anth 4 reader 12

anth 4 reader 12 - 42 SOCIOLINCUISTIC PATTERNS be more...

Info iconThis preview shows pages 1–15. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 42 SOCIOLINCUISTIC PATTERNS be more subtle and complex. Yet the basic approach. of isolating the socially significant variables. and correlating them with the patterns of general social forces, was the same as that which was used on Martha's Vineyard. We can expect that these methods will give us further insight into the mechanism of linguistic change. The Social Stratification of (r) in New York City Department Stores “As this letter is but a jar of the tongue. . . . it is the most imperfect of all the consonants." john Walker. Principles of English Pronunciation. 1791 ANYONE who begins to study language in its social context immediately encounters the classic methodological problem: the means used to gather the data interfere with the data to be gathered. The primary means of obtaining a large body of reliable data on the speech of one person is the individual tape-recorded interview. Interview speech is formal speech—not by any absolute measure, but by comparison with the vernacular of everyday life. On the whole. the interview is public speech—monitored and controlled in response to the preseHCe of an outside observer. But even within that definition, the investigator may wonder if the respOnses in a tapevrecorded interview are not a special product of the interaction between the interviewer and the subject. One way of controlling for this is to study the subject in his own natural social context— interacting with his family or peer group {Labov, Cohen, Robins. and Lewis 1968). Another way is to observe the public use of language in everyday life apart from any interview situation—vto see how people use language in context when there is no explicit observation. This chapter is an account of the systematic use of rapid and anony- mous observations in a study of the sociolinguistic structure of the speech community.1 1. This chapter is based upon Chs. 3 and 9 of The Social Stratification of English in New York City {1966], revised in the light of further work with rapid and anonymous observations. I am indebted to Frank Anshen and Marvin Maverick Harris for refer- ence to illuminating replications of this study (Allen 1965, Harris 1968). 43 44 SOCIOLlNGUlSTlC PATTERNS This chapter is the first of a series of six which deal primarily with the sociolinguistic study of New York City. The main base for that study (Labov 1966a} was a secondary random sample of the Lower East Side, and this data will be considered in the following chapters. But before the systematic study was carried out, there was an extensive series of preliminary investigations. These included 70 individual interviews and a great many anonymous observations in public places. These preliminary studies led to the definition of the major phonological variables which were to be studied, including {1'}: the prestmce or absence of consonantal [r] in postvocalic position in cor, cord, four, fourth, etc. This particular variable appeared to be extraordinarily sensitive to any measure of social or stylistic stratil’ict‘ition. On the basis 0]. the exploratory interviews, it seemed possible to carry out an empirical test of two general notions: first, that the linguistic variable (1') is a social difl'ert-zntiator in all levels of New York City speech, and second, that rapid and anonymous speech events could he used as the basis for a systematic study of language. The study of (r) in New York City department stores which I will report here was conducted in November 1962 as a test of these ideas. We can hardly consider the social distribution of language in New York City without encountering the pattern of social stratification which pervades the life of the city. This concept is analyzed in some detail in the major study of the Lower East Side; here we may briefly consider the definition given by Bernard Barber: social stratification is the product of social difft-rrentiation and social evaluation {1957:1—3). The use of this term does not imply any specific type of class or caste, but simply that. the normal workings of society have produced systematic differences between certain institutions or people, and that these differentiated fortns have been ranked in status or prestige by general agreement. We begin with the general hypothesis suggested by exploratory interr’ierm: if any two subgroups of New York City speakers are melted in c scale of social stratification. then they will be ronlted in the some order by their differential use of (1‘). It would be easy to test this hypothesis by comparing occupational groups, which are among the most important indexes of social stratification. We could, for example, take a group of lawyers. a group of file clerks, and a group of janitors. But this would hardly go beyond the indications of the exploratory interviews, and such an The Social Stratification of (r) 45 extreme example of differentiation would not provide a very exact- ing test of the hypothesis. It should be possible to show that the hypothesis is so general, and the differential use of (r) pervades New York City so thoroughly, that fine social differences will be reflected in the indEX as well as gross ones. It therefore seemed best to construct a very severe test by finding a subtle case of stratification within a single occupational group: in this case, the sales people of large department stores in Manhattan. If we select three large department stores, from the top, middle, and bottom of the price and fashion scale, we can expect that the cus— tomers will be socially stratified. Would we expect the sales people to show a comparable stratification? Such a position would depend upon two correlations: between the status ranking of the stores and the ranking of parallel jobs in the three stores; and between the jobs and the behavior of the persons who hold those jobs. These are not unreasonable assumptions. C. Wright Mills points out that salesgit‘ls in large department stores tend to borrow prestige from their cus- tomers. or at least make an effort in that direction.2 It appears that a person's own occupation is more closely correlated with his lin- guistic behavior—for those working actively—than any other single social characteristic. The evidence presented here indicates that the stores are objectively differentiated in a fixed order, and that jobs in these stores are evaluated by employees in that order. Since the product of social differentiation and evaluation, no matter how minor, is social stratification of the employees in the three stores, the hypothesis will predict the following result: salespeople in the highest-ranked store will have the highest values of (1‘); those in the middle-ranked store will have intermediate values of (r); and those in the lowest-ranked store will show the lowest values. If this result holds true, the hypothesis will have received confirmation in pro- portion to the severity of the test. The three stores which were selected are Saks Fifth Avenue, 2. C. Wright Mills, White Collar (New York: Oxford University Press, 1956], p. 173, See also p. 243: "The tendency of white‘collar people to borrow status from higher elements is so strong that it has carried over to all social contacts and features of the workplace. Salespeople in department stores . . . frequently attempt, although often unsuccessfully, to borrow prestige from their contact with customers. and to cash it in among work colleagues as well as friends off the job. in the big city the girl who works on 34th Street cannot successfully claim as much prestige as the one who works on Fifth Avenue or 5?th Street.” 46 socnouncuzsnc PATTERNS Macy’s, and S. Klein. The differential ranking of these stores may 'be illustrated in many ways. Their locations are one important point: i-lighest-ranking: Saks Fifth Avenue at 50th St. and 5th r’\ve., near the center of the high fashion shopping district, along with other high-prestige stores such as Bonwit Teller. Henri Bendