Linguistic Anthropology statistics n process. Hyndman R J, Bashtannyk D M, Grunwald GK 1996 Es- There hay alizations Computational. and Graphical Statistics 5: 315-36 lynomial Koenker ko, Bassett G 1978 regression quantiles. Econometrica to multivariate response models is a par- Koenker R. Hallock K 2001 Quantile Regression. Journal of Economic, Perspectives, forthcoming important challenge Lehmann E 1974 Nonparametrics: Statistical Methods Based on Ranks. Holden-Day, San franci Manski C 1985 Semipara analysis of discret 7. Conclusion score estimator Classical least squares regression may be viewed as a Portnoy S, Koenker R 1997 The Gaussian hare and the natural way of extending the idea of estimating an Laplacian tortoise: computability of squared-error vs unconditional mean parameter to the problem of absolute-error estimators. with discussion Statistical science step is the formulation of an optimization problem Powell JL! 986 Censored regression quantiles. Journal of that encompasses both problems. Likewise, quantile egression offers an extension of univariate quantile RKoenker estimation to estimation of conditional quantile func- tions via an optimization of a piecewise linear objective function in the residuals. Median regression minimizes the sum of absolute residuals, an idea introduced by Boscovich in the eighteenth century The asymptotic theory of quantile regression closely Linguistic Anthropology parallels the theory of the univariate sample quantiles; computation of quantile regression estimators may be ' Linguistic anthropology'is an interdisciplinary field formulated as a linear programming problem and dedicated to the study of language as a cultural fficiently solved by simplex or barrier methods. A resource and speaking as a cultural practice. It assumes close link to rank based inference has been forged that the human language faculty is a cognitive and a from the theory of the dual regression quantile process, social achievement that provides the intellectual tools or regression rankscore process for thinking and acting in the world. Its study must be gression are provided by Buchinsky (1998)and as they engage in daily social activities. This docu- Koenker and Hallock(2001). Most of the major mentation relies on participant observation and other statistical computing languages now include some methods, including audiovisual recording,annotated capabilities for quantile regression estimation and transcription, and interviews with participants inference Quantile regression packages are available As an interdisciplinary field, linguistic anthropology forrandSplusfromtheRarchivesathttp://hasoftendrawnfromandparticipatedinthede lib.stat.cmu.edu/r/cranandStatlibathttp://velopmentofothertheoreticalparadigmsSomeofits lib.stat.cmu.edu/s, respectively. Stata's central core own history is reflected in the oscillation often found provides quantile regression estimation and inference among a number of terms that are not alway ctions. SAs offers some, rather limited facilities ology, anthropological for quantile regression. linguistics, ethnolinguistics, and sociolinguistics. Its main areas of interest have changed over the years, from an almost exclusive interest in the documentation Bibliography of the grammars of aboriginal languages to the analysis of the uses of talk in everyday interaction and Barrodale,Roberts FD K 1974 Solution of an overdetermined. throughout the life span(Duranti 1997, Foley 1997) This article provides a brief historical account of Buchinsky M 1998 Recent advances in quantile regression linguistic anthropology, and highlights important past for empirical research. Journal of and present issues, theories, and methods Human Resources 33: 88-126 Doksum K 1974 Empirical probability plots and statistical inference for nonlinear models in the two sample case. Annals 1. Linguistic Anthropology within the boasian Gutenbrunner C, Jureckova J 1992 Regression quantile and Tradition in the linear model and derived tatistics. Annals of Statistics 20: 305-30 In the holistic tradition established by Franz boa Hajek J, Sidak Z 1967 Theory of Rank Tests. Academia, Prag 858-1942)in the USa at the beginning of the Czech Republic twentieth century, anthropology was conceived as
statistics based on the full quantile regression process. There have been several proposals dealing with generalizations of quantile regression to nonparametric response functions involving both local polynomial methods and splines. Extension of quantile regression methods to multivariate response models is a particularly important challenge. 7. Conclusion Classical least squares regression may be viewed as a natural way of extending the idea of estimating an unconditional mean parameter to the problem of estimating conditional mean ‘functions,’ the crucial step is the formulation of an optimization problem that encompasses both problems. Likewise, quantile regression offers an extension of univariate quantile estimation to estimation of conditional quantile functions via an optimization of a piecewise linear objective function in the residuals. Median regression minimizes the sum of absolute residuals, an idea introduced by Boscovich in the eighteenth century. The asymptotic theory of quantile regression closely parallels the theory of the univariate sample quantiles; computation of quantile regression estimators may be formulated as a linear programming problem and efficiently solved by simplex or barrier methods. A close link to rank based inference has been forged from the theory of the dual regression quantile process, or regression rankscore process. Recent nontechnical introductions to quantile regression are provided by Buchinsky (1998) and Koenker and Hallock (2001). Most of the major statistical computing languages now include some capabilities for quantile regression estimation and inference. Quantile regression packages are available for R and Splus from the R archives at http:}} lib.stat.cmu.edu}R}CRAN and Statlib at http:}} lib.stat.cmu.edu}S, respectively. Stata’s central core provides quantile regression estimation and inference functions. SAS offers some, rather limited, facilities for quantile regression. Bibliography Barrodale I, Roberts F D K 1974 Solution of an overdetermined system of equations in the l " norm. Communications ACM 17: 319–20 Buchinsky M 1998 Recent advances in quantile regression models: a practical guide for empirical research. Journal of Human Resources 33: 88–126 Doksum K 1974 Empirical probability plots and statistical inference for nonlinear models in the two sample case. Annals of Statistics 2: 267–77 Gutenbrunner C, Jurec)kova! J 1992 Regression