11.

The Social Dimension in Area Linguistics

11.1 Social Dialectology

Until rather recently the techniques of area linguistics have been brought to bear almost exclusively on folk speech. All of the European atlases have, in principle, restricted their selective sampling to this social level of usage. Nevertheless, the sociocultural interpretation of the geographic patterning of variants in folk speech has inevitably led to the recognition of prestige dialects, whether geographic or social, as a potent force of diffusion within the several subareas as well as from area to area. Lacking detailed information on regional variation in prestige dialects, scholars resorted to written records of the literary language to buttress their theories. The chief weakness of this procedure results from the fact that innovations in folk speech are usually not prompted directly by cultivated speech but are mediated by middle class usage, about which so little is known. Only systematic sampling of middle class and upper class speech will provide the evidence for dealing with diffusion realistically.

In the United States, the very first linguistic survey undertook the sampling of usage on three social levels: that of the folk, the middle class, and the cultured. In principle, folk speech and middle class speech were sampled in every community chosen for investigation, cultivated speech in about one out of five. This sampling plan is described in Kurath 1933: 90-95 and Kurath-Bloch 1939: 41-44. The cartographic presentation of two or three social levels community by community offered no serious problems, as one may gather from the maps of the Linguistic Atlas of New England [Kurath-Bloch- Lowman 1939-43: Preface to Vol. I].

Limitation in the number of cultured speakers seemed justified on the ground that their usage might be expected to vary less from place to place because they have wider contacts than the folk and the middle class and are more strongly influenced by literary English. To obtain some insight into differences existing between cultivated persons in larger urban communities, 5 were interviewed in New York City, 4 in Charleston, S. C., 3 each in Philadelphia, Baltimore, and Richmond, and 2 in Boston. These informants include old and middle-aged men and women, members of different professional groups, social organizations, and religious denominations.

Despite the rather small number of cultured speakers included in the American survey, the evidence gathered proved to be sufficient for the following purposes:

(1) to establish the phonemic system of cultivated speech region by region and to convey in synoptic tables the phonic realization of the phonemes by individual speakers [Kurath-McDavid 1961: 31- 100];

(2) to point out some of the more salient features in which cultivated pronunciation differs from the usage of the folk and the middle group, region by region [Kurath-McDavid 1961: Maps 28 mountain, 69 can’t, 72 rather, 77 calm], or agrees with it [ibid.: Maps 38 stairs, 97 creek, 113 root];

(3) to trace the diffusion of certain features on the level of cultivated speech from focal areas to surrounding areas [Kurath in Lunt 1964: 135–44; Van Riper 1957];

(4) to establish the important fact that in the several regions of the Eastern United States the phonemic system of folk speech is in essential agreement with that of cultivated speech; that the differences between them are largely nonstructural, consisting as they do in divergent lexical incidence and in the phonic realization of shared phonemes;

(5) to orient the intensive investigation of social dialects in urban centers and their environs, as several recent studies discussed below have clearly demonstrated.

11.2 The Investigation of Urban Speech

No systematic survey of urban speech comparable with surveys of folk speech in extent or in scientific technique has been undertaken as yet. Not that the importance of urban speech has been ignored or underrated. The dominant influence of urban centers upon the speechways of the countryside and the “vaulting” of urban speech forms from major centers to subordinate centers have been fully recognized. See Chapter VTII above.

The preoccupation with the speechways of the folk as such and as a source for tracing the development of standard or literary usage accounts in part for the neglect of urban speech on the part of dialectologists. But the chief reason would seem to be the bewildering complexity of the linguistic situation in the major cities, most of which have grown like mushrooms since about the middle of the nineteenth century, drawing their population not only from their surrounding but from other dialect areas and from remote countries. These complexities demand refinements in the sampling technique and in the analysis of the field data that are yet to be developed. Tape recording and the computer are devices that will have to be brought to bear upon this important task.

In the United States the investigation of city speech has barely begun, but the importance of it has come to be rather sharply felt. The realization that three-fourths of the American people now live in urban areas, that the cities inevitably dominate the countryside socioculturally, and that the effective teaching of English in the city schools demands reliable information on socially marked differences in everyday speech are three factors that have brought this about [Kurath in PADS No. 49 (1970)].

Fortunately we are not entirely unprepared to face this urgent task. The wide-meshed linguistic surveys of three social levels in large parts of our country, notably the East, the Midwest, and the Pacific states, provide the background for singling out features that vary regionally and by social class and may therefore be expected to have currency within a given city. Preliminary sampling of such features should lead to the construction of effective questionnaires that within the means at our disposal would yield reliable data on significant differences in the speechways of the several social and age groups in the various subdivisions (quarters) of the urban area and its rural surroundings. Inclusion of the adjoining countryside in urban surveys is important because of the rather recent growth of American cities. The linguistic historian will want to determine what features are indigenous to the area in which the city is located, what usages have been imported from other areas, and what elements can be attributed to the literary language studied in the schools.

Recent investigations of the social structure of various cities or some of their subdivisions are another asset to the student of urban speech. Instead of making a more or less intelligent guess as to social types and their interrelations within the urban community he can in such cases be guided by the findings of the sociologists in his choice of informants and, in some instances, even interview the very same socially representative individuals.

