "Projection, focusing and
Fishman, Joshua (1964) "Language maintenance diffusion," in: York Working Paper in Linguistics
and language shift as fields of inquiry," in: Linguis- 9, University of York.
tics 9, 3 2 - 7 0 . Martinet, André (1954) "Dialect," in: Romance
Fishman, Joshua, ed., (1978) Advances in the study Philology 8, 1 - 1 1 .
of societal multilingualism, The Hague. Milroy, Lesley (1980) Language and social net-
Gal, Susan (1979) Language shift, New York. works, Baltimore.
Gal, Susan (1984) "Phonological style in bilingual- Ochs, Elinor/Schieffelin, Bambi (1984) "Language
ism: The interaction of structure and use." in: acquisition and socialization: Three developmental
Schiffrin, D., ed., Meaning form and use in context, stories and their implications," in: Schweder, A./
Washington, DC. LeVine, R., eds., Culture Theory, New York,
GivonjTalmy, ed., Syntax and semantics 12: Dis- 276-322.
course and syntax, New York. Poplack, Shana (1980) "Sometimes I'll start a sen-
Gumperz, John J. (1971) Language in social groups, tence in English y termino en Espagnol: Towards
Standford. a typology of code-switching," in: Linguistics 18,
581-618.
Gumperz, John J. (1982 a) Discourse strategies,
New York. Romaine, Suzanne, ed., (1982) Sociolinguistic varia-
tion in speech communities, London.
Gumperz, John J., ed., (1982 b) Language and social
identity, New York. Sankoff, Gillian (1980) The social life of language,
Philadelphia.
Gumperz, John J./Wilson, Robert (1971) "Conver-
gence and creolization: A case from the Indo-Ar- Scollon, Suzanne/Scollon, Ronald (1979) Linguistic
yan/Dravidian border in India," in: Hymes, D., convergence, New York.
ed., Pidginization and creolization of languages, Scotton, Carol (1976) "Strategies of neutrality:
London. Language choice in uncertain situations," in: Lan-
Heath, Jeffrey G. (1984) "Language contact and guage 52, 9 1 9 - 9 4 1 .
language change," in: Annual Review of Anthro- Slobin, Dan I. (1977) "Language change in child-
pology 13, 3 6 7 - 8 4 . hood and history," in: Macnamara, J., ed., Lan-
Heller, Monica, ed., (in press) Code-switching: An- guage learning and thought, New York.
thropological and linguistic perspectives, Austin. Tannen, Deborah, ed., (1982) Spoken and written
Hill, Jane (1983) "Language death in Uto-Azte- language. Norwood.
can," in: International Journal of American Linguis- Trudgill, Peter (1967/1977) "Creolization in re-
tics 49 (3), 2 5 8 - 7 6 . verse: Reduction and simplification in the Albanian
Hill, Jane/Hill, Kenneth, (1980) "Mixed grammar, dialect of Greece," in: Transactions of the Philologi-
purist grammar and language attitudes in Nahu- cal Society, 3 2 - 5 0 .
atl," in: Language in Society 9, 321—48. Weinreich, Uriel (1954) "Is a structural dialectol-
Hymes, Dell, ed., (1964) Language in culture and ogy possible?" in: Word 14, 3 8 8 - 4 0 0 .
society, New York. Weinreich, Uriel/Labov, William/Herzog, Marvin
Hymes, Dell (1968) "Linguistic problems in defin- (1968) "Empirical foundations for a theory of lan-
ing the concept of tribe," in: Helm, J., ed., Essays guage change." in: Lehmann, W./Malkiel, Y., eds.,
on the problem of tribe, Seattle. Directions in historical linguistics, Austin.
Hymes, Dell (1974) Foundations in sociolinguistics, Woolard, Kathryn A. (1985) "Sociolinguistic varia-
Philadelphia. tion and cultural hegemony: Toward an integration
Kay, Paul (1978) "Variable rules, community gram- of sociolinguistic and social theory." in: American
mar and linguistic change." in: Sankoff, D., ed., Ethnologist 12, 738-748.
Linguistic Variation: Models and Methods, New
York. Susan Gal, New Brunswick, N.J. (USA)