Directions in Sociolinguistics: Report on an Interdisciplinary Seminar

A 1964 summer seminar hosted by the SSRC’s Committee on Sociolinguistics highlighted tensions between sociology and linguistics when scholars gathered to address how their disciplines can deepen research on language’s impact on society. For example, sociologists questioned linguistics’ lack of definition for language or dialect while linguists raised concerns about sociology’s reliance on large quantified data. However, by the end of the seminar, the scholars agreed the encounter had raised important questions and opened new paths of investigation through both sociological and linguistic approaches, including the study of language and social stratification, multilingualism, and language standardization. Every interdisciplinary encounter by its nature involves tensions related to the differing values and expectations of the groups represented. The research seminar on sociolinguistics held in the summer of 1964 at Indiana University showed many evidences of such tensions. The participants in the seminar agreed, however, by the end of the summer that in their own thinking at least the fields of sociology and linguistics had profited from the encounter and new areas of mutual concern had been uncovered. Charles A. Ferguson, University of Washington, USA 1 See Items, December 1963, p. 52, and June 1964, p. 22. In addition to the participants named there, the regular participants in the seminar included Ranier Lang and Leonard Lieberman, PhD candidates in anthropology at the University of California, Los Angeles, and Michigan State University, respectively. John Useem, Michigan State University, participated during half of the session. Three consultants spent about a week each with the seminar: Susan M. Ervin, University of California, Berkeley; Joseph H. Greenberg, Stanford University; Wallace E. Lambert, McGill University. Special presentations to the seminar were made by Allen D. Grimshaw, Indiana University; Michael A. K. Halliday, University of London; Dell Hymes, University of California, Berkeley; Stephan P. Wurm, Australian National University. C. F. Voegelin, Indiana University, was a regular visitor to the seminar. IJSL 2020; 263: 37–44 Open Access. © 2020 Ferguson, published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 Public License. At the beginning of the summer the linguists in the seminar had some definite psychological advantages: They were surrounded by several hundred fellow linguists teaching or studying at the summer Linguistic Institute to which the seminar was attached. Also, all of them had done a considerable amount of research on sociolinguistic phenomena and had acquired some fairly definite notions about the theoretical bases of their work. Finally, they all had the conviction—generally well-hidden but sometimes coming to the surface—that only linguists really understand how language works and consequently sociologists would have to master many of the concepts and techniques of linguistic science in order to do any fruitful work at all on sociolinguistic questions. The sociologists were in a different state: They recognized the need for a deeper understanding of language behavior, but they were doubtful about the social science sophistication of the linguists. On the other hand, the sociologists were perhaps too easily impressed by the linguists’ apparent familiarity with the facts of dozens of languages. During the summer, however, they came to appreciate the insights and knowledge of the linguists and even to acknowledge the importance of the study of linguistics by sociologists; at the same time they realized that as sociologists they had important contributions to make in theory and method as well as in the choice of problems to be investigated. Fairly early in the seminar it became clear that the sociologists could ask embarrassing questions, which pointed up weaknesses in the current theory and practice of linguistics. They asked, for example, for some kind of measure, no matter how crude, of the degree of structural difference between any two languages, since they would find this useful for many kinds of analysis. They were quite baffled to discover that linguists had no such measure and that linguists on the whole had not been interested in devising one. Similarly the sociologists asked for guidance on defining the limits between language and dialect or between one language and another, only to find that the linguists regarded this question as a troublesome problem for which they had no ready solution. One of the greatest obstacles to effective communication between the two groups was the simple fact that the term “data” meant very different things to sociologists and linguists. It is not too strong to say that some of the sociologists were shocked by the linguists’ methods of data collection. They found it hard to believe that most linguists seem to rely on casually chosen informants and poorly controlled introspection for extensive generalizations about the language behavior of large, diversified populations. The linguists, in defending their favorite kinds of data (heard or recorded utterances and documentary attestation of forms), came to see the need for increased sophistication in sampling techniques and generally in statistical treatment of data. On the other hand, the linguists tended to be suspicious of some of the masses of quantified data which the sociologists were 38 Charles A. Ferguson

At the beginning of the summer the linguists in the seminar had some definite psychological advantages: They were surrounded by several hundred fellow linguists teaching or studying at the summer Linguistic Institute to which the seminar was attached. Also, all of them had done a considerable amount of research on sociolinguistic phenomena and had acquired some fairly definite notions about the theoretical bases of their work. Finally, they all had the conviction-generally well-hidden but sometimes coming to the surface-that only linguists really understand how language works and consequently sociologists would have to master many of the concepts and techniques of linguistic science in order to do any fruitful work at all on sociolinguistic questions.