A critical analysis of some of the studies devoted to problems of urban speech in North America and in Europe will make it clear that the investigation of the speech of any major urban area is a difficult task. Even when elements of the population whose first language is other than that of the dominant groups are set aside, the problems of adequately sampling the usage of the social groups from generation to generation and from section to section of the urban complex are formidable.

11.21 The Speech of New York City

A succinct summary of the phonological features of the speech of New York City (NYC) is presented by Kurath-McDavid [1961: 14- 15; 55-57], based upon the evidence gathered from 25 informants in 1941. Three social levels are represented in this sample, which is deliberately restricted to native-born New Yorkers and favors the older generation. Of the 25 informants 6 are cultured, 7 belong to the middle class, and 12 to the lower class; 12 represent the oldest living generation, 8 the middle-aged, and 5 the younger set.

In 1948, Yakira H. Frank, a native of NYC, completed a detailed study of the evidence provided by this sample of the Linguistic Atlas. To gain perspective, she included in her investigation the field records of 21 informants living in the surrounding area—the counties of Nassau and Westchester in New York State and the New Jersey counties adjoining the Hudson River, the bedchamber of many a New Yorker.

It is of some importance to note that the 5 boroughs of NYC— Manhattan, the Bronx, Queens, Brooklyn, and Staten Island—cover some 300 square miles and had a population of nearly 8,000,000 in 1941. Under the circumstances it is obvious that the evidence furnished by 25 informants is hardly sufficient for revealing regional differences within this huge urban complex with its sharply divergent economic and sociocultural subdivisions. For this reason, Frank’s treatment is focused on social and age differences within the urban complex as a whole.

The 25 informants interviewed in the five boroughs of NYC— whatever their social standing or age—have essentially the same system of phonemes. They share 6 “checked” syllabics, as in bit, bet, bat, hot, hut, foot; 8 “free” syllabics, as in bee-beat, bay-bait, buy-bite, bar-heart, and do-boot, no-boat, law-bought, and now-bout; and 24 consonants. The phonic realization of some of these phonemes and their lexical incidence vary by social and/or age groups, as pointed out below.

The only clear-cut systemic difference occurs in the syllabics boy, coil, loin, and burr, curl, learn. Here the cultured speakers have contrastive / ɔI ≠ 3 /, as in coil ≠ curl, while in lower class speech the two phonemes are merged in / 3I /, so that coil and curl, loin and learn are homophonous. Usage of the middle class varies and fluctuates in a complicated manner.

The chief social differences in the articulation of shared phonemes and their lexical incidence revealed by the sample can only be briefly pointed out.

On the folk level, (1) the syllabics in such words as bag and law are raised and prolonged in a majority of the test words; (2) the syllabic of buy-bite has a backed beginning; (3) the suffix of laughing, morning usually ends in /n /; (4) the fricative in three, moth is sometimes replaced by an affricated plosive and that in the, without by /d /.

The following phonological features are largely confined to cultivated speech: (1) the monophthongal articulation of the syllabic in burr, burn, i.e., [3^•] ;(2) the centered beginning of the diphthong in no, boat; (3) the occurrence of [ɑ^•] not only in bar, heart but also— though not consistently—in aunt, can’t, calf and in vase, tomato; (4) the use of [ai] in either; (5) the sequence / ju / in due, new, Tuesday.

Several phonological features are disseminated by age groups. Thus relics of lowered / ɔ /, as in law, and the incidence of the checked vowel /∪/ in root, broom occur only among the old informants. On the other hand, a rather open beginning of the syllabics bee, do, no characterize some of the young speakers.

Two phonological phenomena call for further comment: (1) the varying reflexes of postvocalic / r /, and (2) the merging of the syllabics in loin and learn.

In words like chair, car, door, poor, 14 of the 25 New York City informants had no trace of /r / in 1941, 9 of them had an occasional weakly constricted / r /, and 2 informants living on the periphery (northern Queens and southern Staten Island) used / r / with some frequency, thus agreeing with the usage of the surroundings of NYC, where the incidence of / r / increases with the distance from the city.

The evidence for this variable feature of the speech of New York City and the surrounding area is effectively displayed by Frank in tables IXA and IXB of her dissertation. It is indeed a challenge to further investigation.

The phoneme / ɔI / of boil, join is sharply distinguished from the / 3 / of learn, girl in cultivated NYC speech. At the lower end of the social scale, the / ɔI / is just as regularly merged in / 3 /, mostly articulated as an upgliding diphthong, so that loin and boil rhyme with learn and girl. Usage of the middle class varies. Though merging predominates, there is a trend to introduce the distinction, presumably under the influence of the schools.

The sampling of the speech of New York City for the Linguistic Atlas of the Eastern States has raised questions that can only be answered by more extensive sampling. That, indeed, is one of the functions of a large scale survey, which provides the data for designing effective plans for the investigation of specific problems.

11.22 Stratification on the Lower East Side of Manhattan

Twenty years after the sampling of New York City speech for the Linguistic Atlas, William Labov undertook a systematic investigation of six variable features in the speech of the Lower East Side of Manhattan for his doctoral dissertation at Columbia University (1964). The project was suggested and directed by U. Weinreich. With minor changes, the text of the dissertation has been published under the broad title The Social Stratification of English in New York City (1966). Its publication was sponsored by the Clearinghouse for Social Dialect Study of the Center for Applied Linguistics in Washington, D. C.

The choice of informants was guided by a previous survey of the social structure of the Lower East Side of Manhattan by social scientists whose findings are based upon 988 representatives chosen from a population of about 100,000 inhabitants. For the linguistic survey the sample was reduced from 988 to 81 as follows: 488 were dropped because their native language was not English; 180 had moved out; 125 of the remaining 320 were eliminated by “random” procedure; of the remaining 195, 157 could be reached, and 122 of them were interviewed at length, whereupon 41 of them were eliminated because they were born outside New York City. This left the 81 informants—45 of Jewish, 19 of Italian parentage—who “provided the main body of data.”

To what extent the sample of 81 selected in this manner from the sample of 988 used by the social scientists is representative of the social types inhabiting the Lower East Side is hard to tell. Nevertheless, no other investigation of social gradation in urban speech, in this country or abroad, has been in a position to chose its informants with comparable information and circumspection. Readers of this stimulating and provocative study must of course be aware of the fact that the determination of correlations between speech behavior and social class depends upon the degree to which the informants selected reflect the social structure of the community.

Labov’s immediate purpose is to determine the social dissemination of variant pronunciations and the incidence of these variants in several “styles” of speaking on the part of the social types represented by his sample population. From such data he hopes to infer trends in usage within the community—the Lower East Side of Manhattan—and to gain insight into the social dynamics underlying them.

To secure information on the behavior of the informants under varying circumstances, the author investigates five “styles” of utterance: (A) casual (informal) speech, (B) careful (formal) speech, (C) reading of texts, (D) reading of isolated words, (D¹) reading of minimal pairs.

The variable features chosen for investigation are: (1) the occurrence or loss of postvocalic / r /, as in here, hair, hard, four, sure; (2) the syllabic of bad, half, hand, found to vary all the way from [æ ~ æ^•] to [eə ~ Iə]; (3) the syllabic of law, off, more, said to vary from [ɔ^• ~ ɒ] to ɔə ~ ə]; (4) the consonants of thin and then, articulated as fricatives or as plosives.

The occurrence of the variants among speakers representing the upper middle class, the lower middle class, the working class, and the lower class, and the incidence of these variants in the “styles” of the several social classes are summarily presented in a set of graphs [222, 238, etc.].

The major findings are: (1) postvocalic /r /, as in hard, is increasingly more frequent from “low” to “high” class and from “styles” A to D. (2) Raised variants of the vowel in bad, represented by the digraph / eh /, are less frequent among the middle class than among the lower classes and decrease from “style” A to D. (3) Fricatives in thin and then are consistently more frequent from “low” to “high” class and from “style” A to D. (4) The social stratification of the raised variants of the syllabic of law, off, represented by the digraph / oh /, is less clear; but this variant is clearly more frequent among the lower classes than among the middle class and its frequency drops off sharply from “style” A to D.

If the statistics on which the graphs are based are accepted at face value, the inhabitants of the Lower East Side of Manhattan, whatever their social class, tend to break away from pronunciations current in their casual speech. They introduce the / r / in postvocalic position, tone down or abandon the raised vowels in bad and in law, and avoid the plosive consonants in thin and then. This trend is especially marked among the lower middle class.

Labov suggests that New Yorkers are on the defensive [Bright 1966: 100] and adopt other people’s pronunciations when they want to appear to advantage. That is surely true of the stigmatic plosives in thin and then, the only non-English features. But the adoption of postvocalic / r / may have a more complicated background. Not only is / r / rather regularly heard both before and after syllabics over the radio and seen in print, but also nearly half of the 81 informants seem to be bilingual, speaking Yiddish or Italian in their homes and with friends, languages in which / r / occurs in all positions.

The author attaches great importance to the consistent correlation of the relative frequency of the variants with social class and to the regularity with which the variants either decrease or increase from casual to careful speech, and further to the reading of texts and minimal pairs. The behavior of the informants in “styles” of utterance (or levels of diction) betrays in part an awareness of prestigeous variants and in part the influence of the written word. These findings should stimulate scholars to apply Labov’s technique to other speech communities, so that his view of linguistic change could be tested by other evidence.

Whether linguistic usage in other sections of Manhattan and in the other four boroughs of NYC exhibits the same social and stylistic patterning, as Labov claims, is more than doubtful. To be sure, some of the same “variables” have been observed in other parts of New York City by earlier investigators—from Babbitt [1896] to the Linguistic Atlas [1941; Frank 1948], and from C. K. Thomas [1942] and A. F. Hubbell [1950] to A. Bronstein [1962]. But their social dissemination still remains to be investigated. The very fact that Labov found such a striking correlation between the dissemination of these variables with social indices on the Lower East Side suggests that other sections of Manhattan, characterized by very different social structure, can hardly conform to the same linguistic pattern. Only investigation comparable in method will give us reliable information on other sections of Manhattan—say Harlem and Washington Heights—and on the other boroughs of NYC. There is no substitute for informed carefully planned selective sampling. Until that is done, scholars will do well to set aside Labov’s enthusiastic assertions couched in these words [Bright 1966: 107]:

It is the great merit of Labov’s investigation to have taken full advantage of the results of a systematic sociological sampling operation of a limited section of Manhattan, to have drawn upon this sample of informants to obtain linguistic evidence for establishing correlations between phonological “variables” and social factors, and thus to determine trends in usage in that section of Manhattan. This is a remarkable achievement, destined to be a model for future investigations of urban speech communities. Wherever comparable sociological surveys become available, linguistics will have a golden opportunity to make significant contributions to our understanding of the complex social forces underlying linguistic change in urban areas.

11.23 Social Gradation of Negro Speech in Washington, D. C.

Wide-meshed sampling of Negro speech for the Linguistic Atlas of the South Atlantic States provided evidence for the following statement [Kurath 1949: 6]:

The Gullah dialect spoken along the coast of South Carolina and Georgia appears to be the only exception to this rule (see Section 7.4 above).

That marked social dialects in Negro speech were current long before the abolition of slavery is clear enough: houseservants, separately quartered from the field hands, acquired the language of their masters. The problem of the interrelations between the speech ways of Negroes and Whites has been discussed by the McDavids [1951], with appropriate comments on popular misconceptions.

The social gradation in present-day Negro speech has recently been investigated experimentally by G. N. Putnam and E. M. O’Hern [1955]. Seventy observers were asked to rate the speech behavior of 12 Negroes who had retold a short fable read to them. Recorded separately on tape, the 12 versions of the story were transferred to a master tape in a random order determined by lot. The observers were instructed to place each speaker on a one-dimensional social scale from “high” to “low.” Of the 70 “judges,” 54 were graduate students and 16 teachers; 55 were White and 15 Negro. Few of them were trained in linguistics.

The 12 speakers, of whom 3 were university professors, 2 secretaries, 2 masons, 2 maids, and 3 inhabitants of a slum area in Washington, D. C. had previously been ranked socially by the investigators with reference to an “index of status characteristics” (I. S. C.) based on 4 factors: occupation, source of income, house type, and dwelling area [25–26].

The ranking assigned to the 12 speakers by the “judges” correspondedvery closely to that based upon the “index of status characteristics.” In the words of the investigators: “The product-moment correlation between the I. S. C. scores of the twelve speakers and the mathematical equivalents of the judges’ ratings was +0.80.” This striking correlation is graphically presented in Figure 7 [27].

It is of importance to emphasize the fact that the only information the “judges” had about the 12 speakers was their performance in retelling a short story. Their judgments are based upon a variety of factors, probably in this order of importance: skill in describing the sequence of events, “styling” of the story, prosodie features, syntactic features, verb forms, features of pronunciation. Some insight into the cues on which the judges may have relied can be gained by reading the 12 versions of the story presented in phonetic notation [30–32].

One can readily agree with the concluding remarks of the authors:

11.24 The Negro Community in Memphis, Tenn.

An important contribution to our knowledge of Negro speech in an urban community has been made by J. V. Williamson in her study of the speechways of the Negro community in Memphis, Tenn. [1961]. In this city one-third of the population of half a million belongs to this race, which at the time of the field work had its segregated school system leading from grade school to Le Moyne College, where Miss Williamson teaches. Drawing its population, White and Black, largely from the cotton-growing Delta along the Mississippi, Memphis has had a phenomenal growth, doubling its population every 20 years: a city of 65,000 in 1890, it had reached 500,000 by 1960.

The sampling plan conforms to that of the Linguistic Atlas, but is significantly extended to bring social classes and age groups into relief. Of the 24 informants, 6 are cultured, 10 belong to the middle class, and 8 to the folk; 9 are under 45, 5 over 65, and 10 middleaged. The evidence gathered shows that nonstandard verb forms are largely confined to the folk level and are matched by forms used by Southern Whites. On the other hand, the cultured Negroes employ only the standard literary verb forms, while the usage of the middle class informants varies. A similar situation obtains in phonological matters. There is not a single feature of pronunciation or grammar that lacks its counterpart in the speech of Southern Whites as recorded for the Linguistic Atlas.

Miss Williamson did not investigate the speech of the White population of Memphis, which would indeed be a difficult task for a Negro scholar. One suspects that the socially marked forms of Negro speech are also current among the Whites of the same social level. We shall see that in Chicago, where the Negro population represents a recent mass immigration, the linguistic situation is radically different.

11.25 Negro Speech in Chicago, Ill.

Two sampling surveys of certain aspects of the speech of Chicago have been carried out in recent years. L. A. Pederson’s The Pronunciation of English in Chicago [1964] deals with the usage of the White population in its social gradation and forms the background to the investigation of the speech of the large Negro community within the city carried out by a group of scholars under the general direction of R. I. McDavid, Jr. [1966] and published in mimeographed form under the title Communication Barriers to the Culturally Deprived.

The latter study, supported by the U. S. Office of Education, undertook to establish the existing differences between the speech of Whites and Negroes, to determine the social evaluation of these differences on the part of Chicago Whites and Negroes, and thus to pinpoint those features of Negro speech to which the schools might direct their attention for the good of the Negro.