quantile and regression rank score process in the linear model and derived statistics. Annals of Statistics 20: 305–30 Ha!jek J, Sida!k Z 1967 Theory of Rank Tests. Academia, Prague, Czech Republic Hyndman R J, Bashtannyk D M, Grunwald G K 1996 Estimating and visualizing conditional densities. Journal of Computational. and Graphical Statistics 5: 315–36 Koenker R, Bassett G 1978 regression quantiles. Econometrica 46: 33–50 Koenker R, Hallock K 2001 Quantile Regression. Journal of Economic, Perspecties, forthcoming Lehmann E 1974 Nonparametrics: Statistical Methods Based on Ranks. Holden-Day, San Francisco Manski C 1985 Semiparametric analysis of discrete response: asymptotic properties of the maximum score estimator. Journal of Econometrics 27: 313–34 Portnoy S, Koenker R 1997 The Gaussian hare and the Laplacian tortoise: computability of squared-error vs. absolute-error estimators, with discussion. Statistical Science 12: 279–300 Powell J L 1986 Censored regression quantiles. Journal of Econometrics 32: 143–55 R. Koenker Linguistic Anthropology ‘Linguistic anthropology’ is an interdisciplinary field dedicated to the study of language as a cultural resource and speaking as a cultural practice. It assumes that the human language faculty is a cognitive and a social achievement that provides the intellectual tools for thinking and acting in the world. Its study must be done by detailed documentation of what speakers say as they engage in daily social activities. This documentation relies on participant observation and other methods, including audiovisual recording, annotated transcription, and interviews with participants. As an interdisciplinary field, linguistic anthropology has often drawn from and participated in the development of other theoretical paradigms. Some of its own history is reflected in the oscillation often found among a number of terms that are not always synonyms: linguistic anthropology, anthropological linguistics, ethnolinguistics, and sociolinguistics. Its main areas of interest have changed over the years, from an almost exclusive interest in the documentation of the grammars of aboriginal languages to the analysis of the uses of talk in everyday interaction and throughout the life span (Duranti 1997, Foley 1997). This article provides a brief historical account of linguistic anthropology, and highlights important past and present issues, theories, and methods. 1. Linguistic Anthropology within the Boasian Tradition In the holistic tradition established by Franz Boas (1858–1942) in the USA at the beginning of the twentieth century, anthropology was conceived as 8899 Linguistic Anthropology
Linguistic Anthi comprising four subfields: archaeology, physical (now later became known as'salvage anthropology, that is biological,)anthropology, linguistics(now linguistic the documentation of languages anthropology), and ethnology (now sociocultural ditions that seemed on the verge of disappearing. This nthropology ). This vision of anthropology differs enterprise-a struggle against time due to the great from the one found in the European tradition, where damage already done, by the end of the nineteenth inguistics and social anthropology remained rigidly century, to the indigenoroduced valuable information separate disciplines for most of the twentieth century, despite the emphasis on the use of native languages in on Native American traditions, but it had its method ieldwork among UK anthropologists, and the theor- ological and theoretical drawbacks, the most flagran etical and methodological influence of bronislaw of which was the inability to see or accept the Malinowski (1884-1942), who wrote about the im- cultural contact and colonization portance of linguistic research for an anthropological Through his writing and teaching, Boas brought understanding of human societies. In the 1950s, the scientific rigor to linguistic description and helped adoption of demolish a number of unfounded stereotypes about European preference for ' ethnology'over (cultural) the languages that were then called 'primitive. In an anthropology)for those studies that merged linguistic 1889 article entitled On alternating sounds boas and anthropological interests signaled the intellectual argued that the commonly held view that speakers of recognition, at least in some European academic American Indian languages were less accurate in their circles, of the importance of an 'ethnological side' of renunciation than ers of Indo-Euror istic studies( Cardona 1976), but the institutional languages was false and probably due to the lack of recognition of such a discipline within European linguistic sophistication of those who had first tried nthropology has been slow to come. European to describe indigenous languages. Consistent with his scholars with research interests similar to those of cultural relativism, Boas believed that each language North American linguistic anthropologists are thus should be studied on its own terms rather than more likely to be found in departments of linguistics, according to some preset categories based on the study foreign languages and literatures, folklore, communi- of other, genetically unrelated languages(e.g, Latin) cation, sociology, or psycholog In his 'lntroduction to the Handbook(1911), boas To understand the special role given to the study of provided an overview of the grammatical categories languages in the Boasian tradition, we must go back to and linguistic units necessary for the analysis of the time when anthropology became a profession in American Indian languages and argued against ove the USA, in the period between the last decades of the generalizations that would obscure differences across nineteenth and the first decades of the twentieth languages. He identified the sentence (as opposed to ntury. In that time the study of American Indian the word) as the unit for the expression of ideas, and nguages emerged as an essential part of anthro- listed a number of grammatical categories that are pological research. John Wesley Powell(1834-1902), likely to be found in all languages, while pointing out the founder of the Bureau of Ethnology, later renamed that the material content of words( the meaning of through grants from the US government, linguistic classify reality differently. One language might express eldwork, in the belief that by collecting vocabularies the semantic connections among words pertaining to nd texts from American Indian languages, it would the same semantic field by modifying one basic stem be possible to reconstruct their genetic relations and whereas another language might have words that are thus help in the classification