The sociologists were in a different state: They recognized the need for a deeper understanding of language behavior, but they were doubtful about the social science sophistication of the linguists. On the other hand, the sociologists were perhaps too easily impressed by the linguists' apparent familiarity with the facts of dozens of languages. During the summer, however, they came to appreciate the insights and knowledge of the linguists and even to acknowledge the importance of the study of linguistics by sociologists; at the same time they realized that as sociologists they had important contributions to make in theory and method as well as in the choice of problems to be investigated.
Fairly early in the seminar it became clear that the sociologists could ask embarrassing questions, which pointed up weaknesses in the current theory and practice of linguistics. They asked, for example, for some kind of measure, no matter how crude, of the degree of structural difference between any two languages, since they would find this useful for many kinds of analysis. They were quite baffled to discover that linguists had no such measure and that linguists on the whole had not been interested in devising one. Similarly the sociologists asked for guidance on defining the limits between language and dialect or between one language and another, only to find that the linguists regarded this question as a troublesome problem for which they had no ready solution.
One of the greatest obstacles to effective communication between the two groups was the simple fact that the term "data" meant very different things to sociologists and linguists. It is not too strong to say that some of the sociologists were shocked by the linguists' methods of data collection. They found it hard to believe that most linguists seem to rely on casually chosen informants and poorly controlled introspection for extensive generalizations about the language behavior of large, diversified populations. The linguists, in defending their favorite kinds of data (heard or recorded utterances and documentary attestation of forms), came to see the need for increased sophistication in sampling techniques and generally in statistical treatment of data. On the other hand, the linguists tended to be suspicious of some of the masses of quantified data which the sociologists were accustomed to handle, and were able to demonstrate pitfalls in the use of statistical data on language by showing that even such simple census questions as "What language do you speak?" involve complex attitudinal factors as well as problems of language identification and the evaluation of proficiency. 2 Another source of tension in the discussion was the difference in methods of analysis and argumentation customary in the two groups. The sociologists found quite foreign to their experience what one of them called the linguists' "anecdotal approach" in argumentation, by which a linguist offering a generalization would agree that it was demolished if another linguist present could think of a counterexample from some exotic language which only he knew about. Most of the linguists, on the other hand, were quite unfamiliar with the intricate analysis of covariance or the setting up of alternative conceptual frameworks which the sociologists found congenial.
The first several weeks of the seminar were devoted to the exchange of information, attitudes, and justifications of interest, at first warily and defensively but with increasing mutual respect and acceptance. The procedure was chiefly the presentation of the individual participants' research findings either in oral reports or by circulation of reprints and manuscripts, and then fairly unstructured discussion of some of the issues raised. Also, the first two outside consultants, both experimental psychologists interested in sociolinguistic phenomena, served to focus the attention of the whole group on useful approaches to the field which were neither linguistic nor sociological in emphasis.
