Linguistic Anthropology

Linguistic anthropology is a cover term for mainly Northern American approaches which contextualise language use in socio-cultural terms. According to Hymes (1964:xxiii)

    HYMES, Dell, 1964. Language in Culture and Society. A Reader in Linguistics and Anthropology. New York: Harper and Row. its scope may include problems that fall outside the active concern of linguistics, and always it uniquely includes the problem of integration with the rest of anthropology. In sum, linguistic anthropology can be defined as the study of language within the context of anthropology.

Linguistic anthropology's origins lie in the wider anthropological concerns with indigenous peoples, societies and cultures in the United States, Canada, Meso-America and also farther away. Although its ancestry is in what was initially a US government-funded programme of documentations and descriptions of (mainly) American Indian indigenous languages, myths and historical narratives, linguistic anthropology, in its present form, is the result of a "paradigmatic shift" established in the 1960s (see ethnography of speaking and interactional sociolinguistics). In Duranti's (2001:5) words,

    DURANTI, Alessandro, 2001. 'Linguistic Anthropology: history, ideas, and issues'. In: A. Duranti (ed.), Linguistic Anthropology: A Reader. Oxford: Blackwell. [...] linguistic anthropology as it is practiced today [...] is also more than grammatical description and historical reconstruction, and it is also more than the collection of texts, regardless of whether those texts were collected in one's office or under a tent. It is the understanding of the crucial role played by language (and other semiotic resources) in the constitution of society and its cultural representations.

Nowadays, many linguistic anthropologists have a double agenda:

  • a premium on ethnographic fieldwork and description among indigenous peoples which continue to provide credentials for academic community membership, but with a shift towards contemporary situations of contact (with governments, other communities, private companies, bureaucratic institutions, etc) focusing on the role of language in the formation of a communal identities, literacy projects, language rights movements, in a wage-labour economy, in struggles over economic resources, etc. As an illustration consider Collins' (1998:259) characterisation of his own work among the Tolowa in Northern California in terms of a double shift in perspective:
      COLLINS, James, 1998. 'Our ideologies and theirs.' In: B. Schieffelin, K. Woolard & P. Kroskrity (eds.), Language ideologies. Practice and Theory. Oxford: Oxford University Press, 256-270. to move away from "salvage linguistics" that documents for science another dying language, while tryng to understand what losing a language means for those who face that loss; to move away from a "salvage ethnography" that analyses memory culture, while trying to understand current social dynamics against the backdrop of long-announced and externally perceived cultural death.

    In quite a number of cases, this shift in perspective has foregrounded inequality, power relationships and (language) ideologies - also in the sense of raising issues of appropriation and entitlement in the contact situation between linguistic anthropologist and researched communities.

  • A commitment to the study of language use as situated - institutionalised - practice (for instance, A. Duranti's work on council meetings in Western Samoan villages) and often this is not restricted to "indigenous" contexts (for instance, E. Mertz's work on the contextualisation of legal precedents in US law school classroom discourse as providing critical moments in professional socialisation; another example would be S. Philips, who has published both on linguistic standards of evidentiality in US criminal courts and on judicial practices in Tonga). In this respect, it is true to say that much work in linguistic anthropology has a discourse analytical and/or a pragmatic orientation.

Finally, note that linguistic anthropology (including interactional sociolinguistics) is not to be confused with variationist sociolinguistics. Although at one point, Hymes and Gumperz included Labov's work on language change in the emerging discipline called "ethnography of communication" (using even "sociolinguistics" as a synonym), the two have clearly gone separate ways since the mid-80s. The variationist method is rather exclusively quantitative. It is positivist and tends not to be informed by anthropological theory (e.g. ethnography). Sociological variables such as class, gender, race, etc. tend to be treated as independent situational variables (rather than as culture-specific and situationally-contingent constructs).



Ethnography of speaking develops out of a wider appeal (in the mid 1960's) for "studies that would analyse in detail how language is deployed as a constitutive feature of the indigenous settings and events that constitute the social life of the societies of the world", as anthropological linguistics could "no longer be content with analysing language as an enscapsulated formal system that could be isolated from the rest of a society's culture and social organisation." (Duranti & Goodwin 1992a:1 DURANTI, Alessandro & GOODWIN, Charles, 1992a. 'Rethinking context: an introduction'. In: A. Duranti & C. Goodwin (eds.), 1994, Rethinking Context. Language as an Interactive Phenomenon. Cambridge: Cambridge University Press, 1-42. )).