The director of the project, a native of South Carolina and long a resident in the Midwest, was uniquely qualified to design this research project after years of fieldwork both in the South and in the Midwest. His long interest in social dialect as a reflection of social structure and its implication for the life of the individual [1946, 1948, 1951, 1965, 1966] made him fully aware of the problems to be met. Moreover, he was fortunate in having as his associates a number of excellent linguists.

Previously gathered dialect evidence for the Midwest and the South—the former as background for the usage of the dominant White population of Chicago, the latter as background for the practices of the newcomers—made it possible to select 160 critical items from the worksheets of the Linguistic Atlas to determine the more striking phonological and grammatical differences between the usage of Whites and Negroes.

Thirty Negroes were interviewed, 16 by L. A. Pederson and 14 by J. Willis, a Negro graduate student in anthropology. Pederson, who had previously investigated the usage of Chicago Whites [1964], recorded the responses to the 160 items in phonetic notation, Willis on tapes.

The informants range in age from 30 to 70 and represent all social classes. Six of them were born in Chicago (aged 30 to 60), 22 in the South, chiefly in the South Central States. Of those born in the South, 12 had lived in Chicago at least 20 years.

The Negro respondents are ranked from high to low in accordance with a complex sociocultural index, and in the 10 tables of variants their responses are consistently entered in the established sequence. In the tables, agreement with the usage of the Whites and deviations from it are identified by distinct symbols, so that any marked differences in the behavior of upper-class and lower-class Negroes can readily be seen. So can the complexities of the actual situation.

We may take as an example the reflexes of postvocalic / r /, as in beard, chair, horse, four, which are presented by Pederson in Table 2. As is well known, in the Midwest / r / survives in that position as a constricted consonant [], while it appears as an unsyllabic [] in the speech of Whites and Negroes in the coastal South and large parts of the south-central states. With 2 exceptions, the 17 lower-class informants born in the South rather consistently have [] in these words, even the 7 that have lived in Chicago for more than 20 years. This is what one might expect. The surprising fact is that 4 of the 6 Negroes born in Chicago (age 30 to 60) rather consistently follow this Southern practice. Only 1 of them, a 30-year-old clerk, has adopted the postvocalic / r /. Residential segregation would seem to account for this situation.

The divergent behavior of 3 upper class informants who came to Chicago after age 30 raises some interesting questions: a lawyer born in western Tennessee, aged 57, consistently uses []; a social worker born in eastern Virginia, aged 69, has constricted [] with fair consistency; a minister born in Arkansas, aged 69, uneasily fluctuates between [] and []. Personal attitudes as well as certain contacts conditioned by professional activities are surely involved.

The rating on the part of Chicago Whites of existing differences in pronunciation can be expected to identify features that stigmatize the Negro and thereby put him at a disadvantage within the urban community. In turn, the identification of such “inferior” features furnishes the necessary leads for planning the teaching of the socially superior variants to the Negro, variants that he can use at will when the situation requires it.

In a carefully planned experiment, V. S. and C. H. Larsen undertook to determine an evaluation of some of the more striking phonological differences between the speech of Negroes and Whites. White observers were asked to record on a five-point scale their reaction to specific items, presented to them on tapes, along such dimensions as upper class-lower class, educated-uneducated, pleasantunpleasant, urban-rural, White-Negro. The essential agreement in the judgments of the White observers leaves no doubt about the social standing among Chicago Whites of the variants tested. See, for instance, “exhibits” 1.12 four and 1.14 poor. Negro observers, the Larsens found, largely agree with the reactions of the Whites, although they are less sure of their judgments.

On the assumption that preferred usage tends to spread, one feels tempted to say that tests of this kind could be used to determine probable trends in usage within a speech community.

The magnitude and the difficulty of the task confronting the teaching of a prestigeous type of English to the children of the Chicago Negro community are formidable. Family and neighborhood tradition in a residentially segregated Negro population of 800,000— built up largely since the First World War—are powerful, as this investigation has shown. To make headway at all, the identification of stigmatized features is a necessary preliminary to the teacher’s problem.

11.26 The English of the Mexican Community in San Antonio, Texas

Using a modified version of the Linguistic Atlas questionnaire, J. B. Sawyer interviewed 7 informants of Anglo-American stock and 7 of Mexican extraction in the city of San Antonio, Texas. She discovered that “Anglos” and “Latins,” speaking their vernaculars and living lives apart, do not influence each other to any extent.

Miss Sawyer’s findings, briefly summarized in Word [15.270- 781], are presented in a doctoral dissertation written under the direction of E. B. Atwood (University of Texas, 1957).

The speech of the “Anglos” is characterized by phonological features derived from the Southern and/or the South Midland dialects as described in Kurath-McDavid [1961: 18–22], among them an upgliding diphthong in law, fog; [æu] in cow, house; monophthongal [a^•] in five, night; and contrastive vowels in hoarse ≠ horse. None of these features has been adopted by those whose family language is Spanish, not even by those who speak English with ease.