of American Indian etymologically completely unrelated. As examples of tribes. Boas himself had become fascinated by the the latter type, Boas(1911)mentioned the different grammatical structures of Chinook and other words that are used in English for concepts centered s-nguages of the American northwest coast early on in around the idea of water'-lake, 'river, ' brook, his fieldwork, and seized the opportunity to work for rain, ' dew, wave, foam'-and four different words the bae and edit the Handbook of American Indian for concepts based on snow' in Eskimo. These guage(1911 examples were later taken out of context and the Although Boas, a diffusionist, was quite skeptical of number of words for snowin Eskimo(languages) the possibility of using languages for reconstructing grew larger and larger over the next decades in both genetic relations between tribesand was against any academic and popular publications correlation between language and race--he tried to was Edward Sapir(1884-1939)who, more than transmit to his students a passion for the details of any other of Boas's students, further developed Boas's guistic description and the conviction that interest in grammatical systems and their potential nguages were an important tool ieldwork, implications for the study of culture, and trained a new and(b) the study of culture, especially because the generation of experts of American Indian languages categories and rules of language are largely uncon- (e.g, Mary Haas, Morris Swadesh, Benjamin Lee scious and thus not subject to secondary rational- Whorf, Carl Voegelin). Unlike Boas, however, Sapir izations. Furthermore, Boas was committed to what was not a four-field anthropologist. He wrote and
comprising four subfields: archaeology, physical (now ‘biological’) anthropology, linguistics (now ‘linguistic anthropology’), and ethnology (now ‘sociocultural anthropology’). This vision of anthropology differs from the one found in the European tradition, where linguistics and social anthropology remained rigidly separate disciplines for most of the twentieth century, despite the emphasis on the use of native languages in fieldwork among UK anthropologists, and the theoretical and methodological influence of Bronislaw Malinowski (1884–1942), who wrote about the importance of linguistic research for an anthropological understanding of human societies. In the 1950s, the adoption of the term ‘ethnolinguistics’ (reflecting the European preference for ‘ethnology’ over ‘(cultural) anthropology’) for those studies that merged linguistic and anthropological interests signaled the intellectual recognition, at least in some European academic circles, of the importance of an ‘ethnological side’ of linguistic studies (Cardona 1976), but the institutional recognition of such a discipline within European anthropology has been slow to come. European scholars with research interests similar to those of North American linguistic anthropologists are thus more likely to be found in departments of linguistics, foreign languages and literatures, folklore, communication, sociology, or psychology. To understand the special role given to the study of languages in the Boasian tradition, we must go back to the time when anthropology became a profession in the USA, in the period between the last decades of the nineteenth and the first decades of the twentieth century. In that time, the study of American Indian languages emerged as an essential part of anthropological research. John Wesley Powell (1834–1902), the founder of the Bureau of Ethnology, later renamed Bureau of American Ethnology (BAE), supported, through grants from the US government, linguistic fieldwork, in the belief that by collecting vocabularies and texts from American Indian languages, it would be possible to reconstruct their genetic relations and thus help in the classification of American Indian tribes. Boas himself had become fascinated by the grammatical structures of Chinook and other languages of the American northwest coast early on in his fieldwork, and seized the opportunity to work for the BAE and edit the Handbook of American Indian Languages (1911). Although Boas, a diffusionist, was quite skeptical of the possibility of using languages for reconstructing genetic relations between tribes—and was against any correlation between language and race—he tried to transmit to his students a passion for the details of linguistic description and the conviction that languages were an important tool for (a) fieldwork, and (b) the study of culture, especially because the categories and rules of language are largely unconscious and thus not subject to secondary rationalizations. Furthermore, Boas was committed to what later became known as ‘salvage anthropology,’ that is the documentation of languages and cultural traditions that seemed on the verge of disappearing. This enterprise—a struggle against time due to the great damage already done, by the end of the nineteenth century, to the indigenous cultures of the Americas by European colonizers—produced valuable information on Native American traditions, but it had its methodological and theoretical drawbacks, the most flagrant of which was the inability to see or accept the effects of cultural contact and colonization. Through his writing and teaching, Boas brought scientific rigor to linguistic description and helped demolish a number of unfounded stereotypes about the languages that were then called ‘primitive.’ In an 1889 article entitled ‘On alternating sounds,’ Boas argued that the commonly held view that speakers of American Indian languages were less accurate in their pronunciation than speakers of Indo-European languages was false and probably due to the lack of linguistic sophistication of those who had first tried to describe indigenous languages. Consistent with his cultural relativism, Boas believed that each language should be studied on its own terms rather than according to some preset categories based on the study of other, genetically unrelated languages (e.g., Latin). In his ‘Introduction’ to the Handbook (1911), Boas provided an overview of the grammatical categories and linguistic units necessary for the analysis of American Indian languages and argued against overgeneralizations that would obscure differences across languages. He identified the sentence (as opposed to the word) as the unit for the expression of ideas, and listed a number of grammatical categories that are likely to be found in all languages, while pointing out that the material content of words (the meaning of lexical items) is language-specific and that languages classify reality differently. One