First of all, some notes on ethnography. Ethnography is a principal mode of anthropological enquiry, but there is no unified conception of it. According to Alessandro Duranti (1997:84-5), ethnography is first of all a method. It offers a set of valuable techniques with allows researchers to connect linguistic forms with cultural practices. Its integration within other methods for the documentation of speech patterns sets linguistic anthropologists aside from other researchers into language and communication:

    DURANTI, Alessandro, 1997. Linguistic Anthropology. Cambridge University Press. As a first approximation, we can say that an ethnography is the written description of the social organisation, social activities, symbolic and material resources and interpretative practices characteristic of a particular group of people Such a description is typically produced by prolonged and direct participation in the social life of community and implies two apparently contradictory qualities: (i) an ability to step back and distance oneself from one's own immediate, culturally-biased reactions so to achieve an acceptable degree of "objectivity" and (ii) the propensity to achieve sufficient identification with or empathy for the members of the group in order to provide an insider's perspective."

However, ethnography can also be captured as a particular research ethos (in the sense that expertise in it is generally assumed to come with accumulated experience in fieldwork) with rather far-reaching theoretical implications (in the sense that it presupposes a particular epistemological orientation). One common tenet is undoubtedly the need for involved participation and distanced observation as starting points for all analysis. In Hymes' (1980 HYMES, Dell, 1980. 'What is ethnography?' In: Language and education: ethnolinguistic essays. Washington: Center for Applied Linguistics. ) view, etnography can be characterised as an interactive-adaptive method of enquiry. It

  • involves training in the accumulated comparative knowledge of a subject
  • is based on a desire to be comprehensive (i.e. characterised by a systematic interest in a wide range of related ways of life)
  • may well be topic-driven
  • prioritises comparative and constrastive analysis
  • must lead to hypothesis formation and ultimately generalisation
  • is open-ended and subject to self-correction during the enquiry

Note that Hymes links the latter with a general mission of anthropology, expressed in terms of a desire to "overcome the limitations of the categories and understandings of human life that are part of single civilisation's partial view" (1980:92 HYMES, Dell, 1980. 'What is ethnography?' In: Language and education: ethnolinguistic essays. Washington: Center for Applied Linguistics. ). Note that he also formulates the validity of ethnographic insights in terms of accurate knowledge about the meaning of particular behaviours, objects, institutions for those who participate in it. Ethnography, in Hymes' view, is essentially a participant-driven approach but it would be a mistake to assume that it is based on an illusory metaphor of pure induction: "[T]he more an ethnographer knows on entering the field, the better the result is likely to be" (1980:92 HYMES, Dell, 1980. 'What is ethnography?' In: Language and education: ethnolinguistic essays. Washington: Center for Applied Linguistics. ; compare with some versions of conversation analysis). Ethnography is perhaps best thought of as an epistemology which constantly moves between what is local/specific and the general, between knowledge already acquired and new data. Nor does ethnography exclude critical concerns. Consider in this respect also Hymes (1980:100):

    HYMES, Dell, 1980. 'What is ethnography?' In: Language and education: ethnolinguistic essays. Washington: Center for Applied Linguistics. It [Ethnography] is a mode of enquiry that carries with it a substantial content. Whatever one's focus of inquiry, as a matter of course, one takes into account the local form of general properties of social life - patterns of role and status, rights and duties, differential command of resources, transmitted values, environmental constraints. It locates the local situation in space, time, and kind, and discovers its particular forms and center of gravity, as it were, for the maintenance of social order and the satisfaction of expressive impulse.

 Ethnography stresses the necessity of knowledge that originates in participation, ordinary communication and observation. This provides a major point of discontinuity with many European traditions in discourse analysis - at least to the extent that the latter show a tendency to isolate textual material as "objects" for analysis, drawing reifying boundaries around it. Instead, ethnography values a careful treatment of context (the explicitation of context itself is an epistemological problem), insisting that it is impossible to separate speech data from the history under which it was obtained (see also natural histories of discourse). As Scheper-Hughes (2000:132) points out,

    SCHEPER-HUGHES, Nancy, 2000, 'Ire in Ireland'. Ethnography, 1:1, 117-140. the question often posed to anthropologist-ethnographers about the dangers of 'losing one's objectivity' in the field is really quite beside the point. Our task requires of us only a highly disciplined subjectivity".

 Ethnography stresses connections between sites and media of discourse, in ways which encourage a participant-oriented rather than a more narrowly text-oriented approach to "meaning". Consider in this respect Hymes (1980:95):

    HYMES, Dell, 1980. 'What is ethnography?' In: Language and education: ethnolinguistic essays. Washington: Center for Applied Linguistics. A key to the significance of a type of television programme may not be in the amount of time the family set is on, but in the family pattern of speaking around it. Is the set on, but ignored? Does someone insist on and get silence? Is the program essentially a resource for continuing conversation?

