Linguistics

I. The FieldDell Hymes

BIBLIOGRAPHY

II. Historical LinguisticsYakov Malkiel

BIBLIOGRAPHY

III. The Speech CommunityJohn J. Gumperz

BIBLIOGRAPHY

I THE FIELD

Linguistics has been called the science of language; but this definition begs the questions, What is scientific? What is language? Linguists themselves have not agreed; sometimes their answers have been quite narrow. If we are to consider the gamut of social science interests in language, we must adopt one of two approaches—either interpret “science of language” very broadly or conceive of it as part of some larger field.

In the first approach, “science” is taken in its broadest meaning, Wissenschaft, comprising all serious intellectual disciplines; “language” is taken as implicating all aspects of human speech. Linguistics is then truly the science of language, although as such it embraces work of both linguists and scholars in other disciplines. In the second approach, the “science of language” comprises only those kinds of knowledge for which linguists typically take responsibility. The science of language is then but part of a broader area of inquiry, the study of language; the science of language is linguistics proper, and the study of language may be called the field of linguistics.

The second approach is fairer to the present situation of linguistics. Linguistics is the indispensable basis for serious concern with any aspect of language, and its rapid growth is such that its effective scope may indeed come to equal the whole of the study of language. At present, the student of social science concerned with the place of language in human life must also consult other disciplines, perhaps including his own.

Our standpoint, then, will be that of the field of linguistics. We shall first sketch the development of linguistics within the context of the study of language, then characterize linguistics proper in terms of its scope and content today, and last discuss the import of its work with special reference to the social sciences.

Development of linguistics

The present expanding scope of linguistics is the outcome of a complex history. As a separately organized discipline, so named, linguistics is, However, quite young in English-speaking countries. The Linguistic Society of America was founded in 1924, and the Linguistics Association of Great Britain in 1965. Almost all of the independent departments of linguistics have appeared only since World War II. Indeed, “linguistics” has only recently displaced “philology” as a general name for the study of language, and “linguist” is still widely used by laymen for “polyglot” rather than for a member of a scientific discipline.

Early periods

Behind modern linguistics lies an accumulation of insight and knowledge that reaches to the early stages of human history. No known language is without at least some terms for facts of language and hence an elementary meta-language of a protolinguistic sort. Such terms sometimes show close analysis of structural features; for the most part, however, the internal structures of language remain unconsciously known, and it is terms for uses and varieties of language that are elaborated. Every society indeed has terms, beliefs, attitudes, and knowledge concerning language that may be singled out as its folk linguistics. The character and extent of this folk linguistics are of interest as conditioning linguistic change and as part of the subject matter of ethnoscience. Where a separate discipline of linguistic study has emerged, the interaction between it and the folk linguistics of its society may also set some of the limits and directions of the discipline and hence interest the historian of science and the sociologist of knowledge. Some 2,500 years ago there began to appear disciplined studies of language that can be taken as heralding a stage of national philologies. In China, India, and Greece, valued texts (the cultural arm, as it were, of expanding civilizations and aspiring states) came to be given special attention intended to preserve accurate knowledge of them and to aid in their use. These philologies were preceded by systems of writing, themselves embodiments of analysis of language. In the philologies, techniques of analysis became explicit, if with different emphases and differential success. The greatest achievement occurred in the Indian tradition (independently of writing), culminating in the work of Pānini (c. 500 B.C.). (Pānini was to influence the development of linguistics in Europe, once his work was made known there at the start of the nineteenth century.)

The Western study of language was shaped through most of its history by a tradition of philosophic thought and pedagogic grammars. In the classical world, analysis of language began amid controversies over the regularity or irregularity of language (analogia : anomalia), over language as part of nature or culture (physis : nomos), over relations between language and other subjects; and amid the establishment of Sophistic training and of educational institutions generally (see Robins 1951; Sandys 1903–1908; Marrou 1948). Etymological curiosity played a prominent part. A system of grammatical categories was evolved, based on the nature of Greek and then Latin. A great merit of the classical work was that it treated language structure integrally with language use; indeed, in education the grammarian was subordinate to the rhetor. Exclusive attention to the language (or languages) of an empire was a great limitation, shared by all early philologies. Given the social role of language study, this restriction was perhaps inevitable, but it was crippling to a general knowledge of language itself.

Early modern period

In the medieval period of the West (during the intermittent periods of renaissance before that which bears the capitalized name) theoretical notions about the rational structure of language were elaborated and language’s place in education and human life discussed (see Bolgar 1954; Robins 1951).

Skills that had begun in the classical world in the first renaissance of Greek literature—that of Alexandria—reached new heights in the Renaissance proper, and the tradition of classical philology which it founded provided the model for textual analysis and criticism in all fields, as well as models for grammars and dictionaries of the emerging national philologies in Europe. As the literary use and serious study of modern European languages grew, an empirical knowledge of languages from Asia, Africa, Oceania, and the New World slowly began to accumulate, especially through the field work of missionaries, history’s first organized body of ethnographers. By the eighteenth century the growing body of information was considerable, and there were scientific expeditions whose mission included collection of linguistic data (see Gray 1939; Wonderly & Nida 1963).

Some criteria of relationship among languages had been sought earlier in attempts to harmonize the three great cultural languages (Hebrew, Greek, Latin) and to establish the origin of the peoples of the New World, as well as of modern European peoples. The search for relationships was abetted no doubt by the Biblical account of Babel, which (1) asserted original unity of all languages, and (2) left the details of differentiation open to discovery (see Borst 1957–1963; Metcalf 1964).

In this period the theoretical unity of human language was treated in the Cartesian rationalism of Leibniz and others, in British empiricism, Scottish “common sense,” and French materialism. Indeed, as the intellectual foundations for the social sciences developed, the nature of language, posed as a problem of its origin in history and in the individual, engaged most major theorists “Hobbes, Locke, Rousseau, Adam Smith, Condillac, Herder, etc.—bearing as it did on the fundamental nature of man and the relationship between the natural and cultural worlds.

It is conventional to date the history of linguistics proper from the recognition of Sanskrit and the rise of diachronic linguistics, especially Indo-European studies, in the late eighteenth century and early nineteenth. This view (indeed, origin myth) preserves too much of the self-consciousness of an intellectual climate (romanticism, historicism) set off against the image of its predecessor, the Enlightenment. There was justice in the self-consciousness: in the study of language works were produced that became models and starting points for subsequent scholarship; with specialization, chairs, scholarly organizations, and journals increased quite dramatically. It is almost impossible, however, to fix any one point between the mid-eighteenth century and the mid-nineteenth century as decisive intellectually for the development of the successful methods of Indo-European study; Turgot, Rask, Bopp, Grimm, and Schleicher each has his claim to a contribution. Regarding theoretical views, some would now set aside much of the nineteenth-century effort, emphasizing the preceding rationalist inquiry as a more relevant predecessor. The work of Wilhelm von Humboldt thus comes to seem both an end point to developments of interest in the seventeenth and eighteenth centuries and a starting point for a strain of general linguistics in the nineteenth century.

No doubt the salience of periods of past linguistics will continue to change as modern linguistics changes. If we are to understand the history of linguistics, not for partisanship, but as a case in the comparative study of the general history of science and scholarship, one point is essential. Linguistic research, like social science research, proliferated as a sustained, organized, autonomous activity in the nineteenth century, but from an intellectual and empirical base in the general rise of scholarship and science in Europe as part of the expansion, at once intellectual and commercial, of its known world. With that rise began what can be truly called general linguistics, a linguistics which, while interdependent with the continued development of philologies and of specializations of other sorts, came to take as its compass all the languages of man. (On the place of linguistic thought in intellectual history, especially the place of language in human nature and culture, see Verburg 1952; Chomsky 1966;-Cassirer 1923, chapter 1.)

Nineteenth century

The major study on nineteenth-century linguistics (actually a period approximately from the French Revolution to World War I) records it as the triumph of historical work, especially of comparative Indo-European (Pedersen 1924). From that standpoint the course of the century is one of increasing precision and power of historical method and of its very wide application. Indo-European studies held the center of the stage, partly because of their institutionalization in Prussia and subsequently elsewhere and partly because of the accident of cultural history that the languages prospective Indo-European specialists had to learn in school were the classical “languages of culture.” Greek and Latin, together with their native languages and perhaps some other European language, gave these students intimacy with three or four branches of the language family they were to reconstruct. This cultural support and internal availability of data were not duplicated else-where. The Hungarian, Gyarmathai, had had the methodological insight to relate Hungarian and Finnish, in 1799 (see Pedersen [1924] 1962, pp. 105-106), but the languages of the family (Finno-Ugric) to which Hungarian belongs were scattered in Russia and the north, and scholars had to rely on cumulative field work. Moreover, there was no prestige in working out the relationship of one’s mother tongue to the languages of poor “fish eaters” fthe heroic Turkish conquerors were more attractive as linguistic kin, thus leading much research into a blind alley). Indo-European scholars worked amid the rise of Oriental philology and of interest in the “wisdom from the East” symbolized by Sanskrit.

Among the main accomplishments of the period were the pioneering comparisons of Rask (who was not adequately recognized in his own day); the comparative grammar of Bopp; the recognition of regular patterns of sound change in Germanic by Grimm (and Rask); the etymological dictionaries of Pott and Fick; the method of reconstruction and the family-tree model (Stammbaum) of relationship of Schleicher; the more exact specification of regularities in change in sound and of the role of grammatical analogy of the Jung-grammatiker and others of the 1870s (Ascoli, Verner, de Saussure, Brugmann); the wave model of relationship (Wellentheorie) of Schmidt, complementing that of the family tree. The first basis was laid for the historical study of many language families, both within and outside Indo-European (e.g., comparative grammars of Romance, Celtic, Dravidian, Bantu, Athapaskan). Near the end of the period, the great comparativist Meillet, while calling for attention to regularities independent of specific histories, could rightfully state that the principles already developed for the handling of regular sound change, analogy, and borrowing could continue to be fruitfully extended to all the language families of the world.

The traditions of anthropological philology also became established in this period, beginning perhaps with Herder’s thesis of the individuality and the scientific and humanistic value of the language and literature of every people, whatever its stage of development. This attitude was to be freshly stated many times, for example, by Boas, a century later. Much anthropological study of language has indeed been the philology of peoples without philology of their own. In this sense anthropology has been the third of the great philological enterprises of European civilization, following upon and complementing classical and Oriental philology.

A master term for the study of language was philology, a putative “queen of the sciences” to some, although “linguistics,” linguistique , and Sprachwissenschaft also were in use. Historical linguistics, requiring interpretation of texts, was often termed “comparative philology” in English-speaking countries; and since philologists proper produced the needed grammars and dictionaries, descriptive work also was often identified as “philology.” Analogous to “classical philology,” the study of contemporary European languages was labeled part of “modern philology” (Neuphilologie), and study of a particular set would be, e.g., not “Romance linguistics,” but “Romance philology.” In American anthropology, “philology” remained the common term into this century (e.g., Boas was retained by the Bureau of American Ethnology as “philologist”). In British usage and some learned American usage “philology” continued to serve until very recently. After World War I the trend among practitioners themselves to distinguish sharply between “philology” and “linguistics” was well established, however favorably the relationship between the two might be viewed (Pedersen [1924] 1962, p. 79). Philology might range far, but it remained inseparable from the study of texts and history, and linguistics was being defined as a more general study of language.

Looking back from the standpoint of contemporary linguistics, one sees in the nineteenth century important strands, which were largely neglected in accounts of the triumph of historical work. One is well known and the progenitor of continuing lines of work—linguistic geography and dialectology. It emerged in a rising wave of interest in geographical distributions in all the human sciences during the half-century embracing the turn of the twentieth century as midpoint.[seeLinguistics, article onthe speech community.]

A second major trend in general linguistics stemmed from von Humboldt and others and sought cognitive import in an evolutionary typology of languages. Carried on by Schleicher, Steinthal, and others in various forms, the effort was later discarded amid a general rejection of evolutionary sequences of stages in the cultural sciences. The whole-language labels, such as isolating, agglutinating, inflectional, polysynthetic, and incorporating, were retained only for individual traits that might be compresent within one language.

The concern with typology was the main source of the attention paid to the structural cut of languages and to its possible import (witness the joint launching of the first journal in social psychology by Steinthal and Lazarus in 1860); and the cognitive interest was retained, but associated with grammatical categories and processes of individual languages, quite divorced from broad classes and stages. (Through Boas the interest became an intrinsic part of American linguistic anthropology.)

Typology was thus part of a third trend, the preparation in the nineteenth century of the structural outlook that was to dominate the twentieth. In this outlook much was to depend on the approach taken to the sounds of speech. In the nineteenth century European languages often were taken as a norm, such that the sounds of “primitive” languages seemed odd and even to lack constant units—as so o’ m, projection of a lack of structure here betrayed an inadequate understanding of the nature of the primitive language being studied. Speech sounds were studied officially by a phonetics conceived as a natural science (Naturwissenschaft), not as a part of linguistics. Seeking universal objectivity through finer physical observation, phonetics plumbed a pit of variation, failing to see that universal invariants in language can be attained only through recognition of a patterning within language itself, and that within a language, objectivity of units is in the first instance inter-subjective and qualitative. In the 1880s such a cultural (or psychological) approach to speech sounds was worked out in Kazan by Baudoin de Courtenay and Kruszewski and was independently adumbrated by Boas. Their insights became effective only later—that of the Kazan school in the work of de Saussure, Trubetskoi, and Jakobson, the Boasian heritage in Sapir’s great paper, “Sound Patterns in Language” (1925). Meanwhile, the comparativists, by establishing ever more precisely the principle of invariance in historical reconstruction, anticipated and aided application of this principle in nonhistorical contexts. De Saussure’s brilliant reconstruction of a part of the proto-Indo-European vowel system, for example, was not only confirmed a generation later by the discovery in Hittite of a phonological feature postulated purely on internal evidence in the reconstruction, but its treatment of relationships within and between languages as manifestations of an underlying invariant feature earlier in time is also quite the same in form as the treatment of variants within a language as manifestations of an underlying invariant compresent in time, i.e., as manifestations not of an origin but of a present system.

Twentieth century

Modern synchronic, structural linguistics emerged essentially after World War I. Despite the groundwork that had preceded it, the participants in its emergence saw not continuity but conflict. De Saussure viewed synchronic and diachronic study as antithetical—the one dealing in systems, the other in atomistic traits; and many subsequent structuralists rejected historical work, at least as practiced. At the same time, many historical scholars rejected the new descriptive approaches. Even in American linguistics, where both history and structure have often been of com- mon concern, as the work of Sapir and Bloomfield shows, the issue has been debated (Hockett 1948, p. 188; Hymes 1964a, p. 600). In some circles, especially in Europe, conflict over this issue is perhaps only ending now.

The new movement was part of a general upheaval of intellectual interests after World War I and of the general shift from a primarily historical perspective to an interest in structure and form. With it came the definite triumph of “linguistics” as the general name for the study of language and the development of several different orientations toward the proper tasks of linguistics, often enough excluding portions of the general study of language by way of self-definition.

The first text of structural linguistics, the posthumous Cours de linguistique generate, published from lecture notes by students of de Saussure, ends with the maxim that linguistics has for its true and only object language itself (i.e., not facts of history, psychology, sociology, or whatever). De Saus-sure introduced the distinction la langue : la parole. It subsumes a variety of contrasts and has been variously interpreted and used, but the main import has been that the linguist’s task is the study of la langue, the underlying system and social fact (de Saussure was influenced by Durkheim, as was Meillet), rather than the various aspects of speech and other uses of language that were understood as la parole.

Saussurean linguistics as such was maintained in Switzerland by Sechehaye, Frei, Godel, and others. In Denmark Hjelmslev drew inspiration from de Saussure (and also from Sapir) and developed with Uldall a methodological algebra, glossematics, which postulated the immanence of language and made no assumption as to its empirical reality. In Czechoslovakia two Russian scholars, Trubetskoi and Jakobson, collaborated to lead in the development of a structural outlook uniting synchronic and diachronic work and giving attention both to internal and external functions of language. The accomplishments included a distinctive methodology for phonological and grammatical analysis; basic examination of phonological typology in connection with a search for general laws; new emphasis on the diffusion of linguistic features, structurally interpreted; and analyses of social and poetic varieties of language. In England J. R. Firth advocated a flexible, encompassing descriptive approach, one shaped partly by the English tradition of phonetic research and partly by Malinowski’s notion of the “context of situation.”

In the United States the matrix of structural linguistics was in large part anthropological. Boas had prepared the way by his critique of existing generalizations about the nature of language in the light of American Indian evidence and by his insistence on the description of each language sui generis. (The major statement of his contribution is the 1911 introduction to the Handbook of American Indian Languages.) His student, Sapir, put field work with a multitude of Amerindian language structures to brilliant account in his Language (1921) and in his 1925 study established the basic principle of structural analysis for both American linguistics and ethnography. Around Sapir and Bloomfield gathered the subsequent leaders of professional linguistics. In the United States Bloomfield’s Language (1933) was the first systematic exposition of the new descriptive approach, and it became the standard reference for a generation. Both Sapir and Bloomfield stressed the autonomy of language and linguistics. Sapir’s Language is a masterful set of variations on the theme of the autonomy of grammatical form, and Bloomfield’s book declares that the study of language has many times been approached, but never properly begun, because of failure to focus on language in its own right. Granting this autonomy, both saw larger implications for the results of linguistics, and indeed an interdependence with other disciplines—especially Sapir; in this respect, Bloomfield remained the more austere.

Each of these strands of structuralism has been labeled—the Geneva school, the Copenhagen school, the Prague school, the London school, the Yale school (although the differences between Sapir and his successor Bloomfield perhaps require recognition of two Yale schools). It must be remembered that distinguished linguists worked in the Netherlands, France, and elsewhere, without acquiring separate names and images, and that the labels are poor predictors of the work of individual scholars in the places named. The labels do serve as markers of salient intellectual emphases and outlooks. Differing intellectual contexts and assumptions led to sometimes acerbic exchanges in the first years after World War II; in particular, the stringent behaviorism dominant in the United States was set in method and attitude against the more open-minded empiricism found in Britain and against those influenced by phenomenology, logic, and dialectic on the Continent. There was salient skepticism, if not outright rejection, of the study of meaning, and there was antipathy to statements in terms of process in the Neo-Bloomfieldian camp. In the United States, the effective scope of linguis- tics became especially narrow for a time (see Joos 1957), in contrast to European outlooks and, indeed, the outlook of Boas and Sapir.

Recent trends and structural models

Developments within structural linguistics continue to hold the center of the stage. The main trend has been away from primary concern with phonology (whose conquest by structural methods was a main achievement of the 1920s and 1930s and in terms of which the battles of the day were fought) to concern with morphology, syntax, semantics, and, to some extent, stylistics, or poetics. In the 1940s and early 1950s great attention was given to the analysis of grammar in terms of morphology, the identification and distribution of forms. Especially in the United States there was an attempt to analyze grammar in terms analogous to those used in phonology, and insofar as possible, on the basis of the results of phonological analysis. This strategy of working upward from phonology fitted an emphasis on overt data (rather a “decoding” one) and a reluctance to deal with imputed entities. In the words of Joos, the maxim was “text signals its own structure.” European scholars had usually taken the psychological reality of linguistic structure into account, as had Boas and Sapir; and glossematics began with the higher levels of text and grammar, working downward to phonology. In the late 1950s and the 1960s Noam Chomsky has stressed the centrality of grammar and the underlying mental abilities of users of language. His central point has been that structure in text is largely recognized not because overt signals are perceived but because users of a language apply their knowledge of grammatical structure. Chomsky’s work, along with a general revival of interest in semantic work, has been the major force in bringing general American opinion into agreement to this extent with European and earlier American interests.