language might express the semantic connections among words pertaining to the same semantic field by modifying one basic stem, whereas another language might have words that are etymologically completely unrelated. As examples of the latter type, Boas (1911) mentioned the different words that are used in English for concepts centered around the idea of ‘water’—‘lake,’ ‘river,’ ‘brook,’ ‘rain,’ ‘dew,’ ‘wave,’ ‘foam’—and four different words for concepts based on ‘snow’ in Eskimo. These examples were later taken out of context and the number of words for ‘snow’ in Eskimo (languages) grew larger and larger over the next decades in both academic and popular publications. It was Edward Sapir (1884–1939) who, more than any other of Boas’s students, further developed Boas’s interest in grammatical systems and their potential implications for the study of culture, and trained a new generation of experts of American Indian languages (e.g., Mary Haas, Morris Swadesh, Benjamin Lee Whorf, Carl Voegelin). Unlike Boas, however, Sapir was not a four-field anthropologist. He wrote and 8900 Linguistic Anthropology
Linguistic Anthropolog lectured on culture and personality, but had no interest relativism, if not a corollary of it. It was archaeology or physical anthropology anied by a concern for the proper repre aw as being more appropriately housed in museum of grammatical systems that could not be nan in anthropology departments. Furthermore using the categories of European languages. hereas Boas was skeptical of genetic reconstructio Third. the same antiracist attitude that char nd tended to favor acculturation as a cause of Boas's views on human diversity seemed to milarities between languages, Sapir was a strong the lack of value judgment associated with liever in the power of the comparative hich he used to reconstruct Proto-A and posit the Na-Dene linguistic group(co Athabascan, Tlingit, and Haida) ndividual and society(Mandelbaum 1949). In order While he was at Yale, Sapir encouraged his students to communicate their unique experiences, individuals is no accident that they ended up calling themselves little control. Linguistic rules are usually unconciorve cidenguistics rather than anthropology, and it need to rely on a public code over which they have nthropological linguists. They thought of them- and it is difficult for individual speakers to enter the selves primarily as linguists, as demonstrated by their logic of the linguistic system and alter it to their liking dedication to the study of the grammatical structures In this perspective, linguistic relativity becomes a way of American Indian languages(and other previously of exploring the power that words have over indi undocumented langua phasis on field- viduals and groups. It is thus a precursor to more work and their preference for historical and descriptive recent topics in linguistic anthropology, such as istics kept them separate from the new linguistics language ideologies(see Sect. 4.3 the 1960s, Chomsky's generative grammar. Te Sapir never developed the conceptual framework or anthropological linguists such as Mary Haas the methodology for testing the implications of these advent of this new paradigm was a threat because (a) intuitions about the language faculty. This task was it seemed more committed to linguistic theory than to left to another important figure in the history of languages, and in fact devalued matical descr linguistic anthropology, Benjamin Lee Whorf(1897- tion per se;(b) it was(especially at the beginning) 1941), a chemical engineer who worked as an insurance almost exclusively based on English--Chomsky ar pector, taught himself linguistics, and after 1931 that one could posit universals of language by entered into contact with Sapir and his students at ing on one language; and(c)it predicated the Yale. Although Whorf started out sharing several of for linguists to work on their own intuitions the basic positions held by boas and Sapir on the instead of working with native speakers or inferring nature of linguistic classification, he developed his grammatical rules on the basis of a corpus of elicited own conceptual framework, which included the exts as 1987). This last point was particularly distinction between overt and covert grammatical oblematic for those students of Native American categories, and an important analytical tool for languages who often had only one or two old speakers understanding what kinds of categorical distinctions to work with and could not find younger speakers to speakers are sensitive to-this issue was later further rain in linguistic theory and methods developed in the work on metapragmatics(Whorf 1956). Contrary to popular belief, Whorf was not so much concerned with the number of words for the same referent(e.g, 'snow)in different languages, but I Linguistic Relativity in the History of Linguistic with the implications that different grammatical sys anthropology tems and lexicons have for the way in which speakers Linguistic relativity is a general term used to refer to make inferences about the world. He believed that various hypotheses or positions about the relationship ways of thinking may develop by analogy with between language and culture(see Sapir-Whor fashions of speaking, a concept that was later revived pothesis). Although Sapir and whorf differed in their by Hymes's notion of ways of speakin discussion of the relationship between language and Whorf's work was harshly criticized in the 1960s ulture, and never produced a joint formulation of and 1970s, especially after the publication of Berlin what is meant by linguistic relativity, there is no and Kay's(1969)study of color terminology, in which uestion that the themes and issues often identified as they claimed that lexical labels for basic color terms inguistic relativity are the continuation of the Boasian are not arbitrary but follow universal principles. But paradigm. First, Sapir and Whorf followed Boas's more recent studies have given support to some of ntellectual curiosity for the indigenous languages of Whorf's ideas (Lucy 1992), and even the universality orth America as a way of channeling a more general of basic color terminology and its innate perceptu fascination for alternative ways of being in the world saliency have been questioned(e. g, Levinson 2000) and the desire to make sense of those ways. Second, to Sapir and whorfs ideas about the unconscious aspects the extent to which it started from an emphasis on of linguistic codes continued to play an important part human diversity, linguistic relativity was related to in the history of linguistic anthropology, and re-