In a programmatic vein, ethnography - as it presupposes a dialogic situation of contact as its primary locus for research activity - also carries with it the potential for helping to overcome divisions of society into those "who know" and "those who are known". Because it presupposes an inversely-oriented pair of asymmetries (the ethnographer will be "one-up" on accumulated scientific expertise but will be "one-down" on insiderness - especially in the early stages, and, vice versa, the informant may acquire scientific expertise in the course of the research), ethnography is arguably better-suited for critical research with interventionist and emancipatory ambitions than critical paradigms which are based exclusively on a knowledge-based advantage of the researcher when it comes to determining foci, priorities and desirable goals (see also Slembrouck 2001 SLEMBROUCK, Stef, 2001. 'Explanation, intepretation and critique in the analysis of discourse'. Critique of Anthropology, 21:1, 33-57. ). The "learning" aspect also sets ethnography apart in terms of how one determines what needs to be researched. Agar (1996: 119-120, 126) elaborates the latter point in the general context of "hypothesis-testing":

    AGAR, Michael, 19962, The Professional Stranger. An Informal Introduction to Ethnography. San Diego: Academic Press. It's not necessarily that ethnographers don't want to test hypotheses. It's just that if they do, the variables and operationalisations and sample specifications must grow from an understanding of the group rather than from being hammered on top of it no no matter how poor they fit. [...] things like the learning role, the long-term intensive personal involvement and the holistic perspective are what set ethnography apart - they enable us to learn what people are like rather than seeing if a minute piece of their behaviour in a context we define supports or does not support our ideas of the way they are like."

Finally, ethnography raises issues of representation, in a way which problematises the relationship between "scientific" and "everyday" modes of representing categories, relationships, connections, etc. Note in this respect (Hymes 1980:98):

    HYMES, Dell, 1980. 'What is ethnography?' In: Language and education: ethnolinguistic essays. Washington: Center for Applied Linguistics. The general problem of social knowledge is two-edged: both to increase the accumulated structural knowledge of social life, moving from narrative to structurally precise accounts, as we have commonly understood the process of science, and to bring to light the ineradicable role of narrative accounts. Instead of thinking as narrative accounts as an early stage, we may need to think of them as a permanent stage, whose principles are little understood, and whose role may increase. [...] If narrative accounts have an ineradicable role, this need not be considered a flaw. The problem is not to try to eliminate them, but to discover how to assess them. [...] The question of narrative brings us to another aspect of ethnography. It is continuous with ordinary life.

Ethnography of speaking offers a radically descriptive orientation for the accumulation of data on the nature of ways of speaking within speech communities. Hymes' own formulation of the project is really a preliminary listing of fundamental notions and concepts that must be addressed within an adequate descriptive theory for sociolinguistic enquiry. As he puts it, "what is presented here is quite preliminary [...] one might call it 'toward toward a theory'. Some of it may survive the empirical and analytical work of the decade ahead." (Hymes, 1972:52 HYMES, Dell, 1972. 'Models of the Interaction of Language and Social Life'. In: J. Gumperz & D. Hymes, Directions in Sociolinguistics: the Ethnography of Communication. New York: Holt, Rinehart & Winston, 35-71. ). This looseness, I think, is best understood in terms of drawing attention to relevant concepts that were around at the time, stressing the potential connections between them, but also in terms of not pre-empting the outcomes of empirical and analytical work still to be undertaken. However, as a theoretical position, Hymes' project singles out diversity of speech as the hallmark of sociolinguistic enquiry (1972:39).

    HYMES, Dell, 1972. 'Models of the Interaction of Language and Social Life'. In: J. Gumperz & D. Hymes, Directions in Sociolinguistics: the Ethnography of Communication. New York: Holt, Rinehart & Winston, 35-71. Underlying the diversity of speech within communities and in the conduct of individuals are systematic relations, relations that, just as social and grammatical structure, can be the object of qualitative enquiry. A long-standing failure to recognise and act on this fact puts many now in the position of wishing to apply a basic science that does not yet exist.


    A general theory of the interaction of language and social life must encompass the multiple relations between linguistic means and social meaning. The relations within a particular community or personal repertoire are an empirical problem, calling for a mode of description that is jointly ethnographic and linguistic.