Models of language structure current on the American scene can be noted cursorily as follows:

(1) The Trager-Smith-Joos model, dominant in the early 1950s, treats phonology, grammar, and semology as coordinate and parallel in organization. Relations of language to culture and society are treated as questions of “metalinguistics.” Language is associated with other communicative systems, such as kinesics (gesture and body motion) and paralinguistics (nonlinguistic phenomena of voice), within a general framework for the analysis of culture (Hall 1959).

(2) The tagmemic model, devised by Pike and subsequently developed by Longacre and others, has become the framework for descriptions from many different parts of the world, used especially by the members of the Summer Institute of Linguistics (a group devoted to translating the Bible). It treats phonology, grammar, and lexicon as coordinate hierarchies, parallel in organization. The concept of tagmeme treats as central the relation in grammar between a position and the class of elements that can occur in it, thus reintroducing and generalizing the notion of paradigm. Inspired by Sapir, Pike has sought to generalize linguistic methodology to make it applicable to cultural behavior (see Pike 1954). In doing so he has coined the terms “emic” and “etic” (from “phonemic” and “phonetic”) for the difference between classifications of phenomena that are based on features validated internally in terms of the structure in question and classifications that are based on externally devised or generalized criteria.

Recent descriptive work done by Pike has emphasized use of matrices; by Longacre, analysis of strings and inclusion of transformations (1964).

(3) The “means-ends” model, as Jakobson (1963) has dubbed the original Prague school approach, has diffused into several lines of American research, largely through Jakobson’s presence since the early 1940s. There is widespread acceptance of the notion and importance of distinctive features in phonology: the “letter size” units, or phonemes, are not ultimate constituents but bundles of contrasting components (e.g., voicing, produced by vibration of the vocal chords, versus voicelessness, or absence of vibration, which distinguishes the otherwise identical series of English stops /p t k/ : /b d g/). But not all linguists follow Jakobson in defining phonological features acoustically (in terms of the physical properties of the speech signal), some retaining definition in terms of articulation (where and how the sounds are produced); neither do all linguists follow him in supposing a proposed set of 12 phonological features to be adequate for describing all languages or in regarding distinctive features as always forming oppositions of a binary sort. Jakobson’s analyses of grammatical categories in terms of distinctive semantic features have helped shape componential analysis; his work has influenced literary scholarship, ethnography, and psychology as well.[seeComponential Analysis.]

An enduring importance of the Prague school is that it always keeps the true social complexity of language in view as an object of linguistic study. In its conception a language is a dynamic “system of systems.” A language is never a static, homogeneous system, but a developing set of interdependent, imperfectly adjusted subsystems, some of whose items are disappearing, some emerging. A language, moreover, is a set of means specialized to a multiplicity of communicative and social ends; for example, there are referential, expressive, and directive functions in speech; and literary, standard, and other varieties of a language. (On the history and present development of the Prague school, see Vachek 1964; 19660; 1966b.)

(4) The most influential model both in the United States and abroad is that of transformational—generative grammar. As first formulated in the early 1950s by Zellig Harris (codifier of much post-war American method), transformations were a technique for syntax and text analysis serving to account for relations between sentences. Particularly telling have been the ways in which transformational analyses can account for the ambiguity of a sentence in terms of its derivation from two different underlying sentences; uncover an underlying difference between two superficially quite similar sentences; and integrate the productive systematic relations between different forms of sentence (e.g., actives, passives, negatives).

In the work of Harris’ student, Chomsky, transformations have become part of a new model of language as a whole and of a program for radical recasting of previously held assumptions and goals. Analysis of the underlying “deep structure” of grammar is stressed; much of what was treated by earlier descriptive linguistics is relegated to “surface structure.” Grammar shapes phonology rather than the reverse, and the phonological output of a grammar is specified in terms of distinctive features. (For the first influential monograph, see Chomsky 1957; for subsequent views, see Chomsky 1964; 1965; 1966.) The program is one of attack on behaviorist and positivist outlooks from a stand-point of rationalism and recent philosophy of science. The Neo-Bloomfieldian goal of an algorithm, a mechanical procedure for the analysis of a language, is shown to be futile, and earlier models of syntax are found inadequate to account for the actual facts of languages and the abilities that users of languages must have. Indeed, the goal of accounting for observed data is defined as mere observational adequacy (in effect, merely reporting the data). At the least, linguistics must aim for descriptive adequacy, an account of the knowledge that a fluent user of a language has of its structure, and which he or she can apply in producing and interpreting the infinite set of sentences of which a language is normally capable. What is crucial is the agreement of a grammar not with a corpus but with users’ judgments of acceptable and unacceptable sentences. The true goal of linguistics should be explanatory adequacy; that is, linguistics should characterize the nature of the equipment by means of which a child acquires such knowledge. To achieve the normal yet nearly miraculous result of an infinite capacity from a finite experience in but a few years, a child must be presumed to apply actively a native endowment, formulating theories to account for and go beyond the speech he hears. The rapidity and accuracy of a child’s success, no matter what the language, indicate that all languages are of only one or a few fundamental types and that the contribution of the native endowment must be great. In this light the earlier emphasis in American linguistics on the diversity of languages is reversed in favor of an active search for universals.

The focus of linguistic theory is thus reformulated as linguistic competence, the knowledge of the ideally fluent user of language in an ideally homogeneous speech community. Theory is completed by an account of linguistic performance, comprising the various conditions—psychological, occasional, social—that modify and affect the expression of underlying competence. The critique of learning theory and behaviorism provides a new program for psychologists; the model of language stimulates new work in semantics and to some extent in stylistics and ethnography, although some lines of research into the social role of language are considered misguided or premature. It is fair to say that transformational-generative grammar has replaced Neo-Bloomfieldian work as the focus of attention today. Apart from the close-knit group centered about Chomsky and Halle, some linguists join transformational syntax with other approaches, especially in phonology; and some transformation-alists abjure Chomsky’s views of the psychological and philosophical import of the model (e.g., Harris, Henry Hiz).

(5) The stratificational model owes its name and central formulation to Sydney Lamb. It has attracted several leading linguists (e.g., Hockett, Gleason) and has affinities with the work of Halliday. Originating as a systematic explication of principles in earlier American descriptive work and as a response to needs in computer use, especially mechanical translation, the model treats language as four levels, or strata—semology, lexology, morphology, phonology. The elements of the several strata are related by realization rules, which may be traced either in speaking (semology to phonology) or in hearing (phonology to semology). The model thus is intended to represent the processes of a user of language in either activity (see Gleason 1964 and Lamb 1966 for other formalizations). As with tagmemic analysis and the Harris-Hiz type of transformational analysis, there is interest in units larger than the sentence, such as whole narratives. The sememic analysis has close ties with work in componential analysis and ethnoscience.

(6)System-structure is a generic name for models derived from the Firthian heritage in Great Britain. The most notable of these models is what has come to be known as the “scale-and-category” grammar of M. A. K. Halliday and others (see Halliday et al. 1964; Mclntosh & Halliday 1966). A special interest of the approach is its direct in-corporation of questions of “institutional linguistics” (sociology of language) and its varied application to problems of language teaching, translation, analysis of style, and study of social dialect.

While models of the way grammars should best be written are the center of attention, other lines of linguistic work continue to develop, often revitalized in the light of structural principles; new interests emerge, often as applications and extensions of structural perspective; the concerns of the humanities and social sciences and of social institutions entail selective attention to the results and possible applications of linguistics. In short, controversy focused on structural models should not obscure the great variety of present-day linguistic activity. This variety can be grasped if one considers the three main ways, each cross-cutting the others, in which linguists may be grouped for purposes of congresses, journals, appointments, and so forth. There is affiliation with one or another of the structural approaches (some scholars affiliate with none); there are the subdisciplines, such as the phonetic sciences, lexicography, semantics, stylistics and poetics, onomastics, philology, dialectology, formal and mathematical linguistics, applied linguistics, anthropological linguistics, psycholinguistics, and sociolinguistics; and there are the groupings in terms of languages studied, whether in the sense of language family (Indo-European, Germanic, Algonquian linguistics) or area (African, Indian, Oceanic linguistics).

Distinctive constellations of approach, subdiscipline, and languages studied do occur, coupled with nationality or geographical region; but the overriding trends today are an internationalization and diffusion of interests and outlooks and a broadening and integration of them. Each major structural school, for example, has adherents in several countries, and a variety of links with sub-disciplines and language families and areas. Moreover, each structural school has had a conception of structural linguistics as contributing to a larger enterprise, such as a general science of signs (semi-ology, or semiotics), and as a link in the integration of the natural and social sciences, or at least as a strategic sector of social science; and such contributions are increasingly being realized.

Like many histories, this of linguistics has been rather Hegelian, moving from place to place and topic to topic as the “spirit” of the main advances moved. The main outlines might be caricatured in a Hegelian manner, describing the successive stages as “no language is known” (folk linguistics); “one language is known” (early national philology); “few languages are known” (later philology); and “many languages are known” (general linguistics). Yet any period comprises both the new and emerging and the old and steadily continuing, in a variety of forms, as has been indicated for the present state of linguistics. The nineteenth century was not all historical; the twentieth century is not all structural grammar; and if future linguistics builds on both in pursuit of a dominant functional concern with the place of language in human life, it will be in a spirit not of disregard of other concerns but of their integration. (On current trends, see Mohrmann et al. 1962; International Congress …1963; Ivid 1965; Current Trends in Linguistics; Biennial Review of Anthropology.)

Content of linguistics

We must now take stock more exactly of the present content of linguistics. Linguistics proper can be defined by tasks that remain constant and characteristic. These central tasks are to describe languages, to classify them, and to explain their differences and similarities.

Descriptions of linguistic data

The descriptive task (understood as equivalent to analysis) is primary. In pursuing it, linguists may use a variety of means for determining data and may place data in a variety of frames of reference. To grasp this variety a general framework is needed. We shall adapt and enlarge one put forward by Hockett (1948, pp. 188-190; Hymes 1964a, pp. 600-601), distinguishing four kinds of means by which data are determined (contact, philological, reconstructive, theoretical) and four major frames of reference within which data are placed (syn-

chronic, diachronic, diatopic, syncritic; Table 1 outlines this descriptive mode.

Means for determining linguistic data. Of the four kinds of means, contact subsumes first-hand observation, interviews, surveys, and experimentation in direct contact with users of a language— in short, the various forms of field work and laboratory work, including introspection into one’s own speech habits. Such work has been most characteristic of descriptive linguistics and dialectology, and of anthropology, but the expanding role of psychologists and sociologists in the study of language must be noted (see Lounsbury 1953; Hymes 1959; Samarin 1967; Berko & Brown I960; ErvinTripp 1964).

The philological methods are those used in the interpretation of the nature and transmission of written records, especially texts. Such methods are not limited to the study of classical languages or the languages of civilizations with writing. In the narrower sense that philological methods have come to have for some linguists, a field worker may find himself later in a philological relationship to his own materials; and in areas such as the New World the interpretation of records left by earlier investigators is indispensable for many languages no longer spoken. In the broader, earlier sense of philology, the method leads on into the general interpretation of texts for their cultural content as well as their linguistic form (see Sandys [1903–1908] 1958, vol. 1, pp. 4-13; Bernardini & Righi 1947; Hymes 1965; Malkiel 1966).

By reconstructive methods are meant those methods that use systematic variation in known languages (or dialects) to infer the presence and perhaps the patterning of features unknown in some other. When the evidence is from one language (or dialect) and the inference is to an earlier stage of it, the work is known as internal reconstruction. When the evidence is from two or more dialects or languages and the inference is to an earlier ancestral language, one usually speaks of the comparative method. Notice that in linguistics the term “comparative method” has this specialized use. Scholars in other fields have sometimes been misled into taking it as designating a general method of comparison, or as equivalent to such other techniques as the method of controlled comparison in social anthropology or the comparative method of nineteenth-century social evolution. Analogues to these in the study of language fall within the syncritic frame of reference and are quite distinct from what linguists speak of as the “comparative method” (see Bloomfield 1933; Hoenigswald 1960).

A less common form is that in which the inference is from a series of geographically linked languages or dialects to an intermediate or adjacent one; such use may count as a sort of reconstruction in space rather than time.

Theoretical methods concern aspects of data that are postulated by a general theory or inferred on theoretical grounds. A theoretical component is present in any description, although the extent to which it is recognized, if at all, may vary. Some examples are: that dialects be described in terms of the elements of a fixed overall pattern; that distinctive features be binary; that rules specifying the phonological shapes of words be ordered (or unordered); that of two differing accounts the shorter is to be accepted; that of two differing accounts the more intuitively persuasive is to be accepted; and so forth.

Frames of reference. As frames of reference, linguists commonly distinguish “structural,” “descriptive,” or “synchronic” from “historical,” or “diachronic,” linguistics; linguistic geography, or dialectology, and typological comparison, together with general linguistics, often are distinguished as well. The implicit logic is to separate the primary description of languages from concern with change (through time) and with variation (in space), and from contrast and generalization independently of space and time. In adopting this scheme (somewhat relabeled for symmetry’s sake), we recognize that no concise grouping can be wholly satisfactory and that the names must be somewhat arbitrary, designating as they do gross clusters of work that overlap and are internally diverse.

Synchronic description treats features (dialects, a language) with respect to a particular time, and, by implication, with respect also to a particular place. (“Structural” is used as a surrogate, but descriptions of the sort intended developed long before modern structuralism, and analysis within any frame of reference may be methodologically structural.) While other kinds of work may be strictly speaking “synchronic,” the common connotation of the term in linguistics is that for the purpose at hand, data can be treated as coming from a single source and as essentially homogeneous or unified; in effect, “synchronic” means an idealized single case. The classical presentation of such a synchronic analysis is in the form of a grammar, texts, and dictionary. (For treatments of grammatical description, see Harris 1951; Hockett 1958; Gleason 1955; Bach 1964; Katz & Postal 1964; Longacre 1964; Martinet 1960; Robins 1964; Chomsky 1965; Mclntosh & Halliday 1966; and the several approaches represented in the Monograph Series on Languages and Linguistics [see “Report of the Fifteenth Annual …” 1964] and the journal Language. On texts, see references in Hymes 1964a, p. 365; on dictionaries, p. 209.)

A synchronic structural description must choose or assume a particular norm—say, the standard speech of educated persons or the informal patois of lifelong inhabitants of a village. That is, any serious description must specify its boundary conditions—for whom and when and where it holds. Questions of contexts, purposes, and modes of use must enter (a point most consistently made by Firth and other British linguists such as Robins 1964; Halliday et al. 1964; Dixon 1965). Such questions are often distinguished as “functional,” “structural” serving for the internal makeup of language. This usage, with its internal-external dichotomy, is not happy, however, since use must itself be structurally analyzed and since the point of structural analysis of the makeup of language is to treat features in terms of their functional relationships (e.g., sounds in terms of the contrasts into which they enter to distinguish utterances; see Sapir 1925). Considerations of structure and function apply throughout linguistics, and the extension of this joint scope is of special importance to the social sciences. It is better to specify the exact kind of structure and function in question, rather than to rely on a usage that implies a false discontinuity.

What the classical presentation of a description of use might be like is not yet known. Models for the description of language use have been often postulated, but actual analyses are rare. As prolegomena to such descriptions one can cite the framework employed by Halliday (Halliday et al. 1964) and the concept of an “ethnography of speaking” (Hymes 1962; 1964f b). Restricting attention here to an idealized single case, one can say that an ethnography of speaking would identify and describe the speech events and sequences of speech acts recognized in a community and their distribution; their purposes and interactional norms (e.g., formal/informal); the relationships within speech events—who can participate as speakers and hearers, with regard to what settings, topics, message forms, channels, and codes; the patterning of messages in exchanges and sequences (conversations, curses, narrations, etc.); the hierarchies of functions (referential, expressive, rhetorical, poetic, and other) in such events; and the role of language with respect to other codes. No full accounts yet exist.

Diachronic description treats features (dialects, languages) in terms of time. Often one is concerned with changes in a single line of development, but features that remain stable may be of interest too; and the tracing of borrowings and their etymologies into earlier periods of other languages may become an objective of study. Among products of diachronic description are historical grammars, etymological dictionaries, statements of sound laws connecting stages of languages (e.g., Grimm’s law), depictions of innovations leading to the present vocabularies of members of language groups, and so forth. Diachronic descriptions need not draw a hard line between so-called internal and external aspects of language change, and, indeed, for many purposes cannot do so. Beyond the question of what has occurred, questions of when, where, how, and why arise, requiring reference to uses, contexts, and values for their answers. It is not unusual, however, for a distinction to be drawn between historical linguistics, internal change, and language history, external events. (See chapters on language change in Sapir 1921; Bloomfield 1933; Hockett 1958; Gleason 1955; Martinet I960; 1962; Hymes 1964a, parts 8 and 9.)

Diatopic description treats features (dialects, languages) in terms of space; it is thus inseparably associated with dialectology. Two observations must be made. First, dialectology is often synchronic description; important early work in Europe, for example, was motivated by concern for local forms of speech as against standard languages. There is also a European tradition of local ethnography that includes community speech as well as objects and practices. Second, diatopic description proper has not usually been undertaken for its own sake. Interest in the distribution of phenomena has often derived from diachronic questions of origin, spread, and loss. For dialects and languages, one may seek to infer an earlier location, migration, dispersion, or particular processes of formation. For features, one may seek to infer a particular history and also to sharpen a theory of linguistic change.

It remains that dialectology and linguistic geography require separate recognition, both analytically and for their special traditions of work. Earlier diatopic work most familiarly focused on sounds and words and produced atlases and maps. Recently there has been much recasting of earlier work in structural terms—e.g., analyzing the distribution of sounds in terms of phonological patterns (Hymes 1964a, p. 481).

A second kind of diatopic description is concerned not with geographic space but with social space. Earlier work was most concerned with the social dimension as it “horizontally” tied together several communities. Current work stresses “vertical” relations within communities, talking more of social dialect than of geographical dialect.

Such study may focus on the distribution of specific linguistic variables within a community; in doing so it gives special attention to the social valuation and role of these variables, for the light shed on the community as well as on processes of origin and diffusion (see Fischer 1958; Labov 1963; 1964). Description of variation in social space may also focus on all the varieties of speech to be found in a community—speech levels, men’s speech and women’s, baby talk, argots, and the like. The features of such forms of speech are specifiable structurally, but they appear here as aspects of the use of language within a universe that is defined first of all in social, not linguistic, terms. From an internal linguistic standpoint, cases of, say, men’s versus women’s speech appear only when the differences intrude into the normal description of a language, e.g., as constant differences in the phonological shape of lexical items or as obligatory alternates for inflectional elements. The speech appropriate to men and women presumably is differentiated in every society, however; it should be possible to describe how to “talk like a man” and “talk like a woman” everywhere. From the present standpoint one would inquire into the nature of the differentiation—whether or not it could be ignored in an ordinary grammar.

Such diatopic work is closely linked with description of use within a community. Indeed, in the description of a single community a full diatopic account and a synchronic ethnography of speaking differ only by an element of definition. An idealized synchronic account would restrict itself to a single variety of a language; but in fact every speech community has at least three (formal, conversational, slang). At the level of relations among a series of communities, however, or of relationships within a unit such as the nation, new complexities and types of problems emerge, so that the community-linked perspective of an ethnography of speaking can be considered part of a more general enterprise, the construction of an integrated theory of sociolinguistic description for social units of whatever scale and size.

Syncritic description compares dialects, languages, or features without regard to relative position in space and time. Such work is often called “typological,” especially when the purpose is structural classification of whole languages or their subsystems. “Comparative” would be a natural term, had it not been pre-empted for a reconstructive method and, by extension, the particular kind of work to which that method contributes. Some British linguists indeed seek to recapture “comparative” for comparison of languages generally, including comparison for purposes of translation and language teaching (often called “contrastive linguistics” in the United States; see Halliday et al. 1964, pp. 111-112). Syncritic (from the Greek synkrisis, meaning “comparison”) seems useful as a term that is both general and unambiguous, and so we introduce it here.