lectured on culture and personality, but had no interest in archaeology or physical anthropology, which he saw as being more appropriately housed in museums than in anthropology departments. Furthermore, whereas Boas was skeptical of genetic reconstruction and tended to favor acculturation as a cause of similarities between languages, Sapir was a strong believer in the power of the comparative method, which he used to reconstruct Proto-Athabascan and posit the Na-De!ne! linguistic group (comprising Athabascan, Tlingit, and Haida). While he was at Yale, Sapir encouraged his students to go into linguistics rather than anthropology, and it is no accident that they ended up calling themselves ‘anthropological linguists.’ They thought of themselves primarily as linguists, as demonstrated by their dedication to the study of the grammatical structures of American Indian languages (and other previously undocumented languages). Their emphasis on fieldwork and their preference for historical and descriptive linguistics kept them separate from the new linguistics of the 1960s, Chomsky’s generative grammar. To anthropological linguists such as Mary Haas the advent of this new paradigm was a threat because (a) it seemed more committed to linguistic theory than to languages, and in fact devalued grammatical description per se; (b) it was (especially at the beginning) almost exclusively based on English—Chomsky argued that one could posit universals of language by working on one language; and (c) it predicated the need for linguists to work on their own intuitions instead of working with native speakers or inferring grammatical rules on the basis of a corpus of elicited texts (Haas 1987). This last point was particularly problematic for those students of Native American languages who often had only one or two old speakers to work with and could not find younger speakers to train in linguistic theory and methods. 1.1 Linguistic Relatiity in the History of Linguistic Anthropology Linguistic relativity is a general term used to refer to various hypotheses or positions about the relationship between language and culture (see Sapir–Whorf Hypothesis). Although Sapir and Whorf differed in their discussion of the relationship between language and culture, and never produced a joint formulation of what is meant by linguistic relativity, there is no question that the themes and issues often identified as linguistic relativity are the continuation of the Boasian paradigm. First, Sapir and Whorf followed Boas’s intellectual curiosity for the indigenous languages of North America as a way of channeling a more general fascination for alternative ways of being in the world and the desire to make sense of those ways. Second, to the extent to which it started from an emphasis on human diversity, linguistic relativity was related to cultural relativism, if not a corollary of it. It was accompanied by a concern for the proper representation of grammatical systems that could not be described using the categories of European languages. Third, the same antiracist attitude that characterized Boas’s views on human diversity seemed to motivate the lack of value judgment associated with linguistic diversity. For Sapir, linguistic relativity was a way of articulating what he saw as the struggle between the individual and society (Mandelbaum 1949). In order to communicate their unique experiences, individuals need to rely on a public code over which they have little control. Linguistic rules are usually unconscious, and it is difficult for individual speakers to enter the logic of the linguistic system and alter it to their liking. In this perspective, linguistic relativity becomes a way of exploring the power that words have over individuals and groups. It is thus a precursor to more recent topics in linguistic anthropology, such as language ideologies (see Sect. 4.3). Sapir never developed the conceptual framework or methodology for testing the implications of these intuitions about the language faculty. This task was left to another important figure in the history of linguistic anthropology, Benjamin Lee Whorf (1897– 1941), a chemical engineer who worked as an insurance inspector, taught himself linguistics, and after 1931 entered into contact with Sapir and his students at Yale. Although Whorf started out sharing several of the basic positions held by Boas and Sapir on the nature of linguistic classification, he developed his own conceptual framework, which included the distinction between oert and coert grammatical categories, and an important analytical tool for understanding what kinds of categorical distinctions speakers are sensitive to—this issue was later further developed in the work on metapragmatics (Whorf 1956). Contrary to popular belief, Whorf was not so much concerned with the number of words for the same referent (e.g., ‘snow’) in different languages, but with the implications that different grammatical systems and lexicons have for the way in which speakers make inferences about the world. He believed that ways of thinking may develop by analogy with ‘fashions of speaking,’ a concept that was later revived by Hymes’s notion of ‘ways of speaking.’ Whorf’s work was harshly criticized in the 1960s and 1970s, especially after the publication of Berlin and Kay’s (1969) study of color terminology, in which they claimed that lexical labels for basic color terms are not arbitrary but follow universal principles. But more recent studies have given support to some of Whorf’s ideas (Lucy 1992), and even the universality of basic color terminology and its innate perceptual saliency have been questioned (e.g., Levinson 2000). Sapir and Whorf’s ideas about the unconscious aspects of linguistic codes continued to play an important part in the history of linguistic anthropology, and re- 8901 Linguistic Anthropology