A brief and extremely cursory overview of the fundamental notions which Hymes lists:

  • Ways of speaking is used as the most general term. It is based on the idea that communicative conduct within a community entails determinate patterns of speech activity. The communicative competence of persons comprises knowledge with regard to such patterns.
  • The term fluent speaker draws attention to differences in ability, as well as the need to describe normative notions of ability. Different communities can be expected to hold differing ideals of speaking for different statuses, roles and situations (e.g. they may be based on memorisation, improvisation, quality of voice, etc.).
  • Speech community is a primary concept which postulates the unit of description as a social, rather than a linguistic entity. Rather than start with a "language", one starts with a social group and then begin to consider the entire organisation of linguistic means within it. A speech community is defined tautologically (but radically!) as a community which shares knowledge of rules for the conduct and interpretation of speech.
  • Speech situation are activities which are in some recognisable way bounded or integral. They may have verbal and non-verbal components. They may enter as contexts into statements of rules of speaking (e.g. as an aspect of setting - see below), but they are not in themselves governed by such rules throughout.
  • The term speech event is restricted to (aspects of) activities which are directly governed by rules or norms for the use of speech, with the speech act as the minimal term in the set - for instance, a party (speech situation), a conversation during the party (speech event), a joke within the conversation (speech act).
  • Sixteen components of speech:
    • setting
    • scene
    • key
    • speaker, sender
    • addressor
    • hearer, receiver, audience
    • addressee
    • channels
    • forms of speech
    • purposes - outcomes
    • purposes - goals
    • norms of interaction
    • norms of interpretation
    Act sequences
    • message form
    • message content
    • genre

    Note that Dell Hymes formulated the so-called SPEAKING-framework almost as a footnote, announcing it as a purely mnemonic code word, whose use may have little to do with an eventual theory or model. The grid refomulates the sixteen components, reducing them to the eight letters of the term "speaking".

  • Rules of speaking refer to the observation that shifts in any of the components of speaking may mark the presence of a rule, a structured relation (e.g. from normal tone to whisper, from formal English to slang, correction, praise, embarrassment, withdrawal, evaluative responses, etc.). Differences in the hierarchy of components are also an important part of the taxonomy of sociolinguistic systems.
  • Functions of speech may be statable in terms of relationships among components (e.g. in a given period or society, poetic function may require a particular relationship between choice of code, choice of topic and message form). At the same time, Hymes warns that the definition of function cannot be reduced to or derived directly from other components.



Ethnopoetics offers a particular method for analyzing and interpreting language use in contexts of oral (literary) performance. While the use of the term also signals a literary movement oriented to reading the poetries of "distant others", linguistic anthropological uses of the term revolve around a particular conception of narrative as primarily organized through formal poetic patterns (rather than through choices of content or thematic structure). Especially Hymes's (Hymes 1996 HYMES, Dell, 1996. Ethnography, linguistics, narrative inequality: Toward an understanding of voice. London: Taylor & Francis. ) variant of ethnopoetics is deeply influenced by Roman Jakobson's formulation of the poetic function in language use (Jakobson 1960 JAKOBSON, Roman.,1960, 'Closing statement: Linguistics and poetics'. In: Style in Language, (ed.) T. Sebeok, 1960, Cambridge, Ma: Harvard University Press, 350-377. ). The texture of an oral performance can be transcribed, organized and re-organized carefully to reflect a particular configuration of lines and stanzas. As Hymes (1996: 166 HYMES, Dell, 1996. Ethnography, linguistics, narrative inequality: Toward an understanding of voice. London: Taylor & Francis. ) notes, "the relations between lines and groups of lines are based on the general principle of poetic organization called equivalence" and "equivalence may involve any feature of language". The ethnopoetic method of reading thus invites close attention to similarity and contrast in prosody (e.g. stress, pausing, intonation), sound (e.g. rhyme), in syntax and grammar (e.g. tense markers), the use of particles and clause-connecting devices (contrastive or additive), etc. Units thus identified are held to combine into larger ones, called "verses" and "stanzas", again with patterns of equivalence being the main formal principle for their identification. Interestingly, ethnopoetics largely avails itself of strategies of "(re)writing" and "transcription" to come to terms with the poetic in everyday oral performance. It has been Hymes's ambition to raise ethnopoetic analysis to the level of identifying "emic" patterns which would be characteristic of particular cultures (a programme with Whorfian underpinnings).

In the ethnopoetic view, narrative is seen as a form of verbal action. The meanings it generates are effects of performance (Briggs and Bauman 1990 BAUMAN, Richard & BRIGGS, Charles, 1990. 'Poetics and performance as critical perspectives on language and social life'. Annual Review of Anthropology 19:59-88. ). Oral narratives, seen from this perspective, are transcribed and represented as a whole of lines and stanzas, and the patterning thus revealed counts as constitutive of narrative effect. Narrative content in this view (including a particular evaluative orientation) are considered to be effects of a narrative's formal organization: what there is to be told emerges out of how it is being told. Ethnopoetics' generic scope has included poetry, myths, narrative and folk tales, also with applications in 'forensic' enquiry (e.g. Maryns 2006 MARYNS, Katrijn, 2006. Language in the Belgian Asylum Procedure. Manchester: St-Jerome Publishing. , for whom ethnopoetic narrative analysis offers a strategy for the detection of meaning in asylum seekers' narratives of displacement as part of an institutional asylum procedure: the analyst's attention to poetic structure in the narratives provides a key to more accurate readings of the relationship between narration and narrated events). The latter detail is telling in that it highlights ethnopoetics' orientation to "unearthing" layers of meaning which would otherwise go unnoticed, either to the sociolinguistic analyst or the ordinary institutional language user: "ethnopoetics helps us to see more of what is there" (Hymes 1996: 182 HYMES, Dell, 1996. Ethnography, linguistics, narrative inequality: Toward an understanding of voice. London: Taylor & Francis. ). In this context, it is also worth referring to Hymes' (1996: chapter 8 HYMES, Dell, 1996. Ethnography, linguistics, narrative inequality: Toward an understanding of voice. London: Taylor & Francis. ) deployment of the ethnopoetic method in "second" analyses of narratives analyzed earlier by others as a way to engage with disagreements over point, purpose and evaluative stance. Compare also with the historicizing use of ethnopoetic analysis to restore and resurrect likely aspects of oral performance of narrations that have come to us in written notes or in translation; see also Blommaert's (2001 BLOMMAERT, Jan, 2001. 'Investigating narrative inequality: African asylum seekers' stories in Belgium'. Discourse & Society, 12, 413-449. ) reliance on ethnopoetic analysis as a strategy for laying bare meaning effects in contexts in which disadvantaged speakers compensate for a lack in language proficiency by resorting to explicit structuring devices.



The origins of the concept of indexicality lie in Charles Peirce's distinction between three types of sign-based meaning relationships. Unlike symbols (characterised by an arbitrary form/meaning-relationships) and icons (which reproduce some aspect of a referent), an index is characterised by an existential relationship with the referent (classical examples include: smoke indexes fire). The category of signs that function indexically can easily be extended to a range of linguistic expressions such as demonstrative pronouns (e.g. this, that, those), personal pronouns (e.g. I, you, we), temporal expressions (e.g. now, then, yesterday) and spatial expressions (e.g. up, down, below, here) - in short, deictic elements (sometimes called 'shifters') function indexically. The property of indexicality can be argued to extend to much of linguistic communication - as language is full of examples which are "existentially connected" to particular aspects of social and cultural context.

According to Michael Silverstein (1992:55), indexicality can be understood in spatial imagery:

    SILVERSTEIN, Michael, 1992. 'The indeterminacy of contextualisation: When is enough enough?' In: P. Auer & A. DiLuzio (eds.), The Contextualisation of Language, Amsterdam: John Benjamins. In a topological image, indexicality is by definition what I call a radial or polar-coordinate concept of semiotic relationship: indexical sign-vehicles point from an origin that is established in, by and at their occurring as the here-and-now "center" or tail, as it were, of a semiotic arrow. At the terminus of the radial, or arrowpoint, is their indexical object, no matter what the perceptual and conceptual dimensions or properties of things indexed. Strictly, by virtue of indexical semiosis, the "space" that surrounds the indexical sign-vehicle in unboundedly large (or small), characterisable in unboundedly many different ways, and its indexical establishment (as having-been-brought into being) almost limitlessly defeasible.

One strength of the concept of indexicality is undoubtedly that it draws attention to meaning-making processes which in terms of "existential" connections which users definitely make between sign and the social world which they inhabit but which are hard to pin down precisely in terms which allow firm generalisations. In this reading, speech act force assignment can be understood in indexical terms, i.e. as a matter of a situated establishing of a meaning relationship between a particular utterance and a social act that has come into existence (as different from saying: such and such an utterance have such and such a force in a context-independent way - without having to address the question "for whom?"). A similar analytical radicalisation can be claimed for applications of indexicality in the area of code-selection and switching. What is "existentially assumed" (inferred as "social facts") about a speaker when s/he selects a particular code or switches between codes (e.g. display of competence, display of group membership, oppositional alignment, situational conformity, etc.). In as much as it may be difficult for researchers to pin down what exactly is being indexed, it is unmistakably so that something very consequential is being indexed.