Syncritic description is undertaken sometimes simply to exhibit what diversity of structure may exist among languages, perhaps also to devise a classificatory scheme or measures for a scheme. Two particular purposes can be distinguished as contrastive and generic. A contrastive study sets out to specify differences, as in the characterizations of French and German by Bally or in the setting of Hopi against “Standard Average European” done by Whorf. (Whorf introduced the term “contrastive” in this use.) A generic study seeks significant commonalties in order to establish universals of language—underlying principles either true of all languages or so widely true independently of relationship in time and space as to require a general, rather than a historical, explanation (see Conference on Language Universals 1963; Martinet 1962; Chomsky 1965; references in Hymes 1964a, p. 661).

Syncritic description of uses and contexts also has contrastive and generic purposes, the contrastive being more salient. A notable example is Bernstein’s model of two types of code, elaborated and restricted, which are characterized by differences that include the signaling of subjective intent in verbally explicit form versus its transmission extra-verbally; the focusing of attention on verbal versus extraverbal channels; reliance on verbal persuasion versus authority; orientation toward personal discretion within roles versus status relationships and shared assumptions. The model has been most discussed with regard to class differences in England and the United States, but it has a more general application. For members of any class some situations may call for restricted code behavior; and it seems likely that aspects of socialization pertaining to whole societies can be contrasted in these terms. Other examples of syncritic work of this sort contrast types of verbal interaction as transactional or personal; types of bilingualism as coordinate or compound; types of speech situation as formal or informal; and the like. (These examples and related ones are discussed in articles by Bernstein, Ervin-Tripp, and Gumperz in Gumperz & Hymes 1964, and in articles by Bloomfield, Gumperz, Ferguson, Diebold, and Garvin in Hymes 1964a.)

Each of the contrastive analyses might be directed toward generic purposes. Thus one might have a systematic comparison of baby talk or of men’s and women’s speech with a view to generalizations as to their places in human society as a whole. A further step would be to analyze and compare ranges and uses for whole languages or speech communities, seeking to identify recurrent patterns of function (independent of historical connection) and to integrate such patterns with the results of the contrastive typologies. Such an approach entails a consideration of the place of language among other modes of human communication and the comparison of human communication with communication in other species, treating language as a resource differentially allocated and adapted in human societies on a certain generic base.

Comments. The sketch of kinds of means and frames of reference requires two comments. First, the broad categories of means and frameworks do not conflict or stand in isolation. A way of determining data may contribute to any frame of reference; a frame of reference may make use of any means. Field work, for example, is often thought of as a means to synchronic description, but its purpose may be to place a language genetically or to trace the process of acculturation; to study relations among geographical or social dialects; to determine the distinctive cognitive style of a language or to substantiate a proposed universal of grammar. Synchronic description is often based on contact work, but one may write grammatical rules for philologically interpreted texts or a reconstructed language or specify theoretically the universal parameters of grammar. Desirable as such interplay is in principle, in practice it may be obscured by specialization and controversy (as it has been on some occasions in synchronic and diachronic work); but in the long run interest in knowledge of languages prevails, especially in the work of great scholars. Every means that can contribute information is likely to find a place; frames of reference are likely to prove complementary. Such unity in diversity appears most often in work on a group of languages, be it Romance (see Malkiel 1964 b), Dravidian, or Algonquian. Most linguists being specialists in some language group, such work is a source of stability and strength to the field.

Second, an exact analysis of kinds of work and their interrelationships would require a greater number of dimensions, accurately and systematically named. One set would distinguish study of syntactics (relations of signs within a code), semantics (relations of signs to referents), and pragmatics (relations of signs to their users). (The distinctions are those of Morris; see Greenberg 1948; Hymes 1964a, pp. 27-31.) Our discussion has, in fact, consistently noted study of use (pragmatics) as well as study of codes (syntactics, semantics). A second set of distinctions concerns the underlying dimensions of time, space, social group, system, and function. For these, one can use the Greek forms -chronic, -topic, -gelic, -systemic, -telic together with a prefix to specify in what way each dimension is taken into account. Useful prefixes include a- : without reference to the dimension; mono- : one reference point; bi-, tri-, etc.: two, three, etc., reference points; poly- : multiple reference points; pan- : all reference points; syn- : treatment as having common reference point; dia- : treatment as having continuum, or linked series of reference points. Such dimensions and terms make terminologically convenient a large number of necessary distinctions. Thus, studies that treat the history of a language in terms of discrete stages (e.g., bichronic) can be distinguished from studies that treat it as a continuous development (diachronic). Within the context of synchronic descriptive theory, one can distinguish the complex adequacy of Prague school theory (diatopic, diagelic, polysystemic, polytelic) from descriptive theory that implies a wholly homogeneous description (monotopic, monoagelic, monosystemic, monotelic).

We have broached something of the tasks of classification and explanation; more must now be said about each.

Classification of languages

Languages are often classified by their common internal features (phonological, grammatical, lexical) and in one or another of three ways: as belonging to the same family, the same area, the same type. Although each way implies a different sort of process and explanation, the three need not be mutually exclusive; in a limiting case, a set of languages may all belong to the same family, area, and type at once.

When languages belong to the same family they share a genetic relationship. Specific features of each are explained as due to retention (perhaps much changed) from a common ancestor of all. English, Frisian, Dutch, German, Icelandic, Norwegian, Danish, Swedish (and extinct Gothic), for example, belong to the same family, Germanic, in virtue of their descent from a common ancestor that is called Proto-Germanic. As it happens, that ancestor can be shown to belong to an older fam- ily, Indo-European, whose common ancestor, Proto-Indo-European, may itself someday be shown convincingly to belong in yet an older family. The proportion of features that attest the genetic connection of languages may be quite small; what is required is that the presence of the features be inexplicable by chance or borrowing.

When languages are said to belong to the same language area (German Sprachbund), or convergence area, they share an areal relationship. Continued compresence in the same area may enable genetically related languages to maintain great commonalty of content through shared innovations and mutual borrowing, despite long divergence from their joint ancestor; these languages may even increase their commonalty after an earlier period of differentiation (see Hoenigswald 1960, pp. 155-157; Malkiel 1964a; Kroeber 1955). The most salient cases of areal relationship are those in which the languages are genetically unrelated. Thus the languages of the subcontinent of India belong to three distinct families (Indo-Aryan branch of Indo-European, Dravidian, Munda), but they share significant traits through sustained contact (Emeneau 1956; Hymes 1964a, pp. 642-650).

When languages are said to belong to the same type, one must notice what portion of their features is concerned. A typological relationship may be defined by one or a few traits of interest or by a distinct system or level (phonology, morphology, syntax, lexicon). Attempts to assign languages as wholes to types (a language type proper) have focused on grammatical or semantic characteristics or both. When specific features are investigated, a given language may, of course, fall together in type with quite different sets of other languages, depending on the features in question. (On genetic, areal, and typological classification, see references in Hymes 1964a, pp. 659-661, 651-653, 661-663, respectively.)

Classification of languages in terms of context and use is less well developed, a fact reflected in the absence of a comprehensive conventional terminology. There exists a scattering of individual terms not yet systematized. The dimensions of the subject can be sketched, however, along lines corresponding to those just followed, taking the genetic as concerned primarily with origin, the areal with co-occurrence, the typological as independent of either.

Some terms focus attention on languages as marked by their origin in particular circumstances of use. A koine is a language that has arisen as a lingua franca by a merging of traits among a group of related dialects, as in the Greek koiné of Hellenistic times (from which the term comes). A pidgin arises by drastic reduction of one language, typically with admixture of another; it is by definition a second language to all who use it. A Creole arises if a pidgin becomes and remains the first language for a group, expanding into a normal range of use. (Thus, by definition a Creole was once a pidgin.) By virtue of their common origin in conscious invention, constructed languages intended for international auxiliary use (Esperanto, Interlingua, etc.) belong here.

Some terms group languages according to their relationships within a community or larger population. Some groups of Sephardic Jews in Greece, for example, used Greek at work, Hebrew in religious observances, and Spanish in family conversation. Together the three languages formed their linguistic (or verbal) repertoire. One general classification of the varieties forming a verbal repertoire distinguishes those associated with geographic and social differences and those associated with differences of activity, as dialectal and superposed, respectively. [On this and other aspects of use, see Linguistics,article on THE Speech Community.] Terms often distinguish range of use; “standard language : dialect” and “world language : vernacular” are two such pairs. The use of “language : dialect” has varied and is still unresolved, but the two terms are always correlated in such a way that dialect is the subordinate term—language indicating a variety with higher status or wider use or a set of dialects as a whole.

An important kind of co-occurrence is that analyzed by Ferguson (1959; Hymes 1964a, pp. 429-438) as diglossia: two mutually unintelligible forms of language are in use—one for government, literature, formal religion, and the like (the “High” form) and one for informal conversation, the home, and the like (the “Low” form). The two are part of the verbal repertoire of some, but not all, members of a society, many knowing only the Low form. The general description of language co-occurrence within nations has begun to be studied as a nation’s sociolinguistic profile(Ferguson 1966).

Some terms specify use without necessary reference either to origin or co-occurrence. One such term is lingua franca, a language that serves as a common medium throughout a linguistically diverse region. Standard language, as a consciously codified form of language, belongs here, considered in terms of its intrinsic characteristics and associated attitudes and functions (see Garvin 1959; Hymes 1964a, pp. 521-526, with references). Indeed, all terms that designate components of a verbal repertoire or sociolinguistic profile may be specified and studied syncritically: the High and Low forms of a diglossia situation, languages of religion, trade languages, languages of concealment, slang, etc., can all be studied both in terms of the social circumstances of their origin and in terms of their co-occurrence with other varieties of language.

The uses and imports of the modes of classification are varied but ultimately interrelated. Genetic classification has a certain priority, as a background against which to interpret relationships of area and use and from which to guarantee the historical independence of cases for typological generalizations. As a mode of explanation of resemblances among languages, genetic classification has sometimes been set off against areal classification, as in the Boas-Sapir controversy (see Swadesh 1951; Hymes 1964a, pp. 624-637) and in the earlier California work of Dixon and Kroeber (see Hymes 1961a; 1964a, pp. 689-707). But in fact, the logic underlying the historical modes of classification makes them interdependent. This logic is to determine if corresponding features are (1) of independent origin (universals, convergent, chance), or (2) due to historical connection, and, if historical, whether (a) genetic (“cognate”), due to retention from a common ancestor, or (fc) diffusional (“borrowed”). Neither genetic nor diffusional origin can be assumed; each must be proved, and proof of one excludes the other. Actual historical work must thus attend to both. There remain questions concerning what explanation to assign to particular kinds and amounts of data. Well-integrated grammatical traits and basic vocabulary are the best, though not infallible, test of genetic connection. Despite a priori controversy, the work of the great students of linguistic prehistory is in practice one of cumulative inference from all the available evidence.

Although the languages of the world have been provisionally assigned to genetic groupings, such work is far from complete. For most parts of the world new and better descriptions of languages will permit deeper penetration of the past, as will reconstruction of protolanguages by the comparative method. Anthropologists (Swadesh, Greenberg, Haas, and others) have taken a leading role in this work, dealing with both proof of relationships and development of method for the great time depths and remote relationships that face students of linguistic prehistory. If data permit, beyond proof of relationship lies establishment of relative chronology (subgrouping) among the related languages and of the location and perhaps some of the cultural content of the ancestral language. Proof of borrowing may also lead to knowledge of relative chronology of relationship and the earlier location and cultural content of languages. Such work may provide a framework and hypotheses for prehistoric research with other lines of evidence, and of course it provides many precise examples of regularities and complexities of change, examples that have constantly been posed as problems for psychological and sociological explanation.

Studies of the American Indian languages of the Pacific Coast early revealed phonological areas where many distinct languages share systems containing few vowels and many consonants (including glottalized stops and voiceless laterals); and initial attempts to determine grammatical areas were made early in the century by Dixon and Kroeber. Neither in North America nor in the rest of the world outside Europe, however, has knowledge of areal connections gone beyond some notable individual studies. Increased interest in the structuring of interrelations of communities (such terms as “social field” and “intermediate societies”) may stimulate increased attention to areal relationship as its linguistic counterpart (cf. Gumperz 1961).

Typological relationship may be linked to areal relationship, as when reconstruction of the Proto-Indo-European vowel system suggests a former areal tie with languages of the Caucasus or when it is proposed that proliferation of phonemes is correlated with fewness of speakers in areas of linguistic diversity, since persons in small communities learn the languages of their neighbors as a result either of accommodation or of exogamy, and in either way introduce among themselves phonetic habits that come to enlarge the phonological system of their language. Most typological classification points in one of two directions. It seeks to explain recurrent types in terms of the limited possibilities and internal interdependence of linguistic systems (e.g., laws of the sort, “If A, then B”) and to relate such types to underlying generic properties of the human mind; or it seeks to delineate types in terms of the selective drift within a given culture history, as distinctive of a society or as characteristic of a sociocultural type. (Findings with regard to Hopi, Navajo, and Wintu, for example, may be seen as manifestations of an underlying outlook common to primitive society that Redfield dubbed “participant maintenance.”) The two directions seem to alternate in attention, interest in distinctiveness having given way recently to interest in what is common, considered apart from sociocultural adaptation; but underlying commonalty does not level the projecting dif- ferences that show languages engaged in the histories and lives of those who use them. Both interests are required to explain language.

Classifications as to use (often dubbed functional classifications) are of obvious importance to any concern with the varying roles of languages in culture, society, and personality. Choice and role of language are particularly important in nationalism, political identity, state formation, economic development, and in literacy, education, international communication; and they are also important for changes in the valuation of language itself relative to other modes of experience and communication. New analyses and syntheses of what is known are greatly needed, but they are only slowly beginning to appear (see Weinreich 1953; Ferguson & Gumperz I960; Fishman et al. 1966, pp. 424-458; Ferguson 1966).

Each principal mode of classification in terms of internal content may seem linked to a different frame of reference—genetic to diachronic, areal to diatopic, typological to syncritic—but something of their interconnection has appeared. Any instance of classification can be put into all frames of reference by asking: What are the underlying descriptions? How did the relevant features come about? Where do they occur? What are their defining characteristics? And in pursuit of historical explanation, the emergence, persistence, and sometimes extinction of families, areas, types, and modes of use are interwoven (see Hoenigswald 1960). Pidgins and Creoles, for example, pose problems for theory of genetic and areal relationship and for generic interpretation of typological resemblances. In sum, each mode of classification is useful, indeed indispensable, for particular questions: the major questions of linguistic explanation join together all of the modes. To generalize what Boas once wrote (with genetic classification in mind): “the problem of the study of language is not one of classification.... Our task is to trace the history of the development of human speech” ([1920] 1955, p. 212).

For any one language, its features can in principle be explained in terms of a portion common to all languages as languages (typological-generic), a portion retained from an ancestral stage (genetic), a portion acquired from other languages with which it has come into contact (diffusion al), all of these portions having adapted to each other along distinctive lines (typological-contrastive) in certain circumstances of use. In Sapir’s words, “The formal configuration of speech at any particular time and place is the result of a long and complex historical development, which, in turn, is unintelligible without constant reference to functional factors” ([1924] 1949, p. 152).

Explanation and import

Descriptions and classifications go but part of the way in explaining linguistic data and their import. At the height of the historical approach to language the maxim was offered that the only explanation of a linguistic form is an earlier linguistic form. When the NeoBloomfieldian descriptive approach was dominant, some found it humorous to be asked to lecture on the nature of language. One asked of languages not “why” but “what.” Such particularistic extremes set in relief the more common belief of linguists that in describing and classifying they also are illuminating something beyond the data in hand. How illumination is to come, what it should be, whether it is doggedly sought for or comfortably assumed—all yield much of the drama of the development of linguistics. The quest, most generally put, has been for meaning in particular texts and cultures, for the course of history, for characteristics of the human mind—in effect, for human nature as manifest in the concrete, in history, and in essence.

While the crucial role of language in human life makes its scientific study of perduring relevance to such goals, it remains true that most of the time most linguists seek the illumination of data within their own domain. “Why” questions, explicit or implicit, have a range of answers from the facts of a given language to general principles of structure, from facts of retention and borrowing to general processes of change. Each mode of description and classification implies explanation of some aspect of languages through the relationships it recognizes and discovers.

For our purpose, the critical point is reached when pursuit of explanation leads to relationships extending beyond language, when language is to explain or be explained in relation to other disciplines. Here questions of the unity and future of linguistics are most sharply posed. It is not that there are no questions of unity within linguistics proper, of the integration of different lines of purely linguistic work, but we can consider such questions only as they are entailed by the question of unity within the larger field of linguistics.

The field of linguistics

As must any discipline, linguistics proper must be master in its own house—literally, autonomous; but autonomy can be compatible with either isolation or integration. It is a striking fact that insistence on the independence of linguistics from other disciplines contributes to disunity within linguistics itself, for the independence is defined at the expense of some legitimate mode of studying linguistic data. Recognition of the unity of linguistics as a whole promotes recognition of its interdependence with other disciplines in the broader field of linguistics.

The many aspects of the import of linguistics and other disciplines for each other cannot be reviewed here; rather, we can consider what bases exist for integrating within a unified field of linguistics.

Unity and interdependence have long been recognized in principle, and often in practice, in the pursuit of historical explanation and philological interpretation, where one uses all there is to use. In the particular case, knowledge of customs, artifacts, social conditions, distributions, and environment plays a part inseparable from linguistic knowledge. The disciplines called upon to contribute include archeology, paleobotany, geography, folklore, comparative religion, numismatics, political and social history, and so forth. Where questions of the formation, movement, adaptation, specialization, obsolescence, and extinction of languages are concerned, dependence on social history is patent.

It is fair to say that the situation regarding unity is unclear outside the domain of historical explanation. A unity centered in structural analysis has gone far, integrating a great deal of work in linguistics proper, through recognition that structural formulations are prerequisite to many questions of history and use. There have been several efforts to base integration of a larger field on particular structural models (see references in Hymes 1964a, pp. 61-62), but none has prevailed. In other human sciences relevant to language one finds some use of linguistic models, some picking and choosing of linguistic results, some neglect of connection. The role of such disciplines in structural linguistics is similar; one finds some use of analytic models, some picking and choosing of results, some neglect or even denial of connection.

Such a situation may continue indefinitely. However, a larger unity within the field of linguistics can be attained; linguistic data must remain the focus, but the perspective brought to bear must encompass the gamut of relationships that determine the use of language.

The central requirement of such a perspective is that it focus on the integrity of the verbal message as an act. From such a focus there follows a series of consequences for conceptions of the object of analysis, consequences that have been partly indicated with regard to synchronic, diatopic, and syn-critic description, but that must now be explicitly drawn.

(1) Linguistic description has focused on the form of languages, neglecting the structuring of their use (la langue as opposed to la parole). The social sciences, on the other hand, have usually been concerned with language use, neglecting form. Consequently, most attempts to integrate language with culture or society have inevitably failed, for the terms of the relationship have been conceived as disparate abstractions. A grammar and an ethnography both treat verbal messages as data, but, typically, neither studies messages as having an integral structuring of their own. The one abstracts certain aspects of form, the other certain aspects of content (other kinds of form), as if they were historically disjunct products. Having put asunder, one may try to join together, but the form and process of speech, wherein the relation of language to culture and society is mediated, the cambium, as it were, of both, has not been incorporated into either abstraction. A unified field of linguistics requires study of the patterning of speech as well as of codes.

(2) Structural description has usually defined its object synchronically as a single homogeneous code or the abilities of an ideally fluent user of such a code in an ideally homogeneous community. Such simplification is useful when models of internal structure are being devised and single codes are being described in their terms. Models of the structure of speech must allow for multiplicity of codes —quotation within messages and switching between messages (of bilinguals)—and specialization of codes in different topics, occasions, roles, and institutions. A unified field of linguistics should have as its natural unit of study the speech community rather than the individual code.