At the end of the third week three working groups were set up to meet outside the daily seminar sessions and report back to the whole seminar, respectively on language and social theory, multilingualism, and language standardization (these reflect some of the major common interests of the participants). Each working group had a sociologist as chairman, and each group produced draft documents for consideration by the seminar. A number of the papers started in these groups during the summer are now being revised and expanded by their authors. 3

Language and social stratification
One point on which there was immediate communication between sociologists and linguists was the language aspect of social stratification. The long-standing interest of American sociologists in social stratification has included many dimensions of the subject, but has never been focused in a serious way on its language aspect. Sociologists and other social scientists have been content to observe occasionally that linguistic phenomena seem to be correlated with other social data, coinciding with lines of communication in a community or in effect marking group membership, but they have not investigated these phenomena in a systematic way. On the other hand, scholars in linguistic science have long maintained the importance of social dialects and the social matrix of linguistic differentiation, but the work of the linguists in this field has been characterized by lack of sophistication in techniques of social analysis other than those commonly used in anthropological field work.
William Labov reported to the seminar in some detail on a series of studies, undertaken in New York City during the past several years, which combined the expert knowledge and insights of linguists and sociologists. Working with a population that had previously been surveyed in detail by a team of social scientists, Labov, using appropriate sampling procedures and ingenious modifications of linguistic field methods, showed a strikingly high agreement between linguistic performance on five small phonological points and the socioeconomic ratings, based on occupation, education, and income, made by the social scientists. In fact, for this population the agreement was so high as to suggest that a linguistically trained observer in possession of Labov's results would be able to locate an individual on the socioeconomic scale much faster with linguistic indices than would an interviewer using the nonlinguistic information.
The significance of this kind of research for social scientists was quite clear. A study of social stratification in America-and presumably in other societies as well-that ignores language phenomena is missing one of the most sensitive and accessible indicators of community ranking. Since analysis of the kind of phonological variation that is so important in American social dialectology often requires considerable technical competence in linguistics, a linguist may often be a useful member of a team of social scientists. In many cases, however, once the linguist has found a useful diagnostic feature, it may often be possible for an interviewer or field worker with little specialized linguistic training to employ it. For example, in the analysis of the speech of sales personnel in department stores catering to different socioeconomic strata, it was possible to achieve impressive results simply by eliciting under natural working conditions the pronunciation of the word "fourth" as in "fourth floor" and recording whether the "r" was pronounced.
Anthropologists have been interested in the relation between languages and the cultures in which they are or of which they form a part, and courses on "Language and Culture" are regularly offered at universities. Strangely enough, however, ethnographers have shown little interest in the variation in language use from one society to another or in the sets of beliefs or attitudes toward it which are current in the different societies. Dell Hymes has called this new field of research the "ethnography of communication" and has stimulated a number of linguists and anthropologists to work in the field. As a result, sociologists concerned with the analysis of census data about language or linguists concerned with linguistic analysis in situations of language contact now find their work related to this larger field.

Multilingualism
On the major topic of multilingualism, the seminar discussed on a number of occasions the new approach of John Gumperz, based on studies he had carried out in Norway and India. This approach takes as its starting point a sociolinguistic community as defined by the amount of communicative interaction and delineates the total verbal repertoire of the community, which may include several distinct dialects or styles of the same language or of two or more distinct languages. On the basis of description of the constituent "varieties" of the repertoire and the situational and other factors which bring the respective varieties into play, Gumperz hopes to clarify our understanding of the dynamics of communication and sociolinguistic behavior in the community and to develop a general theory of sociolinguistics.
An example of this approach dealt with a large community of bilingual Hindi-Panjabi speakers in Delhi. The members of this community use both languages with full competence and in a wide variety of situations within the community and outside it. Gumperz' careful study of this community in terms of the forms of each language and the occasions of use casts light not only on technical questions of linguistic interference but also on the present values and social structure of the community and its historical development. The participants in the seminar generally agreed on the value of this approach in study of relatively small communities and explored its applicability to larger societies and nations.