Duranti (2001:26 DURANTI, Alessandro, 2001. 'Linguistic Anthropology: history, ideas, and issues'. In: A. Duranti (ed.), Linguistic Anthropology: A Reader. Oxford: Blackwell. ) makes the point that Silverstein's concept of metapragmatic awareness provides a framework for thinking about the power of specific linguistic forms to reveal or to hide (from speakers' consciousness) their indexical value. This concept of metapragmatic awareness has its origin in theories of linguistic relativity and its underlying assumption about the unconscious nature of linguistic knowledge. In Silverstein's version, features of language structure can either favour or hinder native speakers' ability to interpret pragmatic value. One point to be made in passing here is that this enables a critique of the use of existing lexicalisations of speech acts for inventory of speech acts in a language (there may be speech act forces which cannot be lexicalised in a speech act force verb). But there is more. Linguists have tended to dismissive of native speakers' failure to distinguish between, say, gender as a part of linguistic structure, and, gender in the sense of particular objects/entities being associated with [+/- (fe)male] as part of a "natural order" in the "world out there". The latter has been seen as folk linguistic reasoning (and intrinsically uninteresting to the student of language - except to raise it in order to demystify it). In Silverstein's view, however, one is interested in how speakers make sense of language use and the world. So, rather than disqualify certain informant perceptions as "folk theory", the interesting question is: why do these perceptions occur and why do they show up in this area? It is indexicality which provides the pivotal concept here. In a view which accepts that cultural categories of experience may be extended to areas where no such linkage can be presupposed, the question how speakers orientate themselves (i.e. "where do speakers establish existentially-informed meaning relationships?") matters more than answering accurately the question "does the category at this point really apply?" - in the sense that an answer to the second question should be subsumed in an answer to the first question, rather than being primary.

One question which continues to hover around the profileration of applications of the concept is, of course, what are the limitations of the concept?



According to John Gumperz (Gumperz 1999:453-4), interactional sociolinguistics originated in criticism of earlier work in the etnography of communication which explained cultural diversity in terms of differences between bounded language-culture systems. Instead, interactional sociolinguistics has its origins in

    GUMPERZ, John, 1999. 'On interactional sociolinguistic method'. In: C. Roberts & S. Sarangi (eds.) Talk, Work and Institutional Order. Discourse in Medical, Mediation and Management Settings. Berlin: Mouton de Gruyter, 453-471. the search for replicable methods of qualitative sociolinguistic analysis that can provide insight in the linguistic and cultural diversity characteristic of today's communicative environments, and document its impact on individual's lives.

Gumperz' discourse analytic programme is sociolinguistically motivated: he "did not set out to study interaction in the abstract" (2003:105). In view of his early fieldwork in India and in Europe, the coinage "interactional sociolinguistics" is to be taken literally in positing a different picture of sociolinguistic variability, captured in an interactional sociolinguistic hypothesis which states that accumulated patterns and density of interpersonal contact, ideologies of interpersonal relations and normative principles of appropriacy form are intrinsic to explanations of processes of linguistic convergence, divergence and diversity and often turn out to be more significant than social categorisations.

    GUMPERZ, John, 2003. 'Response essay'. In: S. Eerdmans, C. Prevignano & P. Thibault (eds.), Language and Interaction: Discussions with John J. Gumperz. Amsterdam: John Benjamins, 105-126. Linguistic and cultural boundaries are not just 'naturally' there, they are communicatively and, therefore, socially constructed. Thus, they cannot be essentialised and treated as self-contained islands in research on communicative practices. Apart from interaction as such, ideology, power and history are all central to the way diversity works; depending on how these factors interrelate in specific circumstances, interaction can serve either to accentuate or attenuate the effects of diversity.

Over the years, Gumperz' has also developed a socio-cognitive perspective on interactional research, in early work by insisting on the ideological and constructed nature of communities and thematising the social productivity of discrepancies between the linguistic nature of boundary markers and the social import which these receive in users' perceptions - and, in later work, by concentrating on process of conversational inference in intercultural encounters, often in situations of gatekeeping. Correspondingly one can see a specific orientation to participant understandings in his work (Gumperz 2003: 116-7) oriented to practices of contextualisation,

    GUMPERZ, John, 2003. 'Response essay'. In: S. Eerdmans, C. Prevignano & P. Thibault (eds.), Language and Interaction: Discussions with John J. Gumperz. Amsterdam: John Benjamins, 105-126. a type of ethnography of communication that focuses on what people do in specific types of situations, how they react to others and how they talk about it [...] Basically, I regard myself as a linguistic anthropologist for whom the study of interaction is integral to his more comprehensive ethnographic investigations of the often taken-for-granted ways in which local populations deal with issues they encounter in going on about their affairs.

Thus, one part of his work engages very directly with cognition and inferential processes as depending on culturally-informed but situated inferential processes which play a role in talkers' interpretative constructions of the kind of activity or frame they are engaged in, of a speaker's intention, of what is required next, etc.

The concept of contextualisation is based on a reflexive notion of context, i.e. context is not just given as such in interaction, but it is something which is made available in the course of interaction and its construal depends on inferential practices in accordance with conventions which speakers or may not share. A crucial role in this is how talkers make available and act on so-called contextualisation cues, which John Gumperz (1999:461) defines as

    GUMPERZ, John, 1999. 'On interactional sociolinguistic method'. In: C. Roberts & S. Sarangi (eds.) Talk, Work and Institutional Order. Discourse in Medical, Mediation and Management Settings. Berlin: Mouton de Gruyter, 453-471. any verbal sign which when processed in co-occurrence with symbolic grammatical and lexical signs serves to construct the contextual ground for situated interpretations, and thereby affects how constituent messages are understood.

Typical contextualisation cues are code switching, style switching, prosodic choices, rhythm, particular lexical or syntactic choices, etc. Gumperz distinguishes between two levels of inferencing: (1) global inferences oriented to the "activity type" (what an exchange is about, what topics can be brought up, what should be conveyed (in)directly, etc.) and (2) local inferences oriented towards "preference organisation" (what is intended by one particular move and what is what is required by way of response). By focusing on regularities in intercultural and interethnic encounters (especially in institutional contexts of gatekeeping such as selection interviews), interactional sociolinguistics seeks to explain "unwarranted" institutional outcomes in terms of a speaker/listener's failure to recognise or respond to particular culturally-bound conventions of interpersonal communication - for instance, continuing with the example of selection interviews, how differences in discursive behaviour may have informed judgements of ability and how explanations for this have to be sought in, say, an Asian interviewee's failure to contextualise certain prosodic cues in the native English interviewer's questions but also the gatekeeper's failure to anticipate differential contextualisation practices (CLICK HERE for an example analysis). This way, interactional sociolinguistics also foregrounds how interpretations of talk are interactionally fed by and feed into larger "macro-communicative" orders. More recent versions of the approach also focus on the role of "ideologies of language" in inferential processes.

One of the major strengths of interactional sociolinguistics is its insistence on the occurrence of asymmetries in the communicative background of talkers. It cannot be taken for granted that speakers and hearers share the same inferential procedures or contextualise cues in the same way. Such sharing has to be demonstrated in analysis and one of the main aims indeed is to show how diversity affects interpretation. Compare (and contrast!) with the study of information exchange in analytical philosophical traditions.

A second strength is undoubtedly that, while interactional sociolinguistics takes on board the need to examine in detail the sequential positioning of turns at speaking, it does not share conversation analysis's restrictive concern with overt wording.

A third (and perhaps the most important) strength lies in its "pivotal" outlook - a concern with micro-processes in a way which can throw light on broader social processes and cultural issues - coupled with a dynamic conception of 'context' which recognises open-endedness and resists a neutralisation of a particpant's perspective. The latter is reflected in attendant linguistic anthropological developments of the concept of (re)contextualisation. As Silverstein (1992:75) notes:

    SILVERSTEIN, Michael, 1992. 'The indeterminacy of contextualisation: When is enough enough?' In: P. Auer & A. DiLuzio (eds.), The Contextualisation of Language, Amsterdam: John Benjamins. [...] the indeterminacies that emerge from a broadened and semiotically systematised understanding of 'contextualisation' [are] consistent with Gumperz' insightful recognition that indeterminacy is a shared dilemma of interactional textuality for both the analyst of and the participant in discursive interaction.

For Gumperz (1992:50 GUMPERZ, John, 1992. 'Contextualisation revisited'. In: P. Auer & A. Di Luzio, The Contextualisation of Language. Amsterdam: John Benjamins, 39-53. ), contextualisation cues function indexically (e.g. they share many of the characteristics of shifters; they foreground the discourse-level signalling processes which render human interpreters central to the process of interpretation); however, what sets contextualisation cues apart is that they are not necessarily lexically based (e.g. code-switching and prosody signal relational values quite independently of the 'propositional meaning' of particular lexical strings; CLICK HERE for an illustrative analysis).