(3) Structuralists have usually considered the relation of language to other aspects of life as only supplementing what normally counts as language structure. One looks out from the linguistic account, seeing its features as subject to variation or additional rules and restrictions. The best models do envisage extension of structural description from sentences to larger units—paragraphs, narratives, even conversations, and the many recurrent routines that make daily speech so much more predictable than the infinite potential of language would suggest. Even so, much remains undiscerned until one looks at the linguistic account from the standpoint of its additional functions in social use. In general, one cannot predict such functions from relations of structure as ordinarily described; rather, each level of organization (function) reveals new structural relationships among elements of those below it. Modal particles, for example, may show structural relationships only when seen to join with features of intonation to serve an expressive function. No internal linguistic relationship brings together greetings, terms of address, insults, curses, request forms, and so forth; only social rules can show each to be a set. Thus the lexicon and phrases of a language can be wholly analyzed structurally only from the standpoint of the social level, for some sets within the network of contrasts into which they enter are socially defined. Moreover, the usual structural account, normatively generalizing, omits as ungrammatical some sentences that specifically and acceptably do occur in a community.

In sum, social situations, relationships, and purposes bring into being and maintain linguistic (and nonlinguistic) features and relationships among them. A unified field of linguistics must consider the structures of languages from the standpoint of a description of their contexts of use.

(4) Structural descriptions have usually taken the functions of language for granted. They have focused on the organization of language in the service of reference. The latter term is used here as distinct from both denotation and meaning. One discovers the denotation of an expression in its application on particular occasions; its reference in a dictionary, which states criteria for its application; its meaning in the total import of the situation (see Firth 1935). In effect, most description has based itself on just those speech acts in which grammatical sentences are used with full referential force, neglecting the poetic or expressive facets of speech acts, for example, and the many messages in which (in Sapir’s phrase) it is as if a powerful generator were hooked up to run an electric doorbell.

Descriptive theory has generally taken the social adaptation of language for granted as being every-where the same. The images of one language per community and the infinite potential of any language (as well as the struggle against misconceptions of the adequacy of “primitive” languages) have led some to a militant egalitarianism that refuses to consider the obvious fact that the potentials of languages are not developed equally or in identical directions; that a language is often specialized in certain roles, not all; and that the valuation of even a native language may vary from community to community: free resource here, scarce good there; integral to unity here, easily abandoned there; an object of pride here, without prestige, even disvalued, there; and so forth.

Models of internal structure may ignore these variations in adaptation: reference is indeed the central function underlying grammatical structure. It remains that a unified field of linguistics must take the functions of language (both in speech acts and in communities) as problematic, and it must develop the concepts and methods for their study. (On functions in speech acts, see Jakobson I960; Hymes 1962; 1964b; on functions of languages, see Hymes 1961a; 1966; Ferguson 1966.)

The patterning of speech, from the standpoint of communities and contexts of use and the gamut of functions that speech serves in particular acts and groups, as men enact and transcend their situations, is a dimension of a “totalizing” approach (Sartre 1960) that calls for case studies and analytic comparisons going beyond any line of work familiar to us now; yet the need for such an approach can be indicated readily with regard to several problems.

Understanding the acquisition of language by children is of both theoretical and practical importance. Some seek to account for linguistic competence as the process of a child’s learning to use any and all grammatical sentences in a language; but such a conception of a child’s competence at once omits and idealizes. It omits, in that a child competent in all sentences still would be master of none, not knowing when, where, and how to speak, and about what, to whom, not sharing the attitudes and valuations of the community toward language. It idealizes, in that mastery even of internal structure is a matter of degree, affected at its root by social environment. To explain and affect the communicative competence of children requires the totalizing approach just indicated.

The relation of language to thought is persistently of concern. While the two are far from identical, experience and experimental evidence demonstrate that features of a language do shape behavior and thought. In the long run a language is shaped by the needs of its users, but in the short run the acquisition of experience through a particular language and the need to call on ready linguistic categories partly shape men. All men potentially perceive and think much the same; actually, they notice, store, and recall information mainly in familiar verbal grooves, although the aspects of life for which this is true may vary from society to society. In a multilingual world, moreover, a given language may be the matrix of thought to one person but only its superficial, occasional garb to another. The effect of a language on thought and behavior cannot be inferred from the language alone, but on the basis of a sociolinguistic description of its place in social and personal life (see Hymes 1966).

Verbal art—poetry, narration, oratory, rhetoric, dialectic—is universally made possible by language. A language, indeed, may be viewed as an aesthetic product (Sapir 1921). The forms possible to a verbal art are conditioned by the language, which is also the indispensable means to their study. With the new interest in metrics, poetics, and stylistics as approached through structural linguistics, and with the aid of folklore and anthropology, a truly comparative literature, global in scope, may emerge. And, if seen as not a matter of forms and texts alone but of symbolic action as well, verbal art becomes of special interest to the human sciences generally [see Drama; Interaction,article on Dramatism]. The aspects of human nature that underlie the universality of verbal art; the extent to which abilities are cultivated or left dormant, and why; differences in the valuation of language as an aesthetic medium, relative to others (music, dance, ritual, plastic arts) and to other concerns; the structuring of performances and the possibilities that such structures show major areal groupings, express particular aspects of social life, and reflect particular conceptions of the uses of language—all such concerns require an approach through texts as situated in contexts.

The question of the origin of language has returned to prominence, even though theory in this field is in one sense a myth, a projection into time of assumptions on the essential nature of language and its meaning for man. Linguistics proper can prescribe the elements, generic to all languages, whose origins are to be accounted for. (There is the possibility also that some elements of the last stage of the emergence of true language might be recoverable genetically.) A theory of origin must draw on all possible lines of evidence—biological, psychological, archeological—within a theory of the evolution of man. Since language emerges within an ongoing communication system, it is crucial to specify the conditions of selection that would have been operative. Comparison of human communication systems with those of primates is indispensable; so also is comparison of human communication systems among themselves. Studies that analyze comparatively the uses of language are also necessary, in that they bring into view not only what language may be used for but also what it need not be used for in the transmission of culture and cooperative activity (e.g., some societies seem not to require language for hunting or transmission of tool-making traditions). It is likely that a very limited code, less than true language, sufficed in small homogeneous groups until relatively late in prehistory.

With regard to linguistic conceptions of the unity of man, three perspectives can be distinguished— one envisaging unity through a common origin in the past, one envisaging unity in terms of a common essential nature of language in every time and place, and one envisaging a prospective unity in the context of an emerging world society. The three perspectives are complementary, but the first was more prominent in the nineteenth century (although carried on today in the work of Swadesh and others); the second has been more prominent with the emergence of structural linguistics in this century; and the third is coming into prominence with the increased attention, theoretical and practical, to a sociolinguistic approach. For a long period of human history the differentiation and dispersion of languages was the dominant process, but reintegration and mutual adaptation of languages within more complex social systems have increasingly superseded it. Indeed, genetic differentiation may never occur again. The processes of reintegration and mutual adaptation have accelerated within the same period that general linguistics has emerged, and many of the varied phenomena that attract sociolinguistic attention are aspects of the development of a single modern world—the correlated standardization of national languages; acculturation of dialects and of whole languages; the emergence of pidgins and Creoles at the frontiers of mercantile activity and colonization; the efforts to construct rational international languages; the growth of language academies and bourgeois notions of correctness; the cultivation of intertranslatability among the languages of Europe; the challenge to the stable diglossia of older philological civilizations by proponents of “Low” forms of speech; the extension of writing and literacy; and so forth. If these phenomena are to be related within linguistic theory, that theory must approach communities, functions, uses, and adaptations in a way that indeed takes on the character of an evolutionary perspective.

In one view, to be sure, the concept of evolution does not apply to language after its origin. Certainly no subsequent biological selection is apparent (although suggestions concerning a genetic basis for a few sounds have been made). From the standpoint of sociocultural evolution the matter is differ- ent. In their makeup, use, and survival, languages have been part of the specific adaptations of societies to environments, cultural contacts, and internal changes. And if all languages are equivalent in fundamental structure and potential capacity, languages as actually developed and available to their users have come to differ in ways that correspond, in part, to general stages of the cultural history of man. One mark is development in terms of the metalinguistic function (language about language), as seen in the development of linguistics itself. Some grammatical features and phonological characteristics seem present at one or another level of sociocultural integration and not at others. Most clearly, the differentiation of society is necessarily accompanied by technical elaboration and differentiation in vocabulary and syntax of a novel order. Recently, such development has been carried to the point of providing the linguistic tools for universal science and an incipient world civilization. Science itself is a key factor, for the languages in which it can be conducted (the small subset that may be called “world” languages) share the novel obligation that there be a name for everything in the universe: botany must leave no plant unnamed, ornithology no bird, ethnology no tribe, and so forth. While mathematics and logic have become what may be considered “postlanguage” developments, it remains true that they must be interpreted in natural languages and can be interpreted only in a few of them. The ideal of universal translatability is most nearly realized in these languages (as languages into which translation is made). These languages of course confer no necessary superiority or advantage on any individual. A user of English may be less able to master experience verbally and less skilled in language use than a user of a language quite local in scope. Even so, these observations must be controversial as they stand, offending as they do the widespread belief in the equivalence of all languages in complexity and function (see Hymes 196la; 1964c). The practical importance of such observations is manifest, however, in many issues of education and language policy, both in industrialized and industrializing nations. It should be clear that a modern evolutionary approach to society and culture fails to be adequate, theoretically or practically, if it excludes language.

Linguistics will play a part in the social sciences in the future if only because language so often must be the means of access to other things. Interest in language for its own sake as an aspect of man and society will continue to be an integral part of anthropology and psychology, and, increasingly, of sociology. The novel contributions that linguistic results and linguistic methods can make will be a constant source of such interest, but if an integration within a larger field of linguistics is to be realized, the social sciences themselves will have to contribute results and methods to the study of language. The prominence of the terms “ethno-linguistics,” “psycholinguistics,” and “sociolinguistics” since World War II augurs such a trend. While each term mediates between linguistics and a particular discipline, the set in total mediates between linguistics and the social sciences as a whole, drawing the two together. The outcome of such a unity will be a linguistics that is truly the science of language.

Dell Hymes

BIBLIOGRAPHY

Bach, Emmon W. 1964 An Introduction to Transformational Grammars. New York: Holt.

Berko, Jean; and Brown, Roger W. 1960 Psycholinguistic Research Methods. Pages 517–557 in Paul Mussen (editor), Handbook of Research Methods in Child Development. New York: Wiley.

Bernardini, Antonio; and Righi, Gaetano (1947) 1953 II concetto di filologia e di cultura classica dal Rinascimento ad oggi. 2d ed. Bari (Italy): Laterza.

Biennial Review of Anthropology. → Published since 1955.

Bloomfield, Leonard (1933) 1951 Language. Rev. ed. New York: Holt.

Boas, Franz 1911 Introduction. Part 1, pages 1-83 in Franz Boas (editor), Handbook of American Indian Languages. U.S. Bureau of American Ethnology, Bulletin No. 40. Washington: Government Printing Office.

Boas,franz (1920) 1955 The Classification of American Languages. Pages 211–218 in Franz Boas, Race, Language and Culture. New York: Macmillan. → First published in Volume 22 of American Anthropologist New Series.

Bolgar, R. R. 1954 The Classical Heritage and Its Beneficiaries. Cambridge Univ. Press. → A paperback edition was published in 1964 by Harper.

Borst, Arno 1957–1963 Der Turmbau von Babel: Geschichte der Meinungen iiber Ursprung und Vielfalt der Sprachen und Volker. Vols. 1-4. Stuttgart (Germany): Hiersemann.

Cassirer, Ernst (1923) 1953 The Philosophy of Symbolic Forms. Volume 1: Language. New Haven: Yale Univ. Press.

Chomsky, Noam (1957)1964 Syntactic Structures. The Hague: Mouton.

Chomsky, Noam 1964 Current Issues in Linguistic Theory. Janua Linguarum, Series Minor, No. 38. The Hague: Mouton.

Chomsky, Noam 1965 Aspects of the Theory of Syntax. Massachusetts Institute of Technology, Research Laboratory of Electronics, Special Technical Report, No. 11. Cambridge, Mass.: M.I.T. Press.

Chomsky, Noam 1966 Cartesian Linguistics: A Chapter in the History of Rationalist Thought. New York: Harper.

Conference ON Language Universals, Dobbs Ferry, New York,1961 1963 Universals of Language: Report of a Conference. Edited by Joseph H. Greenberg. Cambridge, Mass.: M.I.T. Press.

Current Trends in Linguistics. → Published since 1963.

Dixon, Robert M. W. 1965 What Is Language? A New Linguistic Approach to Linguistic Description. London: Longmans.

Emeneau, Murray B. 1956 India as a Linguistic Area. Language 32:3-16.

Ervin-tripp, Susan 1964 An Analysis of the Interaction of Language, Topic, and Listener. Pages 86-102 in John Gumperz and Dell Hymes (editors), The Ethnography of Communication. American Anthropologist, New Series, Vol. 66, No. 6, Part 2. Menasha, Wise.: American Anthropological Association.

Ferguson, Charles A. 1959 Diglossia. Word: Journal of the Linguistic Circle of New York 15:325–340.

Ferguson, Charles A. 1966 National Sociolinguistic Profile Formulas. Pages 309–315 in UCLA Sociolinguistics Conference, Los Angeles, 1964, Sociolinguistics. Edited by William Bright. Janua Linguarum, Series Maior, Vol. 20. The Hague: Mouton.

Ferguson, Charles A.; and Gumperz, John J. (editors) 1960 Linguistic Diversity in South Asia: Studies in Regional, Social, and Functional Variation. Indiana Univ., Research Center in Anthropology, Folklore, and Linguistics, Publications, Vol. 13. Bloomington, Ind.: The Center.

Firth, John R. (1935) 1957 The Technique of Semantics. Pages 7-33 in John R. Firth,Papers in Linguistics, 1934–1951. Oxford Univ. Press.

Fischer, John L. 1958 Social Influence on the Choice of a Linguistic Variant. Word: Journal of the Linguistic Circle of New York 14:47-56.

Fishman, Joshua A. et al. 1966 Language Loyalty in the United States: The Maintenance and Perpetuation of Non-English Mother Tongues by American Ethnic and Religious Groups. Janua Linguarum, Series Maior, Vol. 21. The Hague: Mouton.

Garvtn, Paul L. 1959 The Standard Language Problem: Concepts and Methods. Anthropological Linguistics 1, no. 3:28-31.

Gleason, Henry A. (1955) 1961 An Introduction to Descriptive Linguistics. Rev. ed. New York: Holt.

Gleason, Henry A. 1964 The Organization of Language : A Stratificational View. Georgetown University, Washington, D.C., Institute of Languages and Linguistics, Monograph Series on Languages and Linguistics 17:75-95.

Gray, Louis H. 1939 Foundations of Language. New York: Macmillan.

Greenberg, Joseph H. 1948 Linguistics and Ethnology. Southwestern Journal of Anthropology 4:140–147.

Gumperz, John J. 1961 Speech Variation and the Study of Indian Civilization. American Anthropologist New Series 63:976–988.

Gumperz, John J.; and Hymes, Dell (editors) 1964 The Ethnography of Communication. American Anthropologist, New Series, Vol. 66, No. 6, Part 2. Menasha, Wise.: American Anthropological Association.

Hall, Edward T. 1959 The Silent Language. Garden City, N.Y.: Doubleday. → A paperback edition was published in 1961 by Fawcett.

Halliday, Michael A. K.; Mc!ntosh, Angus; and Strevens, Peter (1964) 1965 The Linguistic Sciences and Language Teaching. Bloomington: Indiana Univ. Press.

Harris, Zellig S. 1951 Methods in Structural Linguistics. Univ. of Chicago Press.

Hockett, Charles F. 1948 Implications of Bloomfield’s Algonquian Studies. Language 24:117–131.

Hockett, Charles F. 1958 A Course in Modern Linguistics. New York: Macmillan.

Hoenigswald, Henry M. 1960 Language Change and Linguistic Reconstruction. Univ. of Chicago Press.

Hymes, Dell 1959 Field Work in Linguistics and Anthropology. Studies in Linguistics 14:82-91.

Hymes, Dell 196la Functions of Speech: An Evolutionary Approach. Pages 55-83 in Frederick C. Gruber (editor), Anthropology and Education. Philadelphia: Univ. of Pennsylvania Press.

Hymes, Dell 1961b Linguistic Aspects of Cross-cultural Personality Study. Pages 313–359 in Bert Kaplan (editor), Studying Personality Cross-culturally. New York: Harper.

Hymes, Dell 1961c Alfred Louis Kroeber. Language 37:1-28.

Hymes, Dell 1962 The Ethnography of Speaking. Pages 13-53 in Anthropological Society of Washington, Anthropology and Human Behavior. Washington: The Society.

Hymes, Dell (editor) 1964a Language in Culture and Society: A Reader in Linguistics and Anthropology. New York: Harper.

Hymes, Dell 19641? Directions in (Ethno-) Linguistic Theory. Pages 6-56 in A. Kimball Romney and Roy D’Andrade (editors), Transcultural Studies of Cognition. American Anthropologist, New Series, Vol. 66, No. 3, Part 2. Menasha, Wise.: American Anthropological Association.

Hymes, Dell 1964c A Perspective for Linguistic Anthropology. Pages 92-107 in Sol Tax (editor), Horizons of Anthropology. Chicago: Aldine.

Hymes, Dell 1965 Methods and Tasks of Anthropological Philology (Illustrated with Clackamus Chinook). Romance Philology 19:325–340.

Hymes, Dell 1966 Two Types of Linguistic Relativity. Pages 114–158 in UCLA Sociolinguistics Conference, Los Angeles, 1964, Sociolinguistics. Edited by William Bright. Janua Linguarum, Series Maior, Vol. 20. The Hague: Mouton.

International Congress OF Linguists, Ninth, Cam-bridge, Mass.,1961 1963 Trends in Modern Linguistics. Edited by Christine Mohrmann et al. Utrecht (Netherlands): Spectrum.

Ivie, Milka 1965 Trends in Linguistics. Janua Linguarum, Series Minor, No. 42. The Hague: Mouton.

Jakobson,roman 1960 Closing Statement: Linguistics and Poetics. Pages 350–373 in Conference on Style, Indiana University, 1958, Style in Language. Edited by Thomas A. Sebeok. Cambridge, Mass.: Technology Press of M.I.T.

Jakobson, Roman 1963 Efforts Towards a Means-Ends Model of Language in Inter-war Continental Linguistics. Pages 104–108 in International Congress of Linguists, 9th, Cambridge, Mass., 1961, Trends in Modern Linguistics. Edited by Christine Mohrmann et al. Utrecht (Netherlands): Spectrum.

Joos, Martin (editor) 1957 Readings in Linguistics: The Development of Descriptive Linguistics in Amer- ica Since 1925. Washington: American Council of Learned Societies.

Katz, Jerrold J.; and Postal, Paul M. 1964 An Integrated Theory of Linguistic Descriptions. Cambridge, Mass.: M.I.T. Press.

Kroeber, A. L. 1955 Linguistic Time Depth Results So Far and Their Meaning. International Journal of American Linguistics 21:91-104.

Labov, William 1963 The Social Motivation of a Sound Change. Word: Journal of the Linguistic Circle of New York 19:273–309.

Labov, William 1964 Phonological Correlates of Social Stratification. Pages 164–176 in John Gumperz and Dell Hymes (editors), The Ethnography of Communication. American Anthropologist, New Series, Vol. 66, No. 6, Part 2. Menasha, Wise.: American Anthropological Association.

Lamb, Sydney M. 1966 An Outline of Stratificational Grammar. Unpublished manuscript. Longacre, Robert E. 1964 Grammar Discovery Procedure: A Field Manual. Janua Linguarum, Series Minor, No. 33. The Hague: Mouton.

Lounsbury, Floyd G. 1953 Field Methods and Techniques in Linguistics. Pages 401–416 in International Symposium on Anthropology, New York, 1952, Anthropology Today: An Encyclopedic Inventory. Univ. of Chicago Press.

Mclntosh, Angus; and Halliday, M. A. K. 1966 Patterns of Language. London: Longmans.

Malkiel, Yakov 1964a Some Diachronic Implications of Fluid Speech Communities. Pages 177–186 in John Gumperz and Dell Hymes (editors), The Ethnography of Communication. American Anthropologist, New Series, Vol. 66, No. 6, Part 2. Menasha, Wise.: American Anthropological Association.

Malkiel, Yakov 1964b Distinctive Traits of Romance Linguistics. Pages 671–683 in Dell Hymes (editor), Language in Culture and Society: A Reader in Linguistics and Anthropology. New York: Harper.

Malkiel, Yakov 1966 Is There Room for “General Philology”? Pacific Coast Philology 1:3-11.

Marrou, Henri I. (1948) 1956 A History of Education in Antiquity. London: Sheed & Ward. → First published in French.

Martinet, Andre (1960) 1964 Elements of General Linguistics. Univ. of Chicago Press. → First published in French.

Martinet, Andre 1962 A Functional View of Language. Oxford: Clarendon.

Metcalf, George 1964 The Indo-European Hypothesis in the 16th and 17th Centuries. Paper prepared for Burg-Wartenstein Symposium, 25. Unpublished manu-script.

Mohrmann, Christine et al. (editors) 1962 Trends in European and American Linguistics, 1930–1960. Utrecht (Netherlands): Spectrum.

Pedersen, Holger (1924) 1962 The Discovery of Language: Linguistic Science in the Nineteenth Century. Bloomington: Indiana Univ. Press. → First published in Danish.

Pike, Kenneth L. 1954 Language in Relation to a Unified Theory of the Structure of Human Behavior. Part 1. Preliminary ed. Glendale, Calif.: Summer Institute of Linguistics.

[Report of the Fifteenth Annual Round Table Meeting on Linguistics and Language Studies.] 1964 Georgetown University, Washington, D.C., Institute of Languages and Linguistics, Monograph Series on Languages and Linguistics 17.

Robins, Robert H. 1951 Ancient & Mediaeval Grammatical Theory in Europe With Particular Reference to Modern Linguistic Doctrine. London: Bell.

Robins, Robert H. 1964 General Linguistics: An Introductory Survey. London: Longmans.

Romney, A. Klmball; and D’Andrade, Roy Goodwin (editors) 1964 Transcultural Studies in Cognition. American Anthropologist, New Series, Vol. 66, No. 3, Part 2. Menasha, Wise.: American Anthropological Association.

Samarin, William 1967 Field Linguistics. New York: Holt.

Sandys, John E. (1903–1908) 1958 A History of Classical Scholarship. 3 vols. New York: Hafner.

Sapir, Edward A. 1921 Language: An Introduction to the Study of Speech. New York: Harcourt.

Sapir, Edward A. (1924) 1949 The Grammarian and His Language. Pages 150–159 in Edward A. Sapir, Selected Writings in Language, Culture and Personality. Edited by David G. Mandelbaum. Berkeley: Univ. of California Press.

Sapir, Edward A. 1925 Sound Patterns in Language. Language 1:37-51.

Sapir, Edward A. 1929 The Status of Linguistics as a Science. Language 5:207–214.

Sartre, Jean Paul (1960) 1963 Search for a Method. New York: Knopf. → First published in French. A British edition was published as The Problem of Method.

Swadesh, Morris 1951 DifFusional Cumulation and Archaic Residue as Historical Explanations. Southwestern Journal of Anthropology 7:1-21.

Vachek, Josef (1964) 1966 A Prague School Reader in Linguistics. Bloomington: Indiana Univ. Press.

Vachek, Josef 1966a The Linguistic School of Prague: An Introduction to Its Theory and Practice. Bloomington: Indiana Univ. Press.

Vachek, Josef (editor) 1966b Les problemes du centre et de la périphérie du système de la langue. Travaux linguistiques de Prague, 2. Prague: Editions de 1’Academic Tchecoslovaque des Sciences.

Verburg, Pietr A. 1952 Taal en functionaliteit: Een historisch-critische studie over de opvattingen aangaande de functies der taal vanaf de prae-humanistische philologie van Orleans tot de rationalistische linguistiek van Bopp. Wageningen (Netherlands): Veenman.

Weinreich, Uriel 1953 Languages in Contact: Findings and Problems. New York: Linguistic Circle of New York.

Wonderly, William L.; and Nida, Eugene A. 1963 Linguistics and Christian Missions. Anthropological Linguistics 5, no. 1:104–144.

II HISTORICAL LINGUISTICS

Although consecrated by over a century and a half of use, the term “historical linguistics,” as a designation of a discipline, is something of a misnomer, because the most exciting and controversial operations of that discipline concern the reconstruction of language, i.e., prehistory, rather than documented history. For this reason, perhaps, Saussure, in his search for a label that would neatly contrast with the newly discovered “synchronic” perspective, suggested the qualifier “diachronic,” which, possibly as a result of its paleness, later proved less than successful (1916). In the mid-twentieth century, it might be most apposite to speak of “genetic linguistics” in reference to the entire domain, reserving the alternative designation “glottodynamics” for the hard core of general doctrine governing the analyst’s major operations.

Historical linguistics is very often equated with “comparative linguistics”; to the extent that the tracing of genetic relationships involves some confrontation of an earlier with a later stage of the same language (of Old English, say, with Middle or Modern English), a measure of overlap is indeed unavoidable. For practical purposes, however, it seems advisable to refer to comparative linguistics only where several cognate languages—ideally, they should be observed at the same time level—are jointly analyzed in an effort to arrive at the parental tongue, as when proto-Central Algonquian is reconstructed from available records of Sauk and Fox, Cree, Menomini, and Ojibwa. Of course, it is equally legitimate to engage in the typological comparison of languages with no thought of historical reconstruction and regardless of the presence of any kinship ties—see Bally’s classic confrontation of Modern German and Modern French (1932) and the currently fashionable “contrastive” grammars.

Historical and comparative linguistics reached their first peak of development in the nineteenth century, although there were some rudimentary attempts in western and central Europe from 1500 to 1800. Language history, in contrast, represents a relatively new genre of research; its roots are in broad-gauged introductory chapters to technically worded historical grammars. In terms readily understandable to layman and beginner alike, language history interweaves austere linguistic analyses with discussions—rarely devoid of grace—of social, economic, broadly cultural, demographic, and literary conditions prevailing at the successive time levels, allowing also for remarks on the philological state of transmission. At its best, as in Migliorini’s masterpiece (1960), language history excels at tracing the vicissitudes of a single language viewed within the matrix of the corresponding highly literate national culture.

The individual facts ascertainable through the various analyses devised by historical linguists lend themselves to two entirely unrelated kinds of synthesis. Certain language forms can be lifted out of their original philological context (which alone, in most instances, made their secure identification possible) and can be arranged on a higher plateau of abstraction so as to illustrate broad aspects of a specific linguistic transformation, e.g., the development of sounds, derivational molds, lexical meanings, or syntactic structures from stage A to stage B of the given language. Climbing to a still higher level of generalization, the analyst is at liberty to abandon even the context of the specific language at issue and to cite the modifications observed, for the sake of their illustrative value, in a general methodology of linguistic change. On the other hand, the slivers and nuggets of information obtained through stringent linguistic (in particular, etymological) analysis may be deftly inserted, as highly prized items, in the grandiose mosaics pieced together by patient and versatile historians. These items of information are similar to the fragmentary bits of knowledge collected by physical anthropologists, archeologists, folklorists, and others who attempt to recapture the elusive past.

Traditionally, from the days of such pioneers as the Germanist Jacob Grimm and the Romanist Friedrich Diez to that towering Indo-Europeanist of the early twentieth century, Antoine Meillet, the two conspicuously parallel tools of research in diachronic linguistics have been the manual of historical grammar and the etymological dictionary— the one providing a tightly ordered macrocosm and the other a loose kaleidoscopic array of microcosms. The full-sized historical grammar—not infrequently a multivolume venture—embraced phonology (with excursuses into prosody or accentology), inflection, “word formation” (i.e., affixal derivation and composition or their counterparts in non-Indo-European languages), and syntax. These centered, in ever widening circles, on the word, the phrase, and the sentence. An abridged version was limited to phonology and inflection. Inflection and the “syntax of the word” are so closely adjacent that they tend to merge, and a few scholars have gone so far as to consolidate all of morphology and syntax into the single domain of “morphosyntax,” which forms the hard, inalienable kernel of linguistics. Excursions into semantics, metrics (also, through the inclusion of tropes, rhetoric, or poetics), and stylistics—the last-mentioned more loosely organized and defined in a variety of ways—have at all times been regarded as optional rather than obligatory. Only in recent decades have grammatical and lexical studies drifted apart so sharply in techniques and appeal as to render problematic any joint ventures in the immediate future.

The relative stabilization of historical linguistics in the period 1850–1925 had the advantage of producing a far-reaching standardization in its termi- nology; this, in turn, by virtue of the comparability achieved, has invited and furthered at every step the confrontation of older and more recent studies, a procedure that has become more difficult in the last three decades. The long-unchallenged preeminence of central European scholarship in this field is mirrored by the wide acceptance of such technical terms as “umlaut” (metaphony) and “ablaut” (apophony), while other German labels, potentially just as helpful (e.g., Lehnwort “assimilated borrowing” versus Fremdwort “unassimilated borrowing”), have enjoyed no such popularity. Early standardization was particularly beneficial in certain special types of nonverbal symbolization, e.g., quotation marks for meaning, italics for quoted forms, boldface for transliteration into another alphabet, small capitals for an ancestral language (e.g., Latin versus Romance vernaculars), large capitals for epigraphic material, square brackets for phonetic transcription, asterisks for hypothetical forms, and, above all, the two directional signs: > “changes into” and < “descends from.”

Before long, the success of these symbols led to a temporary staleness, except where the stagnation was relieved by the introduction of signs manufactured by the more aggressively imaginative structuralist school (e.g., slanted lines for “phonemicization”). Thus, few historically oriented scholars have bothered to discriminate typographically between two logically distinguishable hypothetical forms: (a) those undocumented, yet assumed to have existed (*) and (b) those expressly presented as non-existent (*). Again, although few experts would deny the sharp cleavage between phonology and morphology, historical linguists have failed to capitalize on the possibility of contrastive symbolization of phonological versus morphological shifts.

To the extent that genetic linguists are concerned with historical situations, unique by definition, they can resort to the device of “model formation” only on a limited scale. In a way, any reconstruction of genetic relationship between languages or dialects involves a generous measure of schematization aimed at eliminating those details that would tend to blur the broad outline. One can visualize an entirely different kind of model: instead of focusing attention on concrete territories (at historical stages) or avoiding any commitment to the speakers’ habitat (at prehistoric stages), the analyst may decide to invent imaginary countries with a sharply profiled distribution of coastlines, waste-lands, mountain chains, ports of entry, emporia, cultural shrines, etc. He can further posit a certain succession of political, socioeconomic, and strictly linguistic events (say, invasions, retreats into the hilly inland, reconquest of coastal lowlands, splits into dialects) and project them onto the imaginary area, excogitating in abstract terms the likeliest concatenations of linguistic reactions to these pressures. By sharpening the analyst’s alertness to possible and probable intricacies under artificial conditions that are relatively simplified, such schemata can prepare him for successful inquiries into real-life situations, incomparably more complex.

It should be emphasized that the postulate of historical uniqueness is not easy to reconcile with the search for evolutionary universals in the realm of language. However, the prospect of discovering such universals has for many decades been a source of constant titillation. One classic example is the often-observed correlation of word order (and comparable syntactic devices) with the available wealth of inflectional endings. Clarity and economy demand that, if relationships between members of a clause can no longer be expressed unequivocally by means of the endings (e.g., as a result of phonetic erosion), a stiffening of word order should provide an adequate substitute. Also, etymologists have discovered that, of all form classes, adjectives, on balance, tend to present lexical nuclei most resistant to identification. In addition, semanticists report that fluctuations and changes of meaning undergone by verbs exceed, as a rule, those to which a typical noun would be subject. The difficulty with trying to establish absolute universals is that each such attempt presupposes the testing of hundreds of languages. On the other hand, characteristic samples would suffice to identify tendential universals.

As in all evolutionary sciences, the question of purposeful, or oriented, change is at the heart of the philosophy underlying any genetic analysis of linguistic data. Linguists are sharply divided on this matter of teleology: the great Danish theorist Otto Jespersen and the founders of the Prague school categorically affirm the teleological principle (a few visualize a trend toward general improvement achieved through refinement, simplification, and economy); others, particularly Bloomfield (see 1933) and a whole generation of American linguists claiming allegiance to his doctrine, just as vehemently deny it. Discernibly different from the teleological approach, although occasionally confused with it, especially by opponents, is the idealistic slant (characteristic of Benedetto Croce’s school in Italy and Karl Vossler’s in Germany), which stresses the primacy of the speakers’ thinking over their speech habits and grants them in the process a much wider margin of initiative and of control over linguistic change than would be ac- cepted by believers in the pre-eminence of “blind forces” or those (such as Whorf) who view the configuration of a grammatical structure as a prime determinant of thinking and perception. The more literate the speaker and especially the writer, the stronger the case for the idealistic approach; in analyzing “graphemically” the comportment of ancient and medieval scribes, one can hardly fail to distinguish between what they aimed to achieve and what they actually accomplished.

Descriptive and historical linguistics

Basic to all operations in historical linguistics is the view which the analyst holds of the configuration of the speech community under study and of speech communities in general. This was clearly sensed by Bloomfield, who, in his influential book Language (1933), without disregarding the varying density of communication or denying the complexity of certain speech communities, impressed upon his readers the need to reckon with a far-reaching uniformity of speech habits. Similarly, in presenting the comparative method, he leaned toward favoring those situations that exhibit clearcut dialect splits, without denying occasional alternatives. However, many younger scholars have recognized that the link holding together language communities is frequently mere similarity rather than actual homogeneity of speech habits, a point fraught with major implications for the geneticist. It is further held that bilingualism and even trilingualism are more widely disseminated the world over than is strict monolingualism, an assumption that demands flexibility in dealing with a multitude of diversified and changing situations. Thus, two groups speaking language X—one composed of members who have from infancy also mastered language Y and the other containing persons who happen to be constantly using language Z in certain social contexts—are unlikely to react identically to any incipient innovations spreading from a monolingual zone. (One also readily conceives of innovations arising at the intersection of languages.) One final argument in favor of fluidity in the object observed and elasticity in the method applied to its elucidation is the discovery that many areas commonly assigned en bloc to certain languages often lack such “natural boundaries” as might preserve a community of speakers in quasi-hermetic isolation. The emergence of such zones is due rather to conflation, i.e., to successive reapportionments of neighboring territories, each initially sheltering a different language or dialect. Therefore, unless perfect leveling subsequently en-sues, one may detect beneath the present-day “roof” bracketing the dialects remnants and splinters of their original phonic, grammatical, and lexical systems in almost kaleidoscopic confusion.

In linguistics, the relation of the descriptive (or synchronic) to the historical (or diachronic) perspective has been the subject of considerable speculation and discussion, the consensus being that descriptive analysis bears preponderantly on simpler, less opaque situations. From this nearly unanimous opinion several discrepant conclusions may be drawn. Some experts maintain that new techniques, such as the structural method, should be tried out first on horizontal, later on vertical, slices of linguistic material. There are those who visualize a historical structure as a succession of descriptive structures superimposed on one another. The main difficulty in designing such an edifice lies in the fact that certain features structurally less than significant at one evolutionary stage may suddenly acquire conspicuous importance during transition to the next stage. Thus, the descriptivist is free to assert that in words like danc-er, kill-er, the ending -er as the carrier of an identifiably specific meaning (“agent”) represents a derivational morpheme, while in hamm-er, pinc-er, rudd-er the same sequence of sounds plays no comparable role. But the historical linguist, while acknowledging this distinction on certain temporal plateaus, must at all times reckon with the strong possibility of joint actions, inextricably interwoven, by homonymous genuine suffixes (such as the -er of killer) and mere suffixoids (the -er of rudd-er). One consensus is worth mentioning: from the minute inspection of any given state of a single language the experienced analyst can tentatively extract almost as much information on its earlier stages (“internal reconstruction”) as he can from comparing that language “externally” with its congeners.

One way of doing justice to both perspectives has been to engage in a “stairway projection”; among the practitioners of this novel approach one may count such seasoned experimenters as Otto Jespersen (for English), Antoine Meillet (for Latin), and Walther von Wartburg (for French). This particular method of intricate surgery affords glimpses of the consecutive periods of the chosen language, slanted alternately in the descriptive and in the historical direction. The implication of this design is clearly that one may distinguish between periods of relative rest and stability and others marked by spells of stress and strain.

One salient difference between the descriptive and the historical approach in linguistics is that the former in most instances enables the researcher to operate with a finite corpus, an intentional se- lection over whose scope he retains a modicum of control, while the latter often bears on an irremediably fragmentary volume of data. The ability to work with lacunary material and a certain flair for filling in gaps thus become important prerequisites for success in linguistic reconstruction, just as they are for research in geology, paleontology, and paleobotany. Developments are contrastively symbolized by solid lines (documented) or broken lines (hypothesized); however, the latter do not invariably represent initial, prehistoric segments of trajectories. An archaic stage A may very well owe its transparency to the realistic, readily adjustable spelling habits of the scribes concerned; conversely, stage B, although temporally closer to the beholder, may become nebulous, because the scribes of that period, plagued by conservatism, or subject to an inferiority complex, may have endeavored stubbornly to cling to the orthographic norms of their predecessors (“etymological spelling”), while the actual speech processes ran their course with un-abated speed; then again, stage C may mark a vigorous return to graphic realism, entailing the relative translucency of actual speech events. A vivid illustration of these three phases is provided by early Latin, late Latin, certain varieties of ’low” and medieval Latin, and, on an overwhelming scale, the budding Romance vernaculars.

It also happens that some word of unmistakably Latin provenience which, judging from its “normal” transformations, must have been in constant use over two millennia, disappears from written records in the fifth century, only to re-emerge a thousand years later. In cases of this kind, the literary genres of the extant texts act as prisms or filters, often seemingly capricious. They may long repress a word, keeping it submerged until there arises some opportunity—socially or aesthetically controlled—for its definitive surfacing into the standard language.

Trajectories of linguistic change

Systematic inquiry into the configuration of trajectories has not yet outgrown the stage of trial and error. Where regular phonological change occurs, older notions of strictly gradual transitions do not apply. Between, say, the Latin ū, as in pūru, and the French ū, as in pur, it is no longer admissible to posit an infinity of intermediate nuances of the stressed vowel without concurrently accepting some kind of cutoff point at which a vowel already markedly fronted, but still representing no more than an unusual variant within the phoneme /ū/, must have become a member, decreasingly erratic, of the sound family constituting the phoneme /ū/.

In other words, structural thinking forces us to recognize the interaction of slow-working phonetic rapprochements and more or less sudden occasional jumps. This composite schema guarantees the semblance of a close-knit system to a language at any moment of its growth. Thus, the graduality of development—not superseded, but only qualified and hierarchized—remains a vitally important assumption. Significantly, the hypothesis that the shift ū > ū, eminently characteristic of the transformation of provincial Latin into French, is traceable to the contributing agency of Gaulish cannot be refuted by the argument that the Celtic language in question lacked a fully developed /ū/ in its own system. It would have sufficed for the local substratum language, at the start, to have slightly deflected the Latin ū from its original status of a high back vowel in the direction of ū, thus producing a kind of chain reaction or even an accelerated advance along a straight line.

In yet another context, the configuration of a trajectory of linguistic change, properly interpreted, may be revealing. If the changes due to associative interference were to be plotted on a chart, some of them might give the impression of a bizarre zigzagging curve. On such a chart a relatively level line may suddenly start climbing as a result of an outside pressure, a “disturbance,” until it reaches a certain peak. Then the language’s inner mechanism (e.g., the total weight of its inflectional paradigms) may begin to wipe out the irregularity, causing the line to drop until it reverts to its original direction. If in such an up-and-down movement, anteceding the advent of trustworthy written texts, the descending stage completely absorbs the effects of the ascending stage, it is quite impossible to detect the original disturbance. If the down movement falls short of counterbalancing the aberrancy or over-corrects it, there is bound to remain in its wake a residue of startling “exceptions.” In case the irregularity happens to erupt at the very start of the written tradition, it may appear baffling in retrospect that the ancestral language and its eventual modern product should be in perfect mutual agreement while at such sharp variance with the intermediate step, which then, in fact, fails to perform any “mediation.” Thus, the Latin third person singular imperfect ending -ibat (originally -iebat) cast off in early Romance speech -ia(t), which to this day survives in Spanish as -ia, a safely predictable form, but paradoxically it produced instead, in early Old Spanish, -ié, a variant difficult to reconcile either with its antecedent or with its sequel. Investigation (Malkiel 1959; 1964) has disclosed that the rise of -ié simply marks a minor temporary deflection (of ascertainable origin), ultimately neutralized, while the later form ia represents far more faithfully the continuation of a basic trend.

Closely allied to the concern with the convolution of trajectories is the attempt, assuming a certain more or less steady rate of attrition in the core lexicon, to draw from parallel analyses tentative conclusions as to the degree of kinship between congeneric languages and the approximate date of their split. This approach, which rode the crest of a temporary vogue in the 1950s, has become known, broadly, as lexicostatistics and, with special application to dating, as glottochronology. Exaggerated claims, especially the attempt of some practitioners to place these techniques on the same pedestal as the rigorous study of sound correspondences, have led to quick disenchantment and virtual abandonment of the method.

Sound change

For better or worse, the vicissitudes of historical linguistics have been intimately linked with the theories of sound change. The recognition of regularity in these transmutations has been hailed as a milestone along the road to progress (cf. the radical programmatic statements of the “neogrammarians,” or Junggrammatiker, circa 1870) or, more intransigently, as a touchstone of stringent scientific thinking. In the compressed classroom presentation of historical linguistics, lecturers have traditionally inclined toward concentrating on “regular sound changes” as the discipline’s irreducible hard core. In the separate quarters of humanists and anthropologists alike, this rigid attitude has for decades contributed toward producing the impression of linguistics as a highly abstract subject, almost forbiddingly abstruse and, above all, divorced in its style and tone from cultural history, to say nothing of its aloofness from the realm of arts and letters. Moreover, because most provisional rules or “laws” admitted of a few exceptions and some of countless ones, there was for a while a widespread apprehension that the “regularists” were actually propounding some kind of mock science.

In reality, there exist several categories of sound change, each fairly autonomous—but not entirely so—and tied to diverse facets of human comportment and different levels of a speaker’s consciousness. The immediate goal of linguists is to discover one workable formula for presenting these interconnections, however tenuous, and another for dis-covering the elusive ties of sound change categories to discrete mental processes.

The techniques of accurately circumscribing individual sound correspondences that are inherently limited in time and space can be traced to the nineteenth century. By contrasting the French mer (“sea”) and pére (“father”) with their Latin proto-types mare and patre, the analyst learns that the Latin a tended, by and large, to yield e in French. Further refinement of this first approximation is within easy reach. The discrepant first vowels in pére and parrain (“godfather”) (originally parrin, from patrlnus) alert the observer to the possibility that the shift a > e hinges on a crucial accentual condition, while comparison of pére < patre with part < par-te dramatizes the share that the configuration of the stressed syllable may have had in an obvious bifurcation, depending on whether that syllable ended in a vowel or a consonant. By examining with scrupulous care all seemingly aber-rant developments (amenable to observation with optimal results in Old French), the analyst isolates, step by step, the specific phonological (“internal”) factors that must have presided over the evolution, erratic at first glance, of (il)lāc (“there”) >lá paupere (“poor”) > povre (modern pauvre); clauu (“nail”) > clou; aqua (“water”) > eaue (modern eau); palus (“pole”) > pieus; caput (“head”) ‾ ch(i)ef, etc. Comparably detailed breakdowns can be established for all other Latin sounds viewed in their transmission into a chosen “daughter language,” and, by way of effective control, the linguist is free to reverse the perspective and select as a given the basic sound unit of the daughter language, assigning to himself the task of individuating its sources.

But this classification marks only a first step, yielding at best a tidily subdivided inventory of raw facts. The preliminary classification is nonexplicative, lacks statistical underpinning, fails to throw into bold relief parallels, convergences (including some that are partial or have been arrested), and, worse, concatenations of events, and does not begin to take into account other forms of sound shift. Such taxonomy disregards several broad or distinct categories of internal linguistic change. Moreover, it is not elastic enough to do justice to the various external pressures (demographic, social, educational) on evolutionary trends in speech and in the written word as well. It is in these directions— many of them affording fruitful contacts with a whole spectrum of other disciplines—that the chief advances of late-twentieth-century research are bound to lie.

The following are a few illustrations of research in progress and tempting prospects of investigation. Alongside regular phonetic change (akin to Sapir’s “drift”) scholars have placed sporadic shifts (also called spontaneous and saltatory), such as metathesis, the transposition of a sound or intermuta- tion of two sounds, in contact or at a distance, haplology, the elimination of one or two successive segments partially identical, and certain dissimilatory processes. None of these, it has been argued, is confined to a specific locus or span of time, i.e., they are all, at least latently or tendentially, pantopic and panchronic. Granted the fundamental validity of this distinction, there arise several questions and second thoughts. Does the sound system of a language at a given stage—or, alternatively, its pattern of regular changes—typically stimulate or block sporadic shifts, or does it let them take their own course? Could it be true that, for all their uniqueness, regular sound changes, in any random selection, display such strong proclivities in a few characteristic directions that one discerns in them certain universals? Is it legitimate to grade the regularity of sound change (not as an ideal postulate, but as a bit of reality) and to contrast “strong” expectation of outcome, most likely to occur in monolithic societies, with “weak” predictability, attributable to, say, loose conflations of dialects, regional or social? Does such a state of prevalent weakness tend to intensify sporadic shifts and to invite even an excess of lexical contamination? Should frequency of lexical occurrence, or at least of incidence in actual speech, rank as a factor contributing to the degree of regularity, especially where a particularly unusual sequence of sounds falls into no broader pattern of immediate appeal? Can the exigencies of inflectional patterning slow down the pace of a sound change or counteract it to the point of weakening a phonetic “law”? Do other demands of this kind carry sufficient weight to set in motion or to accelerate potential and, especially, incipient sound changes? If the answer to the last two questions is affirmative, can one uphold the view that phonology operates in practically hermetic isolation? Specifically, is it still permissible to resolve the phenomena of genetic phonology into a neat interplay of sound relationships—to be precise, into an alternation of states of equilibrium and states of unrest or tension, to the virtual exclusion of all rival forces? Does it make sense to arrange sound changes in their presumed chronological succession (Richter 1934) without explicit forewarning that such sequences neither invariably presuppose nor necessarily imply the flow of one change from another or from the sum of all others already completed? Can one, in such contexts, ignore with impunity certain extraneous factors such as pressure of morphological paradigms and deflections from the straight course through associative lexical interference?

From earlier incidental mention it is clear that there are other kinds of change affecting linguistic form and, consequently, reflected in the sounds as the obligatory carriers of that form but not here caused by purely phonetic conditions. The most important of these supervenient categories of change is analogical. Speakers make adjustments bearing either on the configuration of a grammatical paradigm or on the shape of a single word; in the latter eventuality, both the radical and the affix are open to modification. Typically, such adjustments follow upon sound change; only by way of exception may one suspect them of impinging, as prime movers, upon sound development. Since analogical changes involve, by definition, associative interference, they seem to occur on a higher level of awareness than straight sound changes; thus they invite psycholinguistic analysis.

Sound symbolism constitutes yet another autonomous category, of slightly controversial status. To the extent that sounds, in symbolic context (and nowhere else), are credited with conveying messages of their own, this marginal category represents a tenuous bridge to semantic change, ordinarily removed from the realms of articulation, acoustics, and auditory perception. Sound symbolism may be absolute or relative. The former category prevails if the analyst attaches, cross-linguistically, an unvarying evocative value to, say, a high front vowel or to a hissing prepalatal consonant; the problem then is to ascertain whether speakers will allow words endowed with major connotative force, through such ingredients, to participate in normal sound shifts at the cost of heavy loss in suggestiveness. The effects of relative sound symbolism are conditioned by the given phonological system; thus, in a language generally averse to long consonants an occasional geminate may boast “expressive” value (which it would otherwise lack). Again, the language historian is curious to learn how speakers can maintain a word enhanced by such a feature in this privileged status amid the welter of pervasive transformations. At this juncture one welcomes contact with information theory.

Pressures for linguistic change

Entirely different from the classes of change are the categories of forces that are apt to produce changes of any kind. But the linguist’s operational procedure in tackling this new problem remains essentially unaltered: again his dual task is first to isolate the forces in question and then to discover the closest available approximation to the formula for their interplay.

It is customary to divorce the internal from the external forces at the outset, notably because the separate inquiries into them seem to appeal to radically different minds. In the former group one can readily distinguish two drives, sometimes acting in polar opposition—one toward economy of effort, the other toward clarity. Economy, syntagmatically conceived, aims at the speech act; in paradigmatic perspective economy relates to the acquisition of, and sustained command over, neuromuscular skills. The former type determines the course of most assimilatory processes, contextual by definition (e.g., Latin actu > Italian atto) and governs the choice of those glides and buffer consonants that serve to smooth away troublesome contiguity (in Old Spanish viewed in its relation to Latin:hōnōrāre > onrar > on-d-rar versus femina > femna > fem-b-ra). The latter type precipitates mergers of phonemes in the system where continued distinction between them would produce only a meager yield (/ā/ and /ē/ in older Parisian, /ē/ and /cē/ with increasing momentum in present-day Parisian). It also dooms to extinction minute groups of words displaying an infrequent sound or combination of sounds.

A groping search for increased clarity may be behind most dissimilatory and haplologic processes. It accounts, as would no other supposition, for the speakers’ readiness to augment their vocabulary (in an effort to reduce lexical polysemy) and to accept longer and more cumbersome syntactic structures (in a recoil from ambiguity). It is perhaps at this point that the newly achieved refinement of transformational grammar would most benefit the classic researches conducted by geneticists. The same urge for increased clarity ultimately justifies the sometimes successful flight from harmful homonymy or from its mere threat—the nearest escape routes being substitute words borrowed from neighboring dialects and reinterpreted within one’s own cultural heritage, and words freely invented.

After one deducts the two-pronged quest for maximum economy and clarity, it is the residue left that threatens to cause serious difficulty; the wisdom of applying to it some such pleasing blanket term as “expressivism” remains to be demonstrated, especially since it is doubtful whether, in the last analysis, one can reduce the remaining forces to a single denominator. One nucleus that cannot by any stretch of the imagination be subsumed under either economy or clarity contains those formations associated with special moods— playful, tender, or festive. In contemporary English the colloquially flavored compositional types, such as hush-hush, ping-pong, riffraff, wishy-washy, pribbles and prabbles, topsy-turvy, mumbo jumbo—sometimes originating in the nursery and displaying a strong admixture of onomatopoeia—admirably fit this description. In Slavic and Romance languages, formations involving strings of hypocoristic suffixes would qualify as a counterpart. The Hebrew spoken in modern Israel, the twentieth century’s linguistic melting pot par excellence, allows speakers the jocose lapse into the Ashkenazic rather than the officially favored Sephardic pronunciation for proper names affectionately uttered, thus proliferating doublets. On the other hand, one runs across a phenomenon such as hypercharacterization (i.e., the sharper, more explicit, even un-economically generous marking of a major grammatical category—say, gender, number, or person). For instance, the Latin socrus (“mother-in-law”), hampered by its conspicuously uncharacteristic -us, is more neatly profiled with regard to gender (and sex) at the Romance stage through a new and far more appropriate ending (Italian,suocera; Spanish, suegra, etc.); cf. also the change of Latin puppis (“poop,” “stern”), marred by an ending indeterminate as to gender, into the clear-cut Spanish pop-a. In such instances, the change is, of course, analogical, but the driving force behind it seems less easy to identify. It certainly can be neither economy nor any overflow of emotion, and one is hesitant, to say the least, to press into service a quest for heightened clarity. The propelling force, one suspects, is the speaker’s endeavor to redesign selected portions of the language, to make the medium of transmission more pointed or so silhouetted as to be aesthetically more satisfying. Sapir’s reference to the “cut” of a language comes to mind here—a fait accompli or a goal toward which speakers may strive.

The most familiar external forces whose impact will produce the various types of linguistic change are those resulting from contact between languages (occasionally one living and the other dead, but preserved in ritual or intensely studied) or regional and social dialects. Typically, a protracted transitional period of thoroughgoing bilingualism or plurilingualism is needed before the contact produces sizable results. In gauging any such impact on a specific language, the historian tries first to determine the principal layer of that language by inspecting the core structure of its grammar and those ingredients of its lexicon best known for their resistance, if not total immunity, to infiltration. Numerals, kinship terms, names of parts of the body, and grammatically functional words are typical examples of such elements. Once this frame of reference has been established, it becomes clear which layers, in the course of further study, will be labeled substrata and which superstrata—eloquent metaphors borrowed from geology and permitting a graphic projection of anteriority. Thus, vis-a-vis Great Russian, the numerous Finno-Ugric languages, now extinct or pushed back to the periphery of eastern Europe, constituted substrata; so did Coptic and, farther down the Nile valley in Egypt, the Greek koine vis-a-vis Arabic, Frisian visa-vis Dutch in Holland, and French plus Canary Island Spanish vis-a-vis English in Louisiana. In the absence of any genuine symbiosis, it is doubtful whether American Indian languages may rank as substrata in relation to English in North America, as they indisputably do in relation to Spanish and Portuguese throughout Latin America. In the twi-light hour between late Antiquity and the Middle Ages, Arabic in southern Spain and Frankish in northern Gaul represented superstrata in relation to divergent varieties of provincial Latin. To the extent that the Greeks tended to form independent cultural nuclei under the aegis of the expanding Roman Republic and later Roman Empire, their settlements qualify as examples of linguistic adstrata vis-a-vis Latin as well as the circum-Mediterranean indigenous languages.

Aside from such “vertical” relations, linguistic pressures operate “horizontally” across political borders and even at long distances through cultural diffusion. Thus, heavy clusters of Gallicisms are found not only in Spanish, Catalan, Italian, German, and Dutch, to say nothing of English, but also in languages occupying nonadjacent areas, such as Rumanian, Polish, Russian, and Swedish. Words are more easily borrowed than sounds, and affixes travel more rapidly than inflectional endings. A classic example of superimposed syntactic, semantic, and (probably) intonational patterns is provided by the multifarious Germanisms observable in eastern Switzerland’s Romansh, a language descended from Raetian Latin.

Against the fairly trivial instances of direct, positive influence one may place the sorely neglected range of indirect or catalytic interferences. Thus, two early medieval Germanic kingdoms carved out of the ruins of the crumbling Roman Empire—that of the Suebi in Galician-Portuguese territory and that of the Burgundians in the Lyon-Geneva area —molded local Latin speech sparingly through loan words but exerted a powerful restraining influence by politically and culturally isolating their territories, at crucial junctures, from such centers of ceaseless linguistic innovation as Toledo and Paris.

An independent kind of external force comprises all sorts of nonlinguistic events potentially rich in linguistic reverberations. The invention of novel tools and machines may breathe new life into a moribund suffix serving to denote instruments. The emancipation of women the world over may develop dormant schemata (affixal, compositional, or otherwise derivational) for the designation of female agentials. A global vogue of formality or familiarity (in clothing, dwellings, human relations, etc.) could hardly fail to revolutionize the system of forms of address and to affect even personal pronouns and possessive adjectives. The sections of the linguistic edifice most vulnerable to these influences are, then, the vocabulary, the derivational machinery (at the midway point between lexicon and grammar), plus a few pieces from the morphosyntactic tool kit.

Far beyond this boundary, the “idealistic” school of thought, entrenched in Italy and Germany only a generation ago (Vossler 1925), tended to assess very liberally the impact of changing modes of thinking on linguistic forms, extending that impact to the foundations of sentence structure. While the consensus of most generations of scholars has ascribed the disappearance of case endings to attrition, recognizing the rise of prepositional para-phrases as a relatively smooth replacement, the “idealists” preferred to view as prime mover the emergence of a new way of thinking (such as analytic rather than synthetic), crediting it with the manufacture of appropriate substitutes which eventually eroded the older grammatical framework. The advent of Christianity figured in these interpretations (especially in H. F. Muller’s) as another favorite determinant of linguistic evolution. The idealistic position is thus diametrically opposed to that of Whorf, who, following Sapir (1921), mused that patterns of thinking may, in the first place, be molded by pre-existent grammatical structures.

The complex interaction of all these isolable forces can be illustrated with the differing, if reconcilable, answers to the classic question: What dooms a word to extinction? Plausible explanations offered either separately or in any number of free combinations include an excess of paradigmatic intricacy or phonological oddity in the fated word; the peril besetting the weaker of two conflicting homonyms; an intolerable dosage of polysemy; a sudden general demand at all levels of the given society for lexical rejuvenation or large-scale overhaul; the obsolescence of a specific cultural element (say, some container or garment) heretofore designated by the word at issue; the ineluctable effect of some socially controlled restriction (taboo, etc.); acceptance, through borrowing from the local prestige language, of a more attractive equivalent, as when the imported Cousine dislodged the native Base in eighteenth-century German.

Theories of linguistic change

For the projection of major phases of linguistic growth, and especially for signaling the relationship between cognate languages, experts in recon-struction have resorted either to the somewhat older “family-tree theory”(Stammbaumtheorie) or to the wave hypothesis. The former is associated with the name of Schleicher, that contemporary and counterpart of the evolutionist Darwin who actually refined rather than launched the “family-tree” concept (1861). It operates with a filiation chart reminiscent of those long favored in the life sciences. The filiation chart, germane in its verbalization and, even more, in its graphic suggestion to the physicists’ and chemists’ views of radiation, cannot be traced to any advocate earlier than Schuchardt (1866) and, in particular, Schmidt (1871). The inherently rigid family-tree diagram presupposes uniform speech communities and their sudden and clear-cut bifurcation. The more elastic wave diagram tends to dissolve any system (or, less orderly, any arsenal) of communication tools into its constituents, granting to each change, whether phonetic, morphosyntactic, or lexical, its own scope and history. Neither the lapse of time it demands nor the area it covers need be exactly identical with those involved in any comparable change. The latest thinking sees in these divergent hypotheses two complementary projections, neither satisfactory if applied in isolation. Regrettably, no theory apt to reconcile them and no technique capable of smoothly integrating their separate findings have so far been devised.

The wave theory has intrinsically tended to give unusual prominence to the territorial expansion of linguistic features, providing the logical justification for linguistic (or dialect) geography. The interest in dialect geography is now past its crest; for many decades it fed on its sentimental motivation, local patriotism, and its partisans’ delight in open-air field work. Practitioners of this approach developed special methods for interviews, oral or written questionnaires, and the cartographic recording of field notes (linguistic atlases). Dialect geographers endowed with historical flair then proceeded to transform the geographic patterns laboriously established into bolder chronological sequences, calling themselves the geologists, paleontologists, or stratigraphers of human speech.