Linguistic Anthi d in the 1980s in the context of a number of Chomsky's notion of ideal speaker-hearer was prob ch projects, including of language lematic for many reasons, including the fact that not logy(Kroskrity 2000) bers of a speech same verbal resources or the same opportunities to use them. In his writings, Hymes shifted the focus from 2. Linguistic Anthropology in the 1960s Universal grammar to diversity. This ethnography of communication andi csenerative change had methodological implications From elicit- grammar was becoming popular in the USA,two ng grammatical patterns and speech genres (e.g ther important programs were also launched: the proverb interaction. Gumperz and Hymes's students went to Through his own qualitatively oriented work and the field with the goal of studying how language was his collaboration with Charles Ferguson on multI- being used in different events(e. g,ceremonies, village notion of language as used by linguists, and intro- street corner, market, dinner table). Linguistic luced the notions of variety, repertoire, and linguistic formance-the use of language (or its significant mmunity or speech community. In the same years, Dell Hymes launched a call for a comparative study of absenceh-became the starting point of any investi- ommunicative events to capture the ways in which gation(Bauman and Sherzer 1974) aking is a cultural activity and should be studied as Hymes and others reconceptualized performance as uch. His collaboration with Gumperz produced a a realm of social action, which emerges out of the interaction with other speakers and as such is not ew paradigm in linguistic anthropology, one in which reducible to the use of the linguistic knowledge and use of languages through ethnographic methods controlled by one individual. The focus on perform- esearchers were expected to study both knowledge nd H was no longer the grammars of indigenous language dimensions of any act of speaking, the role of individuals and groups in the reproduction and trans but communicative events and contextual variation formation of linguistic codes and the institutions they within and across Hymes stressed the need to see linguistics as part support, the responsibility associated with any display of one's verbal skills and the interactive construction anthropology, and insisted on the name linguistic of messages and meanings nthropology' over anthropological linguistics, he also helped defined a domain of inquiry that was in man ways independent of both linguistics and the 3. Context other subdisciplines within anthropology. Whether intended or not, this effort resulted in a type of In an ethnographic approach, language is seen as linguistic anthropology that was much less pre permeable to social situations and social roles while at the same time he g to define those situations and occupied with grammatical description and linguistic roles. This means that linguistic anthropologists need reconstruction than the one practiced by the previous analytical tools and units of analysis to define the generations, and more focused on the language and consequently its aesthetic and political context of speech. With this goal in mind, in the late formance of dimensions. This emphasis 1960s Hymes expanded Jakobson del of the inguage allowed toa gres aiag ab speaking, ber: with each letter of the word'speaking' standing for a left scholars in other fields worrying about linguistic graphers of communication should explore: Situation, ssues that were still central to anthropology, such as Participants, Ends, Act Sequences, Key, Instrumental- language evolution. While Chomsky was leading an ities, Norms, and Genre(Hymes 1967 reprinted in Hymes 1974). This list was meant to provide an etic activity. In this per he notion of commun researche rs, it helped many to identify dimensions of cative competence was central(see Communicative verbal communication that had beer Competence: Linguistic Aspects) en left out of previous investigations. In the meantime, Gumperz refined his analysis of multilingualism and I I From Communicative Competence to contact and became interested in the analysis of face to-face interaction. He expanded the concept of linguistic repertoire to include the range esources The documentation of variation and the observation that speaker-hearers use to make inferences about the of the social life of speech had taught fieldworkers that ongoing context. This line of inquiry resulted in the 8902
appeared in the 1980s in the context of a number of research projects, including the study of language ideology (Kroskrity 2000). 2. Linguistic Anthropology in the 1960s In the same years in which Chomsky’s generative grammar was becoming popular in the USA, two other important programs were also launched: the ethnography of communication and urban sociolinguistics (see Sociolinguistics). Through his own qualitatively oriented work and his collaboration with Charles Ferguson on multilingualism in South Asia, John Gumperz criticized the notion of ‘language’ as used by linguists, and introduced the notions of variety, repertoire, and linguistic community or speech community. In the same years, Dell Hymes launched a call for a comparative study of communicative events to capture the ways in which speaking is a cultural activity and should be studied as such. His collaboration with Gumperz produced a new paradigm in linguistic anthropology, one in which researchers were expected to study both knowledge and use of languages through ethnographic methods (Gumperz and Hymes 1964). The object of inquiry was no longer the grammars of indigenous languages, but communicative events and contextual variation within and across speech communities. Although Hymes stressed the need to see linguistics as part of anthropology, and insisted on the name ‘linguistic anthropology’ over ‘anthropological linguistics,’ he also helped defined a domain of inquiry that was in many ways independent of both linguistics and the other subdisciplines within anthropology. Whether intended or not, this effort resulted in a type of linguistic anthropology that was much less preoccupied with grammatical description and linguistic reconstruction than the one practiced by the previous generations, and more focused on the performance of language and consequently its aesthetic and political dimensions. This emphasis on the actual use of language allowed for great progress in the understanding of the cultural organization of speaking, but left scholars in other fields worrying about linguistic issues that were still central to anthropology, such as language evolution. While Chomsky was leading an antibehavioristic ‘cognitive revolution,’ Gumperz and Hymes were trying to develop a paradigm in which language behavior could be fully explored as social activity. In this perspective, the notion of communicative competence was central (see Communicatie Competence: Linguistic Aspects). 