One uneasy question which surrounds the interactional sociolinguistic framework is: when observing asymmetries, what is "institutional" and what is "cultural"? For instance, routine organisational or institutional 'knowhow' may be taken for granted by one party but remain unknown to the other.



Natural histories of discourse is a discourse-oriented perspective on the study of culture. In their editorial introduction to the volume with the same title, Michael Silverstein and Greg Urban (1996:2-3) link up the term 'natural histories of discourse' with a focus on

    SILVERSTEIN, Michael & URBAN, Greg, 1996. 'The natural history of discourse'. In: M. Silverstein & G. Urban (eds.), Natural Histories of Discourse. Chicago: The University of Chicago Press, 1-20. contextually contingent semiotic processes involved in achieving text - and culture. These are recoverable in some measure only by analytically engaging with textual sedimentations. Each chapter in our collective natural history thus focuses on certain analytic moments in the entextualising/co(n)textualizing process.

The perspective of "natural histories of discourse" takes issue with the anthropological idea of "text-as-culture" to the extent that the latter allows the analyst of culture to extract a portion of ongoing social action and draw a reifying boundary around it, before enquiring into its structure and meaning (cf. the idea that culture is embodied in a set of texts which are handed down from generation to generation). To equate culture with its resultant texts is to miss the point that the thingy-ness of texts is but one stage in ongoing cultural processes and although that thingy-ness may appear to have a quasi-permanent shape (which is by no means always the case), it travels from context to context and as a result, it will enter into new orderings between texts and be surrounded by changed conditions of commentary and explanation. Being sensitive to the natural histories of discourse invites attention to strategies and modes of entextualisation (e.g. transcription, copying, paraphrasing, citing, editing, recording in a new medium, etc.) and how these are constitutive of processes of (re)contextualisation (i.e. create a (new) context for). Further,

  • The perspective of a natural history of discourse raises the question whether there is at all such a thing as text in a durable sense. Consider Silverstein & Urban (1996:4):
      SILVERSTEIN, Michael & URBAN, Greg, 1996. 'The natural history of discourse'. In: M. Silverstein & G. Urban (eds.), Natural Histories of Discourse. Chicago: The University of Chicago Press, 1-20. The image of texts, and also of culture, derving from this latter insight is one a labile and mercurial insubstantiality, in which the text is figured by its always new and present co(n)text. It seems to lose touch with its past, the past, indeed, becoming a projection from the present. But is text entirely an illusion, like a three-dimensional projection from a two-dimensional image? Does textuality completely efface the traces of its antecedent co(n)texts of figuration?
  • In addition to focusing on, say, the recovery of the context in which a text artifact was produced or studying the recovery of an oral performance from text or practices of re-performance, an interest in natural histories of discourse also invites attention to the meta-discursive practices of institutions by focusing, among other things, on how recognised activities involve the insertion of textual artifacts into new contexts. For instance, text-artifacts may be treated as denotationally fixed and as devoid of social interaction (for instance, "reading aloud" exercises in classrooms which are orientated towards the technicalities of pronunciation - see Collins (1996 COLLINS, James, 1996. 'Socialisation to text: structure and contradiction in schooled literacy'. In: M. Silverstein & G. Urban (eds.) Natural Histories of Discourse. Chicago: University of Chicago Press, 203-228. )) or, text-artifacts may be presented as having emerged from social interaction and as strategically re-insertable into social interaction under changed contextual circumstances (for instance, in the reading practices advanced in law school setting, focusing on precedents and case histories as contextually-volatile textual artifacts - see Mertz 1996 MERTZ, Elizabeth, 1996. 'Recontextualisation as socialisation: text and pragmatics in the law school classrooms'. In: M. Silverstein & G. Urban (eds.), Natural Histories of Discourse. Chicago: University of Chicago Press, 229-249. ).
  • The historical perspective on the nature of discourse encourages analytical self-reflection because it draws attention to the role of transcription, recording and representation practices in the construction of social-scientific truths. Such a reflexive turn in analysis, at the same time, foregrounds the inevitable permutations of "data" against the background of shifting role relationships between researcher and researched. (Compare with conversational analysis and its "naive" belief in transcription as objective materiality.)
  • The notion of 'recontextualisation' can be usefully linked up with M. Foucault's concept of an 'order of discourse'. To the extent that people and documents constantly move between situations in an institutional context, 'recontextualisation' permeates all institutional activity. However, the NHD-perspective encourages a reflexive interpretation of this reformulation, i.e. the analysis, interpretation and explanation of institutional practice is in itself a history of x-number of recontextualisations.

[continue to contributions from NEW LITERACY STUDIES]