One extreme formulation of these assumptions, tastes, and techniques (the “age—area hypothesis”) relies chiefly or even exclusively on territorial patterns in piecing together temporal successions. An attempt to schematize these procedures of “areal analysis” was undertaken by the small group of Italian “neolinguists,” a school that produced a short flurry of activity from 1920 to 1950. A point that has hitherto not been satisfactorily investigated and yet clamors for imaginative inquiry is the wisdom of positing, alongside that “outer radiation” dear to dialect geographers and to diffusion-is ts like Boas, some kind of “inner radiation” that might account in undulatory projections for the continuous restructuring of systems.

Yakov Malkiel

[See alsoHistory,article onCulture History; Linguistics, article onThe Speech Community; and the biographies ofBloomfield; Saussure; Whorf.]

BIBLIOGRAPHY

Bally, Charles (1932) 1944 Linguistique générate et linguistique francaise. 2d ed. Bern: Francke.

Benveniste, Meile 1966 Problemes de linguistique generale. Paris: Gallimard.

Bloomfield, Leonard (1933) 1951 Language. Rev. ed. New York: Holt.

Hymes, Dell 1960 Lexicostatistics So Far.Current Anthropology 1:3-44. → Includes eight pages of “Comments” and “References.”

Kurylowicz, Jerzy 1960 Esquisses linguistiques. Wro-claw (Poland): Zaklad Narodowy Imienia Ossolinskich.

Leroy, Maurice 1963 Les grands courants de la linguistique moderne. 2d ed. Brussels, Universite Libre, Faculte de Philosophic et Lettres, Travaux, Vol. 24. Presses Universitaires de Bruxelles. → Contains a reliable account of nineteenth-century research.

Malkiel, Yakov 1959 Toward a Reconsideration of the Old Spanish Imperfect in -ia ¨ -ie. Hispanic Review 27:435–481.

Malkiel, Yakov 1964 Initial Points Versus Initial Segments of Linguistic Trajectories. Pages 402–405 in International Congress of Linguists, Ninth, Cambridge, Mass., 1962,Proceedings. Janua linguarum, Series Maior, Vol. 12. The Hague: Mouton.

Martinet, Andre 1955 Economic des changements phonetiques: Traite de phonologic diachronique. Bern: Francke.

Migliorini, Bruno 1960 Storia della lingua italiana. Florence: Sansoni.

Richter, Elise 1934 Beitrage zur Geschichte der Romanismen. Zeitschrift fur romanische Philologie, Supplement 82. Halle (Germany): Niemeyer.

Sapir, Deward 1921 Language: An Introduction to the Study of Speech. New York: Harcourt.

Saussure, Ferdinand De (1916) 1959 Course in General Linguistics. New York: Philosophical Library. → First published (posthumously) as Cours de linguistique générale.

Schleicher, August (1861) 1874–1877 A Compendium of the Comparative Grammar of the Indo-European, Sanskrit, Greek and Latin Languages. London: Triibner. → Selections from August Schleicher’s Compendium der vergleichenden Grammatik der indogermanischen Sprachen.

Schmidt, Johannes 1871 Die Verwandtschaftsverha.ltnisse der indogermanischen Sprachen. Weimar (Germany): Bohlau.

Schuchardt, Hugo 1866 Der Vokalismus des Vulgtirlateins. Volume 1. Leipzig: Teubner.

Vossler, Karl (1925) 1932 The Spirit of Language in Civilization. London: Routledge. → First published as Geist und Kultur in der Sprache.

III THE SPEECH COMMUNITY

Although not all communication is linguistic, language is by far the most powerful and versatile medium of communication; all known human groups possess language. Unlike other sign systems, the verbal system can, through the minute refinement of its grammatical and semantic structure, be made to refer to a wide variety of objects and concepts. At the same time, verbal interaction is a social process in which utterances are selected in accordance with socially recognized norms and expectations. It follows that linguistic phenomena are analyzable both within the context of language itself and within the broader context of social behavior. In the formal analysis of language the object of attention is a particular body of linguistic data abstracted from the settings in which it occurs and studied primarily from the point of view of its referential function. In analyzing linguistic phenomena within a socially defined universe, however, the study is of language usage as it reflects more general behavior norms. This universe is the speech community: any human aggregate characterized by regular and frequent interaction by means of a shared body of verbal signs and set off from similar aggregates by significant differences in language usage.

Most groups of any permanence, be they small bands bounded by face-to-face contact, modern nations divisible into smaller subregions, or even occupational associations or neighborhood gangs, may be treated as speech communities, provided they show linguistic peculiarities that warrant special study. The verbal behavior of such groups always constitutes a system. It must be based on finite sets of grammatical rules that underlie the production of well-formed sentences, or else messages will not be intelligible. The description of such rules is a precondition for the study of all types of linguistic phenomena. But it is only the starting point in the sociolinguistic analysis of language behavior.

Grammatical rules define the bounds of the linguistically acceptable. For example, they enable us to identify “How do you do?” “How are you?” and “Hi” as proper American English sentences and to reject others like “How do you?” and “How you are?” Yet speech is not constrained by grammatical rules alone. An individual’s choice from among permissible alternates in a particular speech event may reveal his family background and his social intent, may identify him as a Southerner, a Northerner, an urbanite, a rustic, a member of the educated or uneducated classes, and may even indicate whether he wishes to appear friendly or distant, familiar or deferential, superior or inferior.

Just as intelligibility presupposes underlying grammatical rules, the communication of social information presupposes the existence of regular relationships between language usage and social structure. Before we can judge a speaker’s social intent, we must know something about the norms defining the appropriateness of linguistically acceptable alternates for particular types of speakers; these norms vary among subgroups and among social settings. Wherever the relationships between language choice and rules of social appropriateness can be formalized, they allow us to group relevant linguistic forms into distinct dialects, styles, and occupational or other special parlances. The socio-linguistic study of speech communities deals with the linguistic similarities and differences among these speech varieties.

In linguistically homogeneous societies the verbal markers of social distinctions tend to be confined to structurally marginal features of phonology, syntax, and lexicon. Elsewhere they may include both standard literary languages, and grammatically divergent local dialects. In many multilingual societies the choice of one language over another has the same signification as the selection among lexical alternates in linguistically homogeneous societies. In such cases, two or more grammars may be required to cover the entire scope of linguistically acceptable expressions that serve to convey social meanings.

Regardless of the linguistic differences among them, the speech varieties employed within a speech community form a system because they are related to a shared set of social norms. Hence, they can be classified according to their usage, their origins, and the relationship between speech and social action that they reflect. They become indices of social patterns of interaction in the speech community.

Historical orientation in early studies

Systematic linguistic field work began in the middle of the nineteenth century. Prior to 1940 the best-known studies were concerned with dialects, special parlances, national languages, and linguistic acculturation and diffusion.

Dialectology

Among the first students of speech communities were the dialectologists, who charted the distribution of colloquial speech forms in societies dominated by German, French, English, Polish, and other major standard literary tongues. Mapping relevant features of pronunciation, gram- mar, and lexicon in the form of isoglosses, they traced in detail the range and spread of historically documented changes in language habits. Isoglosses were grouped into bundles of two or more and then mapped; from the geographical shape of such isogloss bundles, it was possible to distinguish the focal areas, centers from which innovations radiate into the surrounding regions; relic zones, districts where forms previously known only from old texts were still current; and transition zones, areas of internal diversity marked by the coexistence of linguistic forms identified with competing centers of innovation.

Analysis along these lines clearly established the importance of social factors in language change. The distribution of rural speech patterns was found to be directly related to such factors as political boundaries during the preceding centuries, traditional market networks, the spread of important religious movements, etc. In this fashion dialectology became an important source of evidence for social history.

Special parlances, classical languages

Other scholars dealt with the languages of occupationally specialized minority groups, craft jargons, secret argots, and the like. In some cases, such as the Romany of the gypsies and the Yiddish of Jews, these parlances derive from foreign importations which survive as linguistic islands surrounded by other tongues. Their speakers tend to be bilinguals, using their own idiom for in-group communication and the majority language for interaction with outsiders.

Linguistic distinctness may also result from seemingly intentional processes of distortion. One very common form of secret language, found in a variety of tribal and complex societies, achieves unintelligibility by a process of verbal play with majority speech, in which phonetic or grammatical elements are systematically reordered. The pig Latin of English-speaking schoolchildren, in which initial consonants are transferred to the end of the word and followed by “-ay,” is a relatively simple example of this process. Thieves’ argots, the slang of youth gangs, and the jargon of traveling performers and other occupational groups obtain similar results by assigning special meanings to common nouns, verbs, and adjectives.

Despite their similarities, the classical administrative and liturgical languages—such as the Latin of medieval Europe, the Sanskrit of south Asia, and the Arabic of the Near East—are not ordinarily grouped with special parlances because of the prestige of the cultural traditions associated with them.

They are quite distinct from and often unrelated to popular speech, and the elaborate ritual and etiquette that surround their use can be learned only through many years of special training. Instruction is available only through private tutors and is limited to a privileged few who command the necessary social status or financial resources. As a result, knowledge of these languages in the traditional societies where they are used is limited to relatively small elites, who tend to maintain control of their linguistic skills in somewhat the same way that craft guilds strive for exclusive control of their craft skills.

The standard literary languages of modern nation-states, on the other hand, tend to be representative of majority speech. As a rule they originated in rising urban centers, as a result of the free interaction of speakers of a variety of local dialects, became identified with new urban elites, and in time replaced older administrative languages. Codification of spelling and grammar by means of dictionaries and dissemination of this information through public school systems are characteristic of standard-language societies. Use of mass media and the prestige of their speakers tend to carry idioms far from their sources; such idioms eventually replace many pre-existing local dialects and special parlances.

Linguistic acculturation, language shift

Wherever two or more speech communities maintain prolonged contact within a broad field of communication, there are crosscurrents of diffusion. The result is the formation of a Sprachbund, comprising a group of varieties which coexist in social space as dialects, distinct neighboring languages, or special parlances. Persistent borrowing over long periods creates within such groups similarities in linguistic structure, which tend to obscure pre-existing genetic distinctions; a commonly cited example is the south Asian subcontinent, where speakers of Indo-Aryan, Dravidian, and Munda languages all show significant overlap in their linguistic habits.

It appears that single nouns, verbs, and adjectives are most readily diffused, often in response to a variety of technological innovations and cultural or religious trends. Pronunciation and word order are also frequently affected. The level of phonological and grammatical pattern (i.e., the structural core of a language), however, is more resistant to change, and loanwords tend to be adapted to the patterns of the recipient language. But linguistic barriers to diffusion are never absolute, and in situations of extensive bilingualism— two or more languages being regularly used in the course of the daily routine—even the grammatical cores may be affected.

Cross-cultural influence reaches a maximum in the cases of pidgins and Creoles, idioms combining elements of several distinct languages. These hybrids typically arise in colonial societies or in large trading centers where laborers torn out of their native language environments are forced to work in close cooperation with speakers of different tongues. Cross-cultural influence may also give rise to language shift, the abandonment of one native tongue in favor of another. This phenomenon most frequently occurs when two groups merge, as in tribal absorption, or when minority groups take on the culture of the surrounding majority.

Although the bulk of the research on speech communities that was conducted prior to 1940 is historically oriented, students of speech communities differ markedly from their colleagues who concentrate upon textual analysis. The latter tend to treat languages as independent wholes that branch off from uniform protolanguages in accordance with regular sound laws. The former, on the other hand, regard themselves primarily as students of behavior, interested in linguistic phenomena for their broader sociohistorical significance. By relating dialect boundaries to settlement history, to political and administrative boundaries, and to culture areas and by charting the itineraries of loanwords in relation to technical innovations or cultural movements, they established the primacy of social factors in language change, disproving earlier theories of environmental or biological determinism.

The study of language usage in social communities, furthermore, revealed little of the uniformity ordinarily ascribed to protolanguages and their descendants; many exceptions to the regularity of sound laws were found wherever speakers of genetically related languages were in regular contact. This led students of speech communities to challenge the “family-tree theory,” associated with the neogrammarians of nineteenth-century Europe, who were concerned primarily with the genetic reconstruction of language history. Instead, they favored a theory of diffusion which postulates the spread of linguistic change in intersecting “waves” that emanate from different centers of innovation with an intensity proportionate to the prestige of their human carriers.

Thus, while geneticists regarded modern language distribution as the result of the segmentation of older entities into newer and smaller subgroups, diffusionists viewed the speech community as a dynamic field of action where phonetic change, borrowing, language mixture, and language shift all occur because of social forces, and where genetic origin is secondary to these forces. In recent years linguists have begun to see the two theories as complementary. The assumption of uniformity among protolanguages is regarded as an abstraction necessary to explain existing regularities of sound change and is considered extremely useful for the elucidation of long-term prehistoric relationships, especially since conflicting short-term diffusion currents tend to cancel each other. Speech-community studies, on the other hand, appear better adapted to the explanation of relatively recent changes.

Language behavior and social communication

The shift of emphasis from historical to synchronic problems during the last three decades has brought about some fundamental changes in our theories of language, resulting in the creation of a body of entirely new analytical techniques. Viewed in the light of these fresh insights, the earlier speech-community studies are subject to serious criticism on grounds of both linguistic and sociological methodology. For some time, therefore, linguists oriented toward formal analysis showed very little interest. More recent structural studies, however, show that this criticism does not affect the basic concept of the speech community as a field of action where the distribution of linguistic variants is a reflection of social facts. The relationship between such variants when they are classified in terms of usage rather than of their purely linguistic characteristics can be examined along two dimensions: the dialectal and the superposed.

Dialectal relationships are those in which differences set off the vernaculars of local groups (for example, the language of home and family) from those of other groups within the same, broader culture. Since this classification refers to usage rather than to inherent linguistic traits, relationships between minority languages and majority speech (e.g., between Welsh and English in Britain or French and English in Canada) and between distinct languages found in zones of intensive intertribal contact (e.g., in modern Africa) can also be considered dialectal, because they show characteristics similar to the relationship existing between dialects of the same language.

Whereas dialect variation relates to distinctions in geographical origin and social background, superposed variation refers to distinctions between different types of activities carried on within the same group. The special parlances described above form a linguistic extreme, but similar distinctions in usage are found in all speech communities. The language of formal speechmaking, religious ritual, or technical discussion, for example, is never the same as that employed in informal talk among friends, because each is a style fulfilling particular communicative needs. To some extent the linguistic markers of such activities are directly related to their different technical requirements. Scientific discussion, for instance, requires precisely defined terms and strict limitation on their usage. But in other cases, as in greetings, forms of address, or choosing between “isn”t’ and “ain’t,” the primary determinant is the social relationship between speakers rather than communicative necessity. Language choice in these cases is limited by social barriers; the existence of such barriers lends significance to the sociolinguistic study of superposed variation.

This distinction between dialectal and superposed varieties obviates the usual linguistic distinction between geographically and socially distributed varieties, since the evidence indicates that actual residence patterns are less important as determinants of distribution than social interaction patterns and usage. Thus, there seems to be little need to draw conceptual distinctions upon this basis.

Descriptions of dialectal and superposed variation relate primarily to social groups. Not all individuals within a speech community have equal control of the entire set of superposed variants current there. Control of communicative resources varies sharply with the individual’s position within the social system. The more narrowly confined his sphere of activities, the more homogeneous the social environment within which he interacts, and the less his need for verbal facility. Thus, house-wives, farmers, and laborers, who rarely meet outsiders, often make do with only a narrow range of speech styles, while actors, public speakers, and businessmen command the greatest range of styles. The fact that such individual distinctions are found in multilingual as well as in linguistically homogeneous societies suggests that the common assertion which identifies bilingualism with poor scores in intelligence testing is in urgent need of reexamination, based, as it is, primarily on work with underprivileged groups. Recent work, in fact, indicates that the failure of some self-contained groups to inculcate facility in verbal manipulation is a major factor in failures in their children’s performances in public school systems.

Attitudes to language choice . Social norms of language choice vary from situation to situation and from community to community. Regularities in attitudes to particular speech varieties, however, recur in a number of societies and deserve special comment here. Thieves’ argots, gang jargons, and the like serve typically as group boundary maintaining mechanisms, whose linguistic characteristics are the result of informal group consensus and are subject to continual change in response to changing attitudes. Individuals are accepted as members of the group to the extent that their usage conforms to the practices of the day. Similar attitudes of exclusiveness prevail in the case of many tribal languages spoken in areas of culture contact where other superposed idioms serve as media of public communication. The tribal language here is somewhat akin to a secret ritual, in that it is private knowledge to be kept from outsiders, an attitude which often makes it difficult for casual investigators to collect reliable information about language distribution in such areas.

Because of the elaborate linguistic etiquette and stylistic conventions that surround them, classical, liturgical, and administrative languages function somewhat like secret languages. Mastery of the conventions may be more important in gaining social success than substantive knowledge of the information dispensed through these languages. But unlike the varieties mentioned above, norms of appropriateness are explicit in classical languages; this permits them to remain unchanged over many generations.

In contrast, the attitude to pidgins, trade languages, and similar intergroup media of communication tends to be one of toleration. Here little attention is paid to linguistic markers of social appropriateness. It is the function of such languages to facilitate contact between groups without constituting their respective social cohesiveness; and, as a result, communication in these languages tends to be severely restricted to specific topics or types of interaction. They do not, as a rule, serve as vehicles for personal friendships.

We speak of language loyalty when a literary variety acquires prestige as a symbol of a particular nationality group or social movement. Language loyalty tends to unite diverse local groups and social classes, whose members may continue to speak their own vernaculars within the family circle. The literary idiom serves for reading and for public interaction and embodies the cultural tradition of a nation or a sector thereof. Individuals choose to employ it as a symbol of their allegiance to a broader set of political ideals than that embodied in the family or kin group.

Language loyalty may become a political issue in a modernizing society when hitherto socially isolated minority groups become mobilized. Their demands for closer participation in political affairs are often accompanied by demands for language reform or for the rewriting of the older, official code in their own literary idiom. Such demands often represent political and socioeconomic threats to the established elite, which may control the distribution of administrative positions through examination systems based upon the official code. The replacement of an older official code by another literary idiom in modernizing societies may thus represent the displacement of an established elite by a rising group.

The situation becomes still more complex when socioeconomic competition between several minority groups gives rise to several competing new literary standards, as in many parts of Asia and Africa, where language conflicts have led to civil disturbances and political instability. Although demands for language reform are usually verbalized in terms of communicative needs, it is interesting to observe that such demands do not necessarily reflect important linguistic differences between the idioms in question. Hindi and Urdu, the competing literary standards of north India, or Serbian and Croatian, in Yugoslavia, are grammatically almost identical. They differ in their writing systems, in their lexicons, and in minor aspects of syntax. Nevertheless, their proponents treat them as separate languages. The conflict in language loyalty may even affect mutual intelligibility, as when speakers’ claims that they do not understand each other reflect primarily social attitudes rather than linguistic fact. In other cases serious linguistic differences may be disregarded when minority speakers pay language loyalty to a standard markedly different from their own vernacular. In many parts of Alsace-Lorraine, for example, speakers of German dialects seem to disregard linguistic fact and pay language loyalty to French rather than to German.