A large part of the seminar was spent in broadening the participants' knowledge of the range of possible kinds of multilingual communities. This came partly from the theoretical interest of some and partly from the very varied personal experiences of the participants, which included work in the Philippines, India, Arab countries, Mexico, the Caribbean, Canada, Switzerland, Scandinavia, and Africa. The discussions of the second working group and a number of the seminar sessions as well were devoted to the general problem of language maintenance and language shift in a society and the classification of multilingual societies in terms of the "types" and "functions" of languages in them.
Joshua Fishman, in studying the language retention of ethnic groups in the United States, had focused his attention on the notion of the "domains" of use of a language in a society and the classifications and basic principles related to this. Heinz Kloss, in tracing the emergence of a number of new Germanic literary languages in the nineteenth and twentieth centuries, provided concepts such as Abstand languages (independent by virtue of linguistic distance from others) and Ausbau languages (independent by virtue of the level of use in writing), which hold promise of general validity. William Stewart outlined a sociolinguistic typology for describing national multilingualism which incorporates work of Kloss, Ferguson, and others and offers fresh insights for political scientists and sociologists.
In discussion of multilingualism the seminar participants were wrestling with the problem of developing a frame of reference or set of basic concepts in terms of which they could deal with such questions as these: How does it happen that one language is used as a lingua franca, another not? Why does one immigrant group or submerged indigenous minority retain its language, another not? How many official languages can be supported in a nation of a given size? What are the nonlinguistic correlates of intergroup bilingualism where each subgroup speaks its own language and communication is through a few bilinguals versus intragroup bilingualism where the whole community uses two languages in communication? Does the use of a foreign language in a nation's official life interfere with the economic, political, and social development of the country?

Language standardization
With the proliferation of nationalism after World War I and again after World War II with its generation of new nations, scholars of various disciplines have been led to examine the factors involved in the development of a standard language (i. e., the form of a language serving as a special kind of norm beyond regional and social dialects, often associated with a feeling of nationhood) and the role of language in the development of nationalism. Such men as the French linguists Antoine Meillet and J. Vendryes in the earlier period and more recently the political scientist and sociologist Karl Deutsch made important contributions in this area, but they have not been followed by any appreciable number of workers. Einar Haugen's study of the language controversy in Norway provided the seminar with a detailed case study in language standardization, where the language question has been a highly charged political issue and where the situation for many years seemed to move toward greater disruption and confusion. In contrast with this case where much of the change was brought about consciously and through written media, Ferguson summarized the early standardization of Classical Arabic which was apparently accomplished largely unconsciously, for the limited purposes of poetry and oratory, and with little use of writing. The working group on standardization was able to agree on a number of useful concepts ("homogeneity," "adequacy") and on certain general principles of description, but the chief conclusion was a call for research in areas almost untouched. How do language academies work? How does homogenization take place in a nonliterate society? What are the causal factors in the great variation of routes to standardization?
This listing of highlights connected with the three main areas of discussion at the seminar leaves out some of the most exciting moments of the summer, such as Savitz's account of the speech of juvenile gangs and Friedrich's elaborate analysis of nineteenth-century Russian patterns of address, as well as the presentations by consultants and other special visitors, but it may give a sense of the range of subject matter and the variety of approaches employed.
The success of the sociolinguistic seminar was another evidence of the strength of a growing trend toward the analysis of language as it functions in society. Earlier evidences of the trend include the series of symposia held at annual meetings of the American Anthropological Association, the sociolinguistic articles appearing frequently in the journal Anthropological Linguistics, courses on sociolinguistics offered by linguists in summer programs, 4 the University of California, Los Angeles, Conference on Sociolinguistics in May 1964, and the sociolinguistic session held at the Eastern Sociological Society's annual meeting in April 1964. Even greater evidence of the strength of this trend is found in the publication of volumes of selected readings. Hymes's Language in Culture and Society (Harper & Row, 1964), Fishman's Readings in the Sociology of Language, and the volume of papers from the UCLA Conference, edited by William Bright, (both in press, to be published by Mouton & Co.) are all appearing within a 12-month period, in addition to two special issues of the American Anthropologist devoted to topics in this field.