In similar fashion, Southern regional words used by the “Anglos,” such as clabber “curdled milk,” corn shucks “corn husks,” light-bread “white bread,” pulley bone “wishbone,” snap beans “string beans,” are unfamiliar to most of the “Latins,” who acquire the literary terms taught in the schools.

The same situation obtains with regard to regional folk forms of the verb current among the lower classes of San Antonians of English descent.

Although Spanish is now spoken by nearly half the population of the city, there is no evidence of recent Spanish influence on the English spoken by the “Anglos,” not even in the vocabulary. Such terms as arroyo “dry creek,” mesa “table land,” corral “horse pen,” remuda “group of saddle horses” were adopted during early contacts (ca. 1850) and survive in San Antonio as elsewhere in Texas [Atwood 1962: Maps 15, 19, 21, 22].

Flocking into the city in ever increasing numbers since 1900, chiefly as laborers, San Antonians of Mexican descent constitute a world apart, maintaining contacts with Mexico and preserving their folk culture in their quarters of the city. On the other hand, the English-speaking stock keeps aloof.

11.27 The Relations between Urban and Rural Speech in Two Swiss Cities

The social stratification of urban speech in two Swiss cities is discussed with fine discrimination and careful documentation by H. Baumgartner [1940], associated with Hotzenköcherle in drawing up the plans for the Sprachatlas der deutschen Schweiz. His native city of Biel and the city of Bern, where he conducted his seminars on dialectology, are his two chief observation posts for accumulating the large body of evidence that enabled him to place urban usage in its regional setting, to characterize the behavior of the several social groups, and to outline trends in urban usage.

According to Baumgartner, all social groups in Biel and in Bern speak variants of one and the same regional dialect of Swiss German, that of the Seeland and the Mittelland of the canton of Bern. The speech of the city-born leading families normally has a larger number of words adopted from literary German (taught in the schools) than that of the other social groups, but its systemic phonological and morphological features and the basic vocabulary are essentially the same as those current in the Bernese countryside.

Baumgartner distinguishes three social layers in Swiss urban communities, separates the old stock from the newcomers on all three levels, and points out their linguistic peculiarities in some detail.

In the city of Biel, the speech of the lower class, agreeing with that of the countryside, differs from that of the upper class in such phonological items as chaub ≠ chalb “calf,” finga ≠ finda “find,” doiff ≠ dieff “deep,” föif ≠ ffif “five” [21].

The upper class uses some French and Standard German words that are not current on the lower levels, articulates the /r / in the French way as a velar fricative, replaces the suffix -ig, as in tsitig “zeitung, newspaper” and meining “meinung, opinion” by Standard German -ung, and introduces some “refined” verb forms [22–33]. The middle group, less clearly definable and less sure of itself, wavers between “high” and “low.”

Despite such “class markers” all varieties of the speech of Biel and Bern are mutually intelligible, and many a native is bidialectal and able to shift his “style” with the circumstances [36–38], Children of the upper class usually learn the speech forms of the lower groups from their playmates in the schoolyard.

Diffusion from “high” to “low” is clearly observable in features first adopted by the upper class from Standard German, as pinsel for pämsel “brush,” chirchə for chilchə “church,” sōn for sūn “son,” buttər for ankchə, schinkchə for hammə “ham” [41-42].

Spreading from “low” to “higher,” though deliberately resisted by the upper class, is not unknown, especially in recent times. Thus the phonological type of miuch is replacing milch “milk,” fingə is encroaching upon fində “find,” and i gā is taking the place of i gangə “I am going” among the middle class. Even the children of the upper class are adopting these “low” variants, which are supported by the usage of the countryside [44-46]. For the dissemination of two of these features in Swiss folk speech see Hotzenköcherle 1965: Maps 2.148 folge and 2.119 finden.

Baumgartner summarizes the linguistic dynamics of the Mittelland and the Unterland of the canton of Bern in an interesting diagram [102]. From the city of Bern, with a population of 125,000 in 1940, upper class features spread in all directions to the upper classes in the smaller cities—northward to Biel and Solothurn, southward to Thun, and westward to Freiburg. In these subsidiary centers they infiltrate the lower social levels and may ultimately find their way into the surrounding villages.

11.28 Recent Phonological Innovations in Vienna

The popular speech of Vienna (Wiernerisch), a variety of Central Bavarian similar to that of Munich, has been the object of careful analysis from the structural point of view for half a century. It is spoken in the homes of most native Viennese, regularly used in public by the lower and middle classes, and of course familiar to those among the upper classes who no longer speak it in their homes. This vernacular, often referred to fondly as echt Wienerisch, has been the vehicle of popular plays and songs for many years.

In the schools of Vienna a very different kind of German has been taught since the middle of the nineteenth century: the literary language (Schriftsprache), which is based upon the dialects spoken in east central Germany. As a result, all Viennes are bidialectal in one way or another. Some normally speak the vernacular but readily understand school-taught German; others have full mastery of both dialects and use them at will in accordance with the circumstances; others again understand the vernacular but habitually speak a variety of Standard German. All of them read the Schriftsprache in the daily press and other publications, however they may pronounce the printed word.