Varietal distribution

Superposed and dialectal varieties rarely coincide in their geographical extent. We find the greatest amount of linguistic diversity at the level of local, tribal, peasant, or lower-class urban populations. Tribal areas typically constitute a patchwork of distinct languages, while local speech distribution in many modern nations takes the form of a dialect chain in which the speech of each locality is similar to that of adjoining settlements and in which speech differences increase in proportion to geographical distance. Variety at the local level is bridged by the considerably broader sprea of superposed varieties, serving as media of supralocal communication. The Latin of medieval Europe and the Arabic of the Near East form extreme examples of supra-local spread. Uniformity at the superposed level in their case, however, is achieved at the expense of large gaps in internal communication channels. Standard languages tend to be somewhat more restricted in geographical spread than classical languages, because of their relationship to local dialects. In contrast to a society in which classical languages are used as superposed varieties, However, a standard-language society possesses better developed channels of internal communication, partly because of its greater linguistic homogeneity and partly because of the internal language loyalty that it evokes.

In fact, wherever standard languages are well-established they act as the ultimate referent that determines the association of a given local dialect with one language or another. This may result in the anomalous situation in which two linguistically similar dialects spoken on different sides of a political boundary are regarded as belonging to different languages, not because of any inherent linguistic differences but because their speakers pay language loyalty to different standards. Language boundaries in such cases are defined partly by social and partly by linguistic criteria.

Verbal repertoires

The totality of dialectal and superposed variants regularly employed within a community make up the verbal repertoire of that community. Whereas the bounds of a language, as this term is ordinarily understood, may or may not coincide with that of a social group, verbal repertoires are always specific to particular populations. As an analytical concept the verbal repertoire allows us to establish direct relationships between its constituents arid the socioeconomic complexity of the community.

We measure this relationship in terms of two concepts: linguistic range and degree of compartmentalization. Linguistic range refers to internal language distance between constituent varieties, that is, the total amount of purely linguistic differentiation that exists in a community, thus distinguishing among multilingual, multidialectal, and homogeneous communities. Compartmentalization refers to the sharpness with which vareties are set off from each other, either along the super-posed or the dialectal dimension. We speak of compartmentalized repertoires, therefore, when several languages are spoken without their mixing, when dialects are set off from each other by sharp isogloss bundles, or when special parlances are sharply distinct from other forms of speech. We speak of fluid repertoires, on the other hand, when transitions between adjoining vernaculars are gradual or when one speech style merges into another in such a way that it is difficult to draw clear borderlines.

Initially, the linguistic range of a repertoire is a function of the languages and special parlances employed before contact. But given a certain period of contact, linguistic range becomes dependent upon the amount of internal interaction. The greater the frequency of internal interaction, the greater the tendency for innovations arising in one part of the speech community to diffuse throughout it. Thus, where the flow of communication is dominated by a single all-important center—for example, as Paris dominates central France— linguistic range is relatively small. Political fragmentation, on the other hand, is associated with diversity of languages or of dialects, as in southern Germany, long dominated by many small, semi-independent principalities.

overall frequency in interaction is not, however, the only determinant of uniformity. In highly stratified societies speakers of minority languages or dialects typically live side by side, trading, exchanging services, and often maintaining regular social contact as employer and employee or master and servant. Yet despite this contact, they tend to preserve their own languages, suggesting the existence of social norms that set limits to freedom of intercommunication. Compartmentalization reflects such social norms. The exact nature of these sociolinguistic barriers is not yet clearly understood, although some recent literature suggests new avenues for investigation.

We find, for example, that separate languages maintain themselves most readily in closed tribal systems, in which kinship dominates all activities. Linguistically distinct special parlances, on the other hand, appear most fully developed in highly stratified societies, where the division of labor is maintained by rigidly defined barriers of ascribed status. When social change causes the breakdown of traditional social structures and the formation of new ties, as in urbanization and colonialization, linguistic barriers between varieties also break down. Rapidly changing societies typically show either gradual transition between speech styles or, if the community is bilingual, a range of intermediate varieties bridging the transitions between extremes.

[see alsoLanguage, article onLanguage and culture; andLinguistics, article onHistorical linguistics.]

BIBLIOGRAPHY

Barth, frederik 1964 Ethnic Processes on the PathanBaluch Boundary. Pages 13-20 in Indo-Iranica: Mélanges présentés à Georg Morgenstierne, à Voccasion de son soixante-dixieme anniversaire. Wiesbaden (Germany): Harrassowitz.

Bernstein, basil (1958) 1961 Social Class and Linguistic Development: A Theory of Social Learning. Pages 288–314 in A. H. Halsey et al. (editors), Education, Economy, and Society. New York: Free Press. → First published in Volume 9 of the British Journal of Sociology.

Bloomfield, leonard (1933) 1951 Language. Rev. ed New York: Holt.

Brown, roger w. 1965 Social Psychology. New York: Free Press.

Gumperz, john j.; and hymes, dell h. (editors) 1964 The Ethnography of Communication. American Anthropologist New Series 66, no. 6, part 2.

Halliday, michael a. k.; mc!ntosh, angus; and Strevens, peter (1964) 1965 The Linguistic Sciences and Language Teaching. Bloomington: Indiana Univ. Press.

Haugen, einar i. 1956 Bilingualism in the Americas:

A Bibliography and Research Guide. University, Ala.: American Dialect Society.

Haugen, Einar I. 1966 Language Conflict and Language Planning. Cambridge, Mass.: Harvard Univ. Press.

Hertzler,joyce O. 1965 A Sociology of Language. New York: Random House.

Hymes,dell H. (editor) 1964 Language in Culture and Society: A Reader in Linguistics and Anthropology. New York: Harper.

Jespersen,otto (1925) 1964 Mankind, Nation and the Individual, From a Linguistic Point of View. Bloomington: Indiana Univ. Press. → First published as Menneskehed, nasjon og individ i sproget. Kurath, hans (editor) 1939–1943 Linguistic Atlas of New England. 3 vols. and a handbook. Providence, R.I.: Brown Univ. Press.

Labov, William 1966 The Social Stratification of English in New York City. Unpublished manuscript, Center for Applied Linguistics.

Passin, Herbert 1963 Writer and Journalist in the Transitional Society. Pages 82-123 in Conference on Communication and Political Development, Dobbs Ferry, N.Y., 1961, Communications and Political Development. Edited by Lucian W. Pye. Princeton Univ. Press. → Contains a discussion of the relationship of national languages to political development.

Weinreich, uriel 1953 Languages in Contact: Findings and Problems. New York: Linguistic Circle of New York.

247. Linguistics

See also 186. GRAMMAR ; 236. LANGUAGE ; 330. PRONUNCIATION ; 382. SPEECH .

betacism

1. excessive use of the sound b.

2. improper articulation of this sound. —betacist , n.

biolinguistics

the study of the relations between physiology and speech. —biolinguist , n.

cherology

the description and analysis of the distinctive units used in the sign language of the deaf. —cherologist , n. —cherologic, cherological , adj.

diachronism, diachrony

the study and description of the change or development in the structural systems of a language over a stated period of time. Also called historical linguistics . Cf. synchronic linguistics . —diachronic , adj.

dialect

a variety of a language peculiar to a particular region or group within a larger community, usually but not always existing in the spoken form only. —dialectal , adj.

dialect geography

the study of dialects with regard to their geographic distribution, as well as how their distribution may be affected by geography, e.g., the spread of a particular dialect being halted at a mountain range, forest belt, body of water, etc.

dialectology

1. the study of dialects and dialect features.

2. the linguistic features of a dialect. —dialectician, dialectologist , n. —dialectologie, dialectological , adj.

echoism

1. the formation of sounds like those in nature; onomatopoesis.

2. the tendency of paired sounds to become more similar phonetically, as the d sound in iced tea which has become a t; assimilation. —echoic , adj.

etymology

the study of the origin and history of individual words. —etymologist , n. —etymological , adj.

folk etymology

the reanalysis of a word by native speakers into a new element or elements, e.g. hamburger (properly ‘from Hamburg’) being split into ham- and -burger; and the subsequent combination of -burger with a number of words in which it is used to mean ‘ground patty.’

gammacism, gammacismus

the inability to pronounce the soft palatal consonants such as g and k.

geolinguistics

the study or science of linguistics in relation to geography. —geolinguist , n. —geolinguistic , adj.

glossematics

the science or study of glossemes, the smallest unit of linguistic communication. —glossematic , adj.

glossologist

Archaic.

1. a linguist; a philologist.

2. one who compiles glossaries.

glossology

Archaic.

linguistics.

glottochronology

a statistical and lexical study of two languages deriving from a common source to determine the time of their divergence, as English and German. Cf. lexicostatistics . —glottochronologist , n. —glottochronological , adj.

glottology

the science of linguistics.

grammar

1. the study of the formal system of a language, especially the aspects of sound, forms, and syntax.

2. a work detailing such an analysis. —grammarian , n. —grammatic, grammatical , adj.

graphemics

the study of systems of writing and their relationship to the systems of the languages they represent. Also called graphonomy . —graphemic , adj.

hybridism

1. a word formed from elements drawn from different languages.

2. the practice of coining such words. —hybrid , n., adj.

idiolect

a person’s individual speech habits.

lallation

Phonetics.

1. the replacement of l for r in speech.

2. the mispronunciation of l. Cf. lambdacism .

lambdacism

Phonetics.

the mispronunciation of double l, giving it the sound of y or ly.

2. Cf. rhotacism . substitution of the sound l for another sound, as that of r. Also labdacism . Cf. lallation .

lexicography

the writing, editing, or compiling of dictionaries. —lexicographer , n. —lexicographic, lexicographical , adj.

lexicology

the study of the meanings of words and of idiomatic combinations. —lexicologist , n. —lexicologic, lexicological , adj.

lexicostatistics

the study of languages and their vocabularies by statistical methods for historical purposes. Cf. glottochronology . —lexicostatistic, lexicostatistical , adj.

lexigraphy

Rare. the art of defining words or compiling lexicons. —lexigraphic , adj.

linguistic typology

the classification of languages by structural similarity, e.g., similarity of syntactic or phonemic features, as opposed to classification on the basis of shared linguistic ancestry.

metalinguistics

the science or study of language in relation to its cultural context. —metalinguist , n. —metalinguistic, metalinguistical , adj.

morphemics

the study and description of the morphemes of a language, i.e., its minimum grammatical units, as wait and -ed in waited. —morphemicist , n.

morphology

1. a branch of linguistics that studies and describes patterns of word formation, including inflection, derivation, and compounding of a language.

2. such patterns of a particular language. —morphologist , n. —morphological , adj.

morphophonemics

1. the study of the relations between morphemes and their phonetic realizations, components, or distribution contexts.

2. the body of data concerning these relations in a specific language. —morphophonemicist , n. —morphophonemic , adj.

nasalism

a tendency toward nasality in pronouncing words. Also nasality .

onomasiology

onomastics. —onomasiologist , n. —onomasiologic, onomasiological , adj.

onomastics

the study of names and their origins. —onomastic , adj. —onomastician , n.

orthoepy

the study of correct pronunciation. —orthoepist , n. —orthoepic, orthoepical, orthoepistic , adj.

paronymy

the state or condition of containing the same root or stem, as perilous and parlous. —paronym , n.

philology

1. the study of written records to determine their authenticity, original form, and meaning.

2. linguistics, especially historical linguistics. —philologist, philologer , n. —philologic, philological , adj.

phonematics

phonemics.

phonemics

1. the study and description of phonemes, i.e., the set of basic units of sound used in a language and phonemic systems.

2. the phonemic system of a given language. Also phonematics . —phonemicist , n.

phonetics

1. the science or study of speech sounds and their production, transmission, and perception, and their analysis, classification, and transcription.

2. the science or study of speech sounds with respect to their role in distinguishing meanings among words.

3. the phonetic system of a particular language. Cf. phonology . —phonetician , n. —phonetic, phonetical , adj.

phonology

1. the study of the history and theory of sound changes in a language or in two or more languages comparatively.

2. the phonetics and phonemics of a language at a stated time; synchronic phonology. —phonologist , n. —phonological , adj.

psycholinguistics

the study of the relationships between language and the behavioral mechanisms of its users, especially in language learning by children. —psycholinguist , n. —psycholinguistic , adj.

rhotacism

Phonetics.

1. a misarticulation of the sound r or the substitution of another sound for it.

2. Cf. lambdacism . substitution of the sound sound r for another sound, as that of l.

2. the excessive use of the sound r.

3. Phonology. replacement of the sound z or s by r in Indo-European languages, as German wesen, English were. —rhotacize , v. —rhotacistic , adj.

semantics

1. the study of the meaning of words.

2. the study of linguistic development by examining and classifying changes in meaning. Also called semasiology, sematology, semology . —semanticist , n. —semantic , adj.

semasiology, sematology

semantics.

semeiology, semiology, semology

the study or science of signs; semantics. —semeiologist, semiologist , n. —semeiologic, semiologic, semeiological, semiological , adj.

semiotics, semiotic

the study of the relationship between symbology and language. —semiotician, semioticist , n.

sigmatism

a faulty pronunciation of sibilant sounds.

structuralism

an emphasis in research and description upon the systematic relations of formal distinctions in a given language. Also called structural linguistics . —structuralist , n.

synchronic linguistics

the study of the phonological, morphological, and syntactic features of a language at a stated time. Also called descriptive linguistics . Cf. diachronism .

syntax

the study of the principles by which words are used in phrases and sentences to construct meaningful combinations. —syntactic, syntactical , adj.

tagmemics

the study of the tagmemes of a language, i.e., the minimal units of grammatical construction, embodying such phenomena as distinctive word order and grammatical agreement. —tagmemic , adj.

tonetics

the phonetic study and science of the tonal aspects of language. —tonetician , n. —tonetic , adj.

transformationalist

an advocate or student of the theory of transformational grammar, a system of grammatical analysis that uses transformations of base sentences to explain the relations between thought and its syntactic manifestation and to express the relations between elements in a sentence, clause, or phrase, or between different forms of a word or phrase, as active or passive forms of a verb.

vocalism

Phonetics. the system of vowels in a given language. —vocalic , adj.

LINGUISTICS The systematic study of LANGUAGE. Its aim is to look at language objectively, as a human phenomenon, in order to uncover general principles and provide reliable descriptions.

History

Although the formal study of language dates from at least the middle of the first millennium BC in India and ancient Greece, the era of scientific language study is commonly dated from the end of the 18c, when English was discovered to have the same ancestor as a number of European and Asian languages. This discovery initiated at least a century of intense interest in COMPARATIVE PHILOLOGY, which involved uncovering links between languages, writing comparative grammars of related languages, and reconstructing their common ‘ancestors’. These activities stimulated a search for the mechanisms underlying LANGUAGE CHANGE. In the 20c, a change of emphasis occurred, largely through the work of the Swiss linguist Ferdinand de Saussure, sometimes regarded as ‘the father of modern linguistics’. He advocated separating DIACHRONIC (historical) from synchronic (contemporary or co-occurring) aspects of language study. He argued that language at any point in time is an interlocking structure, in which all items are interdependent, an insight which is now taken for granted in linguistics and forms the basis of 20c STRUCTURALISM. In the 1930s and 1940s, descriptive linguistics was developed largely in the US, as linguists sought to describe the fast-disappearing American-Indian languages, with Edward Sapir and Leonard BLOOMFIELD being regarded jointly as the ‘fathers of American linguistics’. Midway through the 20c, Noam CHOMSKY triggered another change of direction, when he instigated work in generative linguistics, a concern for the principles in the minds of speakers which could generate language (account for their knowledge of language in an explicit way).

Branches

Linguistics comprises a large number of branches, several of them hybrids with other disciplines. Although PHONETICS, the scientific study of speech sounds, is usually regarded as an intrinsic part of linguistics, it is often taken to be a discipline in its own right, especially by phoneticians, who point to its 19c origins. Linguistics and phonetics together are therefore often referred to as the linguistic sciences. At its core, linguistics can be said to have three classic subdivisions: (1) PHONOLOGY the study of sound patterns; (2) MORPHOLOGY and SYNTAX the composition of words and sentences; (3) SEMANTICS the study of meaning. Some linguists consider that morphology and syntax can be subsumed under the traditional term GRAMMAR; others argue that phonology, morphology, syntax, and semantics all constitute the grammar of a language. Each can be studied synchronically or diachronically (or both together) and the order in which they have been dealt with within a grammar has fluctuated over the years. In the last quartercentury, some previously fringe areas have become increasingly important, notably: PRAGMATICS the study of language usage; SOCIOLINGUISTICS the study of the relationship between language and society; PSYCHOLINGUISTICS language and the mind; LINGUISTIC TYPOLOGY the analysis of languages into types; computational linguistics the use of computers to simulate language processes; stylistics linguistic analysis applied to literature and style; APPLIED LINGUISTICS linguistics in relation to such practical activities as language teaching, LEXICOGRAPHY, and speech therapy. See CORPUS.

LINGUISTICS

LINGUISTICS. The early discipline of linguistics in the United States consisted in large part of the work of three eminent scholars—Franz Boas, who studied Native American languages; Edward Sapir, the most prolific of Boas's students; and Leonard Bloomfield, who was trained in Germanic philology and taught languages. Boas, Sapir, and Bloomfield were among the founders in 1924 of the Linguistic Society of America, the leading professional organization and publisher of the discipline's journal.

Bloomfield and Sapir were leaders in descriptive linguistics, now often referred to as structural linguistics. According to them, languages should be described as interlocking assemblages of basic units and as functioning wholes independent of earlier developmental stages. Such descriptions might then form the basis for comparing related languages and reconstructing their common origin. Sapir identified the phoneme as a basic unit of sound patterning and offered evidence for its psychological reality. Bloomfield, on the other hand, advocated indirect observation to identify the distinct meanings associated with units of form. His followers developed a mandatory set of discovery procedures for all valid analyses that built upon the sequential distribution of units of sound. These procedures, and strictures against mixing comparison with description, were in practice often violated, with good reason. Linguists were prepared to assume that languages might differ from one another without limit; thus, one could assume no commonalities. They were reacting in part to clumsy attempts to superimpose categories of classical grammar on descriptions of New World languages. Many of them thought that the grammatical categories of language might shape perceptions of reality.

Beginning in the mid-twentieth century, Noam Chomsky revised these ideas—including the supposed necessity of phonetically based discovery—in what became known as generative grammar. Language was for him a hypothetico-deductive system centered about Syntactic Structures, the title of his 1957 treatise. According to Chomsky, language and human cognition evolve together. Language is innate, its categories universal among humankind. It is part of children's normal development, rather than a skill learned by some and not by others, such as playing a musical instrument or driving a car. Children must ascertain the particular sound-meaning combinations and parameter settings used in their environment. The linguist can learn more about this innate capability from probing a single language rather than surveying multiple languages.

Whereas generative grammar was autonomous, with many of its constructs presuming homogeneous speech communities of identical idealized hearer-speakers, William Labov developed methods for sampling and quantifying the variation in usage by members of actual communities and by given individuals on different occasions. He showed the influence of social attitudes on language within speech communities. Some of his studies using the sound spectrograph demonstrated that speakers perpetuate distinctions they are unable to recognize.

Even as they considered the existence of a universal grammar, however, linguists in the 1990s became concerned with the high rate of language death in the modern world. An increasing number of young linguists committed themselves to studying language ecology, in hopes of preventing or curtailing the incidence of language death, and to recording and analyzing little-studied endangered languages to preserve at least a record of what had been lost. It was almost as if the discipline had come full circle from the efforts of Boas and his students nearly a century earlier.

BIBLIOGRAPHY

Hymes, Dell, and John Fought. American Structuralism. The Hague and New York: Mouton, 1981.

Newmeyer, Frederick J. Linguistic Theory in America. 2d ed. Orlando, Fla.: Academic Press, 1986.

Newmeyer, Frederick J., ed. Linguistics: The Cambridge Survey. 4 vols. New York: Cambridge University Press, 1988.

Byron W.Bender/c. w.

See alsoBehaviorism ; Post-structuralism ; Psychology .

linguistics Systematic study of language, its nature, structure, constituent elements, and changes. As a discipline, linguistics embraces phonetics, phonology (the study of sound systems within languages), grammar (including syntax), semantics and pragmatics (the study of language use). See also Chomsky; Saussure; structuralism