As Duranti  has noted (and the next paragraphs summarize his article), three paradigms have emerged over the history of the subdiscipline. The first paradigm, known as “linguistics” (later, “anthropological linguistics”), was devoted to themes unique to the subdiscipline—linguistic documentation of languages then seen as doomed to extinction (these were the languages of native North America on which the first members of the subdiscipline focused), grammatical description, typological classification, and the problem/question of linguistic relativity (associated with Edward Sapir and Benjamin Whorf but actually developed by Franz Boas and, before him, by a long line of European thinkers from Vico to Herder to Humboldt).
Dell Hymes was largely responsible for launching the second paradigm that fixed the name “linguistic anthropology” in the 1960s, though he also coined the term “ethnography of speaking” (or “ethnography of communication”) to describe the agenda he envisioned for the field. It would involve taking advantage of new developments in technology, including new forms of mechanical recording. Hymes also introduced a new unit of analysis. Whereas the first paradigm focused on ostensibly distinct “languages” (scare quotes indicate that contemporary linguistic anthropologists treat the concept of “a language” as an ideal construction covering up complexities within and “across” so-called linguistic boundaries), the unit of analysis in the second paradigm was new — the “speech event.” (The speech event is an event defined by the speech occurring in it — a lecture, for example — so that a dinner is not a speech event, but a speech situation, a situation in which speech may or may not occur.)
Hymes had hoped to link linguistic anthropology more closely with the mother discipline. The name certainly stresses that the primary identity is with anthropology, whereas anthropological linguistics conveys a sense that the primary identity of *its* practitioners was with linguistics, which is a separate academic discipline on most universities campuses today (not in the days of Boas and Sapir). However, Hymes’ ambition in a sense backfired; the second paradigm in fact marked a further distancing of the subdiscipline from the rest of anthropology.
In the current paradigm, which has emerged since the late 1980s, instead of continuing to pursue agendas that come from a discipline alien to anthropology, linguistic anthropologists have systematically addressed themselves to problems posed by the larger discipline of anthropology — but using linguistic data and methods.
So, for example, they investigate questions of sociocultural identity linguistically. Linguistic anthropologist Don Kulick has done this in relation to identity, for example, in a series of settings, first in a village called Gapun in Papua New Guinea. Kulick explored how the use of two languages with and around children in Gapun village — the traditional language (Taiap) not spoken anywhere but in their own village and thus primordially “indexical” of Gapuner identity, and Tok Pisin (the widely circulating official language of New Guinea). (Linguistic anthropologists use “indexical” to mean indicative, though some indexical signs create their indexical meanings on the fly, so to speak . To speak Taiap is associated with one identity — not only local but “Backward” and also an identity based on the display of *hed* (personal autonomy). To speak Tok Pisin is to index a Modern, Christian (Catholic) identity, based not on *hed* but on *save*, that is an identity linked with the will and the skill to cooperate. In later work , Kulick demonstrates that certain loud speech performances called *um escândalo*, Brazilian travesti (roughly, ‘transvestite’) prostitutes shame clients. The travesti community, the argument goes, ends up at least making a powerful attempt to transcend the shame the larger Brazilian public might try to foist off on them—again, through loud public discourse and other modes of performance.
Linguistic anthropologists Elinor Ochs and Bambi Schieffelin, in a remarkable series of studies  addressed the important anthropological topic of socialization (the process by which infants, children, and foreigners become members of a community, learning to participate in its culture), using linguistic as well as ethnographic methods. They discovered that the processes of enculturation and socialization do not occur apart from the process of language acquisition, but that children acquire language and culture together in what amounts to an integrated process. Ochs and Schieffelin demonstrated that baby talk is not universal, that the direction of adaptation (whether the child is made to adapt to the ongoing situation of speech around it or vice versa) was variable that correlated, for example, with the direction it was held vis-a-vis a caregiver’s body. In many societies caregivers hold a child facing outward so as to orient it to a network of kin whom it must learn to recognize early in life.
Ochs and Schieffelin demonstrated that members of all societies socialize children both *to* and *through* the use of language. Ochs and Taylor uncovered how, through naturally occurring stories told during dinners in white middle class households in southern California, both mothers and fathers participated in replicating male dominance (the “father knows best” syndrome) by the distribution of participant roles such as protagonist (often a child but sometimes mother and almost never the father) and “problematizer” (often the father, who raised uncomfortable questions or challenged the competence of the protagonist). When mothers collaborated with children to get their stories told they unwittingly set themselves up to be subject to this process.
Schieffelin’s more recent research  has uncovered the socializing role of pastors and other fairly new Bosavi converts in the Papua New Guinea community she studies. Pastors have introduced new ways of conveying knowledge— i.e. new linguistic epistemic markers —and new ways of speaking about time  opened the way, there has been an efflorescence of work done by linguistic anthropologists on the major anthropological theme of ideologies—in this case “linguistic ideologies,” sometimes defined as “shared bodies of commonsense notions about the nature of language in the world” . Silverstein  has demonstrated that these ideologies are not mere false consciousness but actually influence the evolution of linguistic structures, including the dropping of “thee” and “thou” from everyday English usage. Woolard , in her overview of “code switching,” or the systematic practice of alternating linguistic varieties within a conversation or even a single utterance, finds the underlying question anthropologists ask of the practice—Why do they do that?—reflects a dominant linguistic ideology. It is the ideology that people should “really” be monoglot and efficiently targeted toward referential clarity rather than diverting themselves with the messiness of multiple codes in play at a single time.
In a final example of this third paradigm, a group of linguistic anthropologists has done very creative work on the idea of social space. Duranti published a ground breaking article on Samoan greetings and their use and transformation of social space . Prior to that, Indonesianist Joseph Errington  — making use of earlier work by Indonesianists not necessarily concerned with language issues per se — brought linguistic anthropological methods (and semiotic theory) to bear on the notion of the “exemplarly center,” or the center of political and ritual power from which emanated exemplary behavior. Errington demonstrated how the Javanese *priyayi*, whose ancestors served at the Javanese royal courts, became emissaries, so to speak, long after those courts had ceased to exist, representing throughout Java the highest example of ‘refined speech.’ The work of Joel Kuipers further develops this theme vis-a-vis on the island of Sumba, Indonesia. And, even though it pertains to Tewa Indians in Arizona rather than Indonesians, Paul Kroskrity’s argument that speech forms originating in the Tewa *kiva* (or underground ceremonial space) forms the dominant model for all Tewa speech can be seen as a rather direct