Language Culture and Society: 1924-2014
Roger W Shuy, LSA January 2014
As LSA was founded in 1924, the interrelationships of language, culture and society were already evident. Saussure had declared language to be a social fact, foreshadowing future evolving perspectives about language use, change, and variability. There were also others who viewed language as an autonomous, self-contained structure, independent of any social context. In the past 90 years, contrasting views of language as social fact versus language independent of social fact have endured, although not totally in the same ways and still not always in harmony.
The inheritance: Standing on the shoulders of giants
Before Saussure, manifestations of the study of language in social context were very different from those of the modern era. But the early developments of the comparative method eventually led to the notion of Indo-European in the 19th Century (Grimm, Humbolt, Leibniz, Jones), at which point the study of language, culture and society not only became apparent but also produced those who disagreed. The Neogrammarians rejected language variation and espoused the slogan, “sound laws suffer no exceptions.” In response, dialectologists of that era came up with their own opposing slogan: “every word has its own history.”
A half-century before LSA was founded William Dwight Whitney had proposed a close relationship of language, culture and society and wrote:
Speech is not a personal possession but a social: it belongs, not to the individual, but to the member of society. No item of existing language is the work of an individual; for what we may severally choose to say is not language until it is accepted and employed by our fellows. The whole development of speech, though initiated by the acts of individuals, is wrought out by the community (1867, 404).
Whitney’s concept was echoed nine years after LSA began, when Bloomfield’s monumental Language echoed Whitney’s concept and devoted an entire chapter to speech communities and presaged later studies of social dialect, gender differences, and age grading by linguists such as Meillet, Martinet, Weinreich, and eventually lead up to creative refinements of Labov and others. By the end of WWII, while descriptive linguists were eliciting information directly from speakers and producing phonological and grammatical descriptions of languages as unitary, homogeneous wholes, it was becoming clear that some research traditions were evolving in the area of language, culture, and society:
(1) the dialectology tradition, in which native speakers were interviewed in a variety of communities with the purpose of discovering the regional diffusion of selected linguistic features;
(2) the language contact tradition, which analyzed and compared what happens to languages when they bump up against each other; and
(3) the anthropological linguistics tradition, which included ethnographic analyses of the relationship of cultural variability and language use.
These three traditions were later joined by
(4) the sociology tradition, even though to that point sociologists had dealt only marginally with language as a critical social factor in group interaction and stratification,and by
(5) the psychological tradition that had begun to deal with language and cognition. These five research traditions focused on face-to-face interaction using the linguistic analysis of tapes and transcripts. Some were differentiated as “macro-sociolinguistics,” the concern with larger social units, and others as “micro-sociolinguistics,” the concern with smaller language units.
1. The dialectology tradition
Despite its emphasis on only regional dialect variation, the early twentieth century dialectologists in Europe had kept the study of language variation alive before the broader and more sophisticated relationships of language and society emerged. Whitney’searlier ideas about the how language and society were intimately connected were taken seriously by linguistic geographers in Germany, France and other European countries. Questionnaires were developed and field workers were sent out (some on bicycles) to describe the speech of large numbers of people who represented their geographical areas. The goal the American Dialect Society (founded in 1899) was to produce a dialect dictionary of American speech, but up to the time of the LSA was created, dialectologists had made little or no progress in producing such a dictionary. Many dialectologists, following the descriptive linguistic tradition of the time, had already turned their attention away from creating a dialect dictionary to carrying out field research for a mammoth regional dialect atlas of the US and Canada. Although the unfinished dialect dictionary had produced some 26,000 dialect terms, it provided little or no information about the regional distribution in which these terms existed. To address this missing information, in 1931 Hans Kurath was appointed the director of The Linguistic Atlas of The United States and Canada, leading a very ambitious effort to associate many of these already collected terms by geographical region and employed an unsophisticated measure of their social status. It is noteworthy, however, that social indicators were collected without concern for the social variability related to age, gender, ethnicity, or race of the informants and with an unsophisticated assessments of the informants’ social status.
Among the original leaders closely associated with the Atlas project were Kurath, McDavid, Allen, Cassidy and Marckwardt, who carried out some of the fieldwork themselves and enrolled their graduate students to do the same. For example, as one of McDavid’s students, my assignment was to carry out five or six hour Atlas interviews in every county of Illinois except in highly-urbanized Cook County, which McDavid assigned to my fellow student, Lee Pederson. Following Atlas protocol, in each county I used the lengthy Atlas interview to tape-recorded only the oldest, mostly male, mostly rural inhabitants who had not lived anywhere else in their lives. This provided some useful historical data, for it gave an indication of how English was spoken in those areas during the past hundred years, but it showed little or nothing about current variability related to age, gender, ethnicity, and race. And the only indicator of social status was the number of years the informants attended school.
By the 1960s a great deal of the Atlas fieldwork had been done but relatively little was published except for the work on the dialects of New England and the Upper Midwest. The Atlas project suffered from the slow pace at which the data were gathered, so by the time the first publications came out, the Atlas methodology had already come under serious review. Today, regional variability is still studied, but with vastly different methods and linguistic analysis.
2. The language contact tradition
In the 1940s Einer Haugen was producing exciting evidence about what happens when languages come in contact with each other, especially about bilingualism and loanwords. Linguistic interest in multilingualism increased in the 1960s and 70s, spearheaded by the work of Fishman, who examined the role of language in distinguishing between nationalism (government and education) and nationism (language, culture, history, religion). He concluded that multilingualism tends to work against nationalism unless one of the languages is chosen as the national or official language. He also discovered that different languages were assigned different tasks in multilingual societies. This association of language form with language function was developed and elaborated by Ferguson, who named this phenomenon diglossia, a label that combined function, prestige, literary heritage, acquisition, standardization, and stability, with the conventional linguistic tools of grammar, lexicon, and phonology. Up to that time, bilingualism had been considered the territory of psychology, but diglossia went beyond this to what was beginning to be known as sociolinguistics.
3. The anthropological linguistics tradition
Boas, Sapir, Malinowski, and Firth provided a tradition of anthropological linguistics presaging the studies of language, culture and society by later scholars such as Bright, Hymes, and Gumperz. We owe to Gumperz the unit of analysis called “speech communities,” much of the early research on “code switching,” and eventually the methodology of “interactional sociolinguistics,” all of which stress that social relationships underpin the language forms of discourse. Reacting to Chomsky’s ideas about language competence, Hymes created the concept of “communicative competence,” specifying what speakers needed in order to operate as a full member of a language community. This included knowledge of all the appropriate ways to use the language, as identified in his now famous SPEAKING model, which accounted for the speech situation, setting and scene (S), the participants (P), the speaker (S), the speech event’s outcome, purpose and goals (E), the act sequence of message form and content (A), the key, meaning the manner and spirit of the communication, instrumentality and channel (K), the norms of interaction and interpretation(N), and the genre, which included categories like lectures, poems, advertisement, etc(G). These features were central to what came to be called the ethnography of communication.
4. The sociological tradition
Sociologists discover the orderly phenomena of groups that emerge out of aggregate behavior. Fishman was an early leader in what he called the sociology of language. He studied sociological divisions such as small group interaction, sociocultureal factors, and stratification to address issues of bilingualism, language maintenance, and language planning. At about the same time the ethnomethodology studies of Sacks, Shegloff, Garfinkel, Heritage, and others took as their goal the systematic analysis of conversational structure and focused on the social organization associated with smaller units of spoken language interaction. The ethnomethodologists discovered diverse particulars of everyday speech, such as turn taking behavior, adjacency pair organization, repair activities, opening and closing activities, and evidence of preferred and dispreferred language.
5. The psychological tradition
By the 1950s various psychologists had come to understand George Miller’s belief that verbal behavior permeates much of human psychology and that a strong link to linguistics was necessary. The development of cognitive psychology was a good match for generative linguistics and thus cross-disciplinary work grew, particularly with the work of Bellugi, Berko-Gleason, Bever, Ervin-Tripp, and others who also associated their work with bilingualism and language style. The 1960 ground breaking paper by Brown and Gillman, “The Pronouns of Power and Solidarity,” aptly illustrated the connection of psychology to linguistics, culture and society. And psychological work on language attitudes by Fred Williams and others began to attract linguists, as illustrated by Labov’s early studies and Preston’s more recent work on perceptual dialects.
The defining moments of 1964
In the history of disciplines, sometimes there is a single pivotal point, one series of events that brings unity and clarity to diverse strands of thought. In my opinion, 1964 was the critical year for the current synthesis of language, culture and society. Early in that year the Social Science Research Council held a special meeting that brought linguists and social scientists together, Bright did the same at a meeting at UCLA, and a similar event took place at the LSA summer institute at Indiana U that same year. Among the participants at some or all of these meetings were Hymes, Gumperz, Bright, Ferguson, Haugen, Goffman, McDavid, Marckwardt, Francis, Fishman, and the rising star, Bill Labov, who at that point was finishing his PhD at Columbia. I was fortunate enough to have a post doc grant to that LSA institute to try to help figure out how to computerize Linguistic Atlas records, but hearing, meeting, and associating with this group of giants was a game changer. It was then that the term, “sociolinguistics,” began to be used to describe the discussions and publications growing out of these meetings, even though Labov and some others said that there was no real need for this term because linguistics cannot be isolated from culture and society. Instead, he believed, linguistics should embrace and include all three.