2.1 From Communicatie Competence to Performance The documentation of variation and the observation of the social life of speech had taught fieldworkers that Chomsky’s notion of ideal speaker-hearer was problematic for many reasons, including the fact that not all members of a speech community have access to the same verbal resources or the same opportunities to use them. In his writings, Hymes shifted the focus from Universal Grammar to linguistic diversity. This change had methodological implications. From eliciting grammatical patterns and speech genres (e.g., myths, traditional stories, proverbs, riddles) from native speakers, linguistic anthropologists shifted their attention to the actual use of language in social interaction. Gumperz and Hymes’s students went to the field with the goal of studying how language was being used in different events (e.g., ceremonies, village councils, classrooms) and different places (e.g., at a street corner, market, dinner table). Linguistic performance—the use of language (or its significant absence)—became the starting point of any investigation (Bauman and Sherzer 1974). Hymes and others reconceptualized performance as a realm of social action, which emerges out of the interaction with other speakers and as such is not reducible to the use of the linguistic knowledge controlled by one individual. The focus on performance allowed fieldworkers to recognize the creative dimensions of any act of speaking, the role of individuals and groups in the reproduction and transformation of linguistic codes and the institutions they support, the responsibility associated with any display of one’s verbal skills, and the interactive construction of messages and meanings. 3. Context In an ethnographic approach, language is seen as permeable to social situations and social roles while at the same time helping to define those situations and roles. This means that linguistic anthropologists need analytical tools and units of analysis to define the context of speech. With this goal in mind, in the late 1960s Hymes expanded Jakobson’s model of the speech event, and proposed the ‘SPEAKING Model,’ with each letter of the word ‘speaking’ standing for a cluster of dimensions of speech events that ethnographers of communication should explore: Situation, Participants, Ends, Act Sequences, Key, Instrumentalities, Norms, and Genre (Hymes 1967 reprinted in Hymes 1974). This list was meant to provide an etic grid for comparative purposes, and although it was rarely used in its entirety or formally adopted by researchers, it helped many to identify dimensions of verbal communication that had been left out of previous investigations. In the meantime, Gumperz refined his analysis of multilingualism and language contact, and became interested in the analysis of faceto-face interaction. He expanded the concept of linguistic repertoire to include the range of resources that speaker-hearers use to make inferences about the ongoing context. This line of inquiry resulted in the 8902 Linguistic Anthropology
Linguistic Anthropology notion of contextualization cues, linguistic Anthony Giddens, Mikhail Bakhtin, Michel Foucault through which'speakers signal and listeners Clifford Geertz, Paul Ricoeur);(c) the use of the what the activity is, how semantic content concept of ideology in trying to understand how derstood and how each sentence relates speakers conceptualize what constitutes appropriate precedes or follows(Gumperz 1982, p 131). Typical and interpretable language; and (d) the introduction examples of contextualization cues are intonational and wider adoption of new recording technologies patterns, paralinguistic features(e.g, tempo, pausing),(e. g, video, digitized images ), and their implications choice of code(e.g, English vs. Spanish), use of key for the definition of what constitutes an empirically words, and formulaic expressions. They can be studied adequate representation of speaking or, more broadly order to make sense of both successful and communication successful communication-crosstalk' was the ame with which miscommunication between people with different cultural backgrounds came to be known 4. New trends Gumperz's research on South Asian speakers of The last two decac des have seen the development of English in the UK In the late 1970s and early 1980s, the notion of language and other cultural resources. Three of the context was also being revisited by a number of other most important ones are language socialization, multi esearchers within linguistic anthropol michael lingualism, and the study of the linguistic dimensions Silverstein(1976) proposed a model of indexicalit of power and control that could be adapted to the sociocultural study language in and as context. He identified linguistic (i.e, they can be interpreted only on the basis of an 4. Language Socialization existential connection with some independently es- Although the acquisition of communicative com- tablished aspect of context) to highly creative (i.e, petence was always meant to be an important part of constitutive of their own context). Deictic terms like the program in the ethnography of communication, this'in'this room is cold are presupposing because it the field of language socialization did not fully develol is assumed that the room must exist for the utterance until the mid-1980s when elinor ochs and Bambi to be interpretable, whereas second-person pronouns Schieffelin(1984)defined it as(a)the process of getting (e.g, you in English, tu in Spanish)are creative, given socialized through language and (b)the process of nat they establish the identity of the addressee/ getting socialized to language, and offered some recipient while simultaneously creating the role of specific directions for research. By applying an an- addressee/recipient in the ongoing speech event. thropological reading to prior work on language anguages that have socially differentiated second- acquisition, they reframed it as embedded in culturally person pronouns (e.g, the classic T/V type of dis- specific expectations about the role of children and Spanish tu/Usted, German guages, French tu/cous, adults in Western societies and, particularly, in white tinction of many European lan or tu/Lei)are more extreme examples of systems in neither one of the two speech communities they had which words are used to activate or establish the studied (i.e in Papua New guinea and in Samoa) relevant social coordinates of equality /inequality, have a register corresponding to what is known as olidarity /power. The study of indexicality has be- baby talk'or'motherese, Ochs and Schieffelin not come a major focus of interest in contemporary only demonstrated that simplification in talking to ustic anthropology, and has been accompanied by infants, contrary to what was suggested by some a renewed interest in the role of the human body in the linguists and psycholinguists, is not universal, but stablishment of the referential grounding of most also, and more importantly, that simplification in communicative acts(Hanks 1990) talking to infants correlates with other forms of A more recent effort toward the definition of context accommodation to children, and local conceptual- is Elinor