The phonological structure of the Viennese vernacular has been briefly outlined above [8.6]. Here we shall consider some of the changes in the system of syllabics that have occurred since the turn of the century or are now taking place. They are discussed by E. Kranzmayer in a spirited article entitled Lautwandlungen und Lautverschiebungen im gegenwärtigen Wienerischen [1953].

At the turn of the century, the Viennese vernacular had a symmetrical four-level system of oral monophthongs. The author calls this stage of the dialect Altwienerisch [209]. By the time of the First World War, the monophthongization of the syllabics in weit, Maul, Haus among the younger generation produced the five-level system of oral monophthongs described by A. Pfalz [1918] and reported by B. J. Koekoek [1953: 12] as characteristic of the older generation in midcentury.

A generation later—at the time of the Second World War—the symmetry of this five-level system was “shattered.” While contrastive close /o / and open /o / were preserved, the close front vowels / e / and / 5 / were merged in the open vowels / Õ / and / 0 /, respectively, as shown by the arrows in the vowel diagram above. How this asymmetry, this “hole” in the system, was mended is discussed at length by the author [219ff.].

Kranzmayer attributes the merging of the two sets of mid front vowels to the influence of Standard German (Hochsprache), taught in the schools of Vienna, but without analyzing the complicated background. A simple listing of the conflicting correspondences between Standard German (SG) and the Viennese vernacular (W) in typical words reveals the problems faced by the teacher of SG:

Under the circumstances the teacher would presumably tell his pupils to pronounce the vowels in accordance with the spelling, sounding the graphemes e, eh, ä, âh without rounding the lips and the graphemes ö, öh with rounding. In the process the distinction between close and open mid front vowels would be eliminated in reading Standard German texts—whence ultimately also in the vernacular of the younger generation.

Two other changes in the phonemic structure of Wienerisch were in progress by 1950: denasalization of nasal syllables as in hin /hĩ /, schön / šẽ /, Bein /bã /, Sohn / sũ / [231].

Kranzmayer is inclined to attribute denasalization to “Czech infiltration” [233]; and yet, a similar change has recently taken place in the Central Bavarian dialect of Munich [Kufner 1962: 71–73], where such an influence is unthinkable. Why not simply the effect of school-taught Standard German?

The merging of the strong postvocalic fricative /ff, ss, xx /, as in offen, wissen, machen, with their weak counterparts, as in Ofen, Wiesen, weihen, is even more recent. It is said to occur in the speech of less than 15 percent of persons under 25 and to vary from word to word [227–28].

Kranzmayer links these striking changes in the phonology of Wienerisch since ca. 1900 to the social upheavals resulting from the two World Wars. The impoverishment of the linguistically conservative upper classes are said to have given free rein to trends inherent in the Wienerisch of the lower classes and to the intrusion of features of Standard German. This seems plausible enough as a general theory. And yet, some of the author’s attributions to specific social forces and to certain inherent linguistic drives, advanced with no little enthusiasm, fail to carry conviction. Disagreeing sharply with the views of the Sprachatlas group at Marburg, he underrates diffusion through communication and overstates the case of parallel spontaneous phonological change emanating from drives inherent in the phonological system, which he calls “polygenesis.”

The author’s challenging presentation of striking phonological changes in the former capital of the Austro-Hungarian Monarchy lays the foundation for planning a systematic sampling survey of the speech of Vienna in its social and regional manifestations. The 500 informants interviewed by the author over a three-year period [200]—representing different age groups and several social classes in the 21 wards of the metropolis—could hardly provide the evidence required for a detailed description of the varieties of German current among social and age groups in the several wards (Bezirke) of the city. A set questionnaire levelled at known or expected variants and a systematic selection of representative speakers are essential in dealing with complex linguistic situations.

11.29 Phonological Variants in the Speech of Young Parisians

In 1957 R. Reichstein sampled the usage of 556 girls enrolled in public and private schools of 5 of the 20 wards (arrondissements) of Paris with respect to three variable phonological features: (1) preservation or loss of contrastive low front and low back / á ≠ à /, as in pattes “paws” vs. pates “pastries”; (2) preservation or loss of contrastive short and long / ε ≠ ε̄/, as in belle “beautiful” vs. bêle “bleats”; (3) preservation or loss of contrastive unrounded and rounded //, as in brin “stalk” vs. brun “brown.”

The examples were presented to the informants on slips in such sentences as il march à quatre pattes “he walks on all four” and il mange des pâtes “he eats pastries.” They read the sentences and then repeated them by heart.

The informants were assigned to three social levels on the basis of the occupations of their fathers [84]. In the statistical treatment of the responses only the usage of the daughters of “professionals” and of “workingmen” are taken into account, the middle group being set aside. Even with this drastic simplification of the statistical problem the results of the investigation are highly significant.

These are Miss Reichstem’s chief findings:

(1) The preservation of the three contrastive pairs of syllabics differs markedly from ward to ward, being consistently highest in Ward 16 and lowest in Ward 18 [71].

(2) In the total sample of 556 informants, the preservation of contrastive /á ≠ à /ranges from 72 to 43 percent; that of / ε ≠ ε̄/ from 58 to 23 percent; that of // from 30 to 10 percent. It is suggested that the isolated quantitative distinction represented in /ε≠ ε̄/ and the small “functional load” of // favor the loss of these two contrasts.