After a few years of continuing development of this synthesis, the 1972 Georgetown Round Table brought together the then major recognized leaders to try to outline the principles of what Hymes called “socially constituted linguistics.” Topics of the symposium ranged from optional rules in grammar and polylectal grammars, to conversational analysis, the sociology of language, language planning, and survey methodology. Fillmore stressed that although armchair linguistics also had much to tell us, sentences were not the maximal units of linguistic analysis, indicating that the functions of language noted by Hymes must surely be included.
Methodological development in language variability
Almost single-handedly, Labov altered the direction of methods in dialect study and at the same time the influences of anthropology, sociology and psychology have had profound effects on the methodology of what has come to be called sociolinguistics. Sociologists like Fishman and others had criticized the simplistic nature of the social structures used by dialectologists to describe language variation. Labov provided useful answers to this criticism, as evidenced first by his earlier New York City study and throughout his career ever since. He brought sophistication to sampling techniques, social categorization, and statistical measurements that have endured in his more recent research on phonological change in progress.
All of this took place during the height of the generative grammar revolution and descriptivists were not deaf to this. Some, including traditional dialectologists, fought a rear-guard battle, rejecting the new theoretical positions. Others tried to keep the best of the old ways while at the same time accommodating them with what they considered the best of the new ideas. Various attempts to accommodate generative grammar with language variability in social contexts were devised during the 1960s and 1970s, such as DeCamp’s introduction of implicational scales, Stewart’s concepts of “acrolect” and “basilect,” and Bailey’s “wave theory.”
Improvements in sampling techniques and statistical analyses were first introduce by Labov in his stratified sample of New York City speakers, a technique later used by Shuy and Wolfram in their study of Detroit speech and by Trudgill in his Norwich research. Labov created the concept of variable rules. Cedegren and Sankoff developed statistical models to fit rule probabilities into the competence and frequency rules of performance. Despite the growing sophistication of methodology, some critics complained that sociolectal grammars are merely abstractions derived from a group of idiolects and a potential weakness of the new efforts to describe language variability came from what Labov himself identified as “the observer’s paradox,” referring to the difficulty of observing behavior without affecting it. The alternative to this, of course, would be to surreptitiously record speakers representing various social categories. This idea was quickly rejected by academic researchers on ethical grounds.
Although the ethics of academic research prohibited our gathering data surreptitiously, law enforcement agencies had begun to carry out undercover sting operations in which there were no such ethical restrictions. Thanks to technological developments in recording equipment in the late 1970s and continuing to this day, various police departments and government agencies like the FBI began to record suspects in many sting operations. This has produced a treasure trove of naturalistic language that usually avoids the observer’s paradox. Unfortunately, the language research field has not yet found ways to make use of it. Although the makers of these tapes did not control for the usual social, regional, or ethnic categories, it is possible to discover them after the natural language recordings had been made. The storage area of my home contains thousands of hours of such recordings made by law enforcement, waiting for researchers to make some kind of use of them if and when some way can be discovered to do so. Unfortunately, to this point I have been unable to find an academic home for them where scholars can make use of them.
Today, the willingness of the formerly separated disciplines of linguistics, anthropology, sociology, and psychology to share their interests, information, methods, and goals has led to the current closer relationship of these fields in what can be called the area of language, culture, and society.
From the dialectology tradition we preserved the long held focus on individual and group variability of language, and thanks to Labov and others, many changes have taken place in terms of sampling, statistical measures, and other matters. From the anthropology tradition we can now identify the macro features of language variation in terms of the speech events in which the language occurs, for as Gumperz pointed out, speech events are “tacitly understood rules of preference, unspoken conventions as to what counts as valid and what information may or may not be introduced (1990, 9). Hymes’s units of channel, code, topic, and setting provide additional measures of the social and cultural aspects of language. From sociology’s tradition of micro analysis tradition we can now identify the regularities of conversational structure, including mechanisms of repair, opening and closing routines, adjacency pairs, and turn taking behaviors. From the language contact and psychological traditions we can now better understand the process of bilingualism, pidgins and creoles, sign language, stylistic variability, and the power structures of language use.