Ochs's(1996)model for the construction of izations of children and their place in society social identities. which is based on a number of Although related to child language acquisition situational dimensions established through language studies, language socialization studies examine the use:social acts, activities(a sequence of two or more cultural implications of what is being done with, to, acts), and affective and epistemic stances around, and through talk to children, with the theor The 1990s saw a rethinking of the concept of context etical assumption that learning is a two-way street in part due to: (a) a renewed awareness of the role of and that both experts and novices may come out of theory and methodology in defining the difference routine social encounters with new ways of thinking between the message and its context(Duranti and acting, and feelin Goodwin 1992);(b) the influence of a number of Language socialization is conceptualized as a never theorists from other disciplines(e. g, Pierre Bourdieu, ending process because speakers never stop learning 8903
notion of contextualization cues, linguistic features through which ‘speakers signal and listeners interpret what the activity is, how semantic content is to be understood and how each sentence relates to what precedes or follows’ (Gumperz 1982, p. 131). Typical examples of contextualization cues are intonational patterns, paralinguistic features (e.g., tempo, pausing), choice of code (e.g., English vs. Spanish), use of key words, and formulaic expressions. They can be studied in order to make sense of both successful and unsuccessful communication—‘crosstalk’ was the name with which miscommunication between people with different cultural backgrounds came to be known (thanks especially to a BBC program centered around Gumperz’s research on South Asian speakers of English in the UK). In the late 1970s and early 1980s, the notion of context was also being revisited by a number of other researchers within linguistic anthropology. Michael Silverstein (1976) proposed a model of indexicality that could be adapted to the sociocultural study of language in and as context. He identified linguistic signs along a continuum from highly presupposing (i.e., they can be interpreted only on the basis of an existential connection with some independently established aspect of context) to highly creative (i.e., constitutive of their own context). Deictic terms like ‘this’ in ‘this room is cold’ are presupposing because it is assumed that the room must exist for the utterance to be interpretable, whereas second-person pronouns (e.g., ‘you’ in English, tu in Spanish) are creative, given that they establish the identity of the addressee} recipient while simultaneously creating the role of addressee}recipient in the ongoing speech event. Languages that have socially differentiated secondperson pronouns (e.g., the classic T}V type of distinction of many European languages, French tu}ous, Spanish tu}Usted, German du}Sie, and Italian tu}Voi or tu}Lei) are more extreme examples of systems in which words are used to activate or establish the relevant social coordinates of equality}inequality, solidarity}power. The study of indexicality has become a major focus of interest in contemporary linguistic anthropology, and has been accompanied by a renewed interest in the role of the human body in the establishment of the referential grounding of most communicative acts (Hanks 1990). A more recent effort toward the definition of context is Elinor Ochs’s (1996) model for the construction of social identities, which is based on a number of situational dimensions established through language use: social acts, activities (a sequence of two or more acts), and affective and epistemic stances. The 1990s saw a rethinking of the concept of context in part due to: (a) a renewed awareness of the role of theory and methodology in defining the difference between the message and its context (Duranti and Goodwin 1992); (b) the influence of a number of theorists from other disciplines (e.g., Pierre Bourdieu, Anthony Giddens, Mikhail Bakhtin, Michel Foucault, Clifford Geertz, Paul Ricoeur); (c) the use of the concept of ideology in trying to understand how speakers conceptualize what constitutes appropriate and interpretable language; and (d) the introduction and wider adoption of new recording technologies (e.g., video, digitized images), and their implications for the definition of what constitutes an empirically adequate representation of speaking or, more broadly, communication. 4. New Trends The last two decades have seen the development of several new projects involving the interface between language and other cultural resources. Three of the most important ones are language socialization, multilingualism, and the study of the linguistic dimensions of power and control. 4.1 Language Socialization Although the acquisition of communicative competence was always meant to be an important part of the program in the ethnography of communication, the field of language socialization did not fully develop until the mid-1980s, when Elinor Ochs and Bambi B. Schieffelin (1984) defined it as (a) the process of getting socialized through language and (b) the process of getting socialized to language, and offered some specific directions for research. By applying an anthropological reading to prior work on language acquisition, they reframed it as embedded in culturally specific expectations about the role of children and adults in Western societies and, particularly, in white middle-class families. Using their discovery that neither one of the two speech communities they had studied (i.e., in Papua New Guinea and in Samoa) have a register corresponding to what is known as ‘baby talk’ or ‘motherese,’ Ochs and Schieffelin not only demonstrated that simplification in talking to infants, contrary to what was suggested by some linguists and psycholinguists, is not universal, but also, and more importantly, that simplification in talking to infants correlates with other forms of accommodation to children, and local conceptualizations of children and their place in society. Although related to child language acquisition studies, language socialization studies examine the cultural implications of what is being done with, to, around, and through talk to children, with the theoretical assumption that learning is a two-way street, and that both experts and novices may come out of routine social encounters with new ways of thinking, acting, and feeling. Language socialization is conceptualized as a neverending process because speakers never stop learning 8903 Linguistic Anthropology