(3) There are marked differences between the social extremes and between pupils in public and private schools [85–86], but without any clear-cut pattern from ward to ward.

(4) The preservation of contrastive / á ≠ à / varies from word to word, being 50 percent in pattes vs. pâtes but only 10 percent in rat “rat” vs. ras “shaved” [61].

This may be enough to suggest the complexity of the social and regional behavior of the variants of the three Parisian features investigated by the author. It is more than likely that phonological changes in progress in such cities as Vienna and New York exhibit similar complexities in their social and regional dissemination.

Reichstein rests content with a statistical analysis of the linguistic data in their relation to social groups and schooling in the 5 wards of Paris she chose for investigation—and wisely so. As A. Martinet points out in his concluding remarks to this article [96-99], all of the 20 wards of Paris as well as the suburbs will have to be investigated before the social drives and the controlling structural factors can be established with some degree of probability.

11.3 Concluding Remarks

In recent years there has been a quickening interest in social dialects and the relations between different languages spoken within political domains, and in determining their relations to social structures and sociocultural forces. In America such publications as those of C. A. Ferguson [1959], D. Hymes [1964], J. Gumperz and D. Hymes [1964], and W. Bright [1966] bear witness to a spirited discussion of a field of research that lay all but dormant for the two or three decades devoted to the structural analysis of a great variety of languages and the refinement of appropriate theories and techniques for the pursuit of purely synchronic studies.

The results achieved in structural linguistics will be an invaluable asset in dealing with the complicated problems of social dialectology. It is obvious that the differences between the dialects current in a community or area must be identified as structural and substructural, and that incidental variants must be evaluated as such if the description of the dialects is to be meaningful, i.e., scientifically sound. Whether the study of language in a variety of social settings will lead to modifications of some of the concepts underlying synchronic structural linguistics remains to be seen.

The student of social dialects has also at his disposal the theories and the results of diachronie linguistics, a major achievement of the nineteenth century. Here the central concept is the regularity of phonemic change as finally formulated by the Neo-Grammarians. Based upon regular phonemic correspondences existing between several stages of one and the same language or between languages of the same linguistic stock, it implies the recognition of the fact, so clearly demonstrated by synchronic linguistics, that every language or dialect is structured phonemically. Being so structured, it is to be expected that the units of the phonemic system will change in the same way in all the morphemes in which they occur, unless prosodie factors or cross influences in the grammatical or dérivational system interfere. Deviations from the normal change of a phoneme that cannot be accounted for by such internal processes are attributable to external influence, i.e., to adoption from another language or dialect. Unless the student of social dialects chooses to be content with a purely synchronic description, he will rely upon the theory of the regularity of phonemic change to distinguish between internal developments and importations.

Since the adoption or adaptation of features of another dialect is a common phenomenon, the dialectologist is in a strategic position to elucidate the nature of such processes and the social conditions underlying them. He may even be able to demonstrate that under certain circumstances regular phonemic change can result from consistent replacement of a native phoneme by that of a dominant dialect.

Some of the techniques that have been employed in determining the geographic (spatial) dissemination of dialect features and their interpretation are directly applicable to research in social dialectology, and others can be rather easily adapted, such as preliminary sampling, the construction of a questionnaire, the conduct of the interview, the determination of correlations between speech areas and sociocultural domains leading up to historical interpretation. But some innovations and refinements in procedure are required, especially a prior investigation of the social structure of the area or community on which the choice of representative informants must be based. Recording on tape of elicited responses as well as spontaneous conversation or narration will play a major part in coping with the usage of bidialectal speakers.

Although little systematic work has been done in social dialectology, awareness of the currency of two or more social dialects in one and the same community or geographic area has of course not been wanting. In German-speaking Europe three major levels, more or less clearly defined area by area, are generally recognized [Henzen 1938: esp. 179-94; Moser I960]: (1) cultivated speech (Hochsprache), which tends to approximate the national standard taught in the schools; (2) folk speech (Mundart), peculiar to the farming population and to country folk in general, which is used in the home and in dealing with one’s neighbors; (3) regional colloquial speech (Umgangssprache, Verkehrssprache), which serves as the medium of informal public communication within a major dialect area or political domain.

The relation between these three social levels of speech varies strikingly from area to area. Thus in Switzerland regional colloquial speech is rather clearly set off from cultivated and from folk speech. Here many a native is tridialectal, using each dialect in its proper social setting. In Württemberg, Bavaria, and Austria the situation is similar, although folk speech is not as widely used as in Switzerland. Also in Scotland three social dialects are current: folk speech, standard Scots, and standard English [Grant 1921: xx–xxii; Mcintosh 1952: 28–33].

Many of the problems with which research in social dialectology is faced are also encountered by the student of bilingual communities, such as the correlation between the several languages with social groups, the conditions that prompt the choice of the medium of communication on the part of bilingual speakers, and diffusion from one medium to the other.

Since the lines are more sharply drawn between languages than between dialects, the student of bilingual areas has certain advantages in dealing with such problems. For this reason it seems highly probable that the social dialectologist can improve his techniques and interpretations by keeping in touch with research now underway in the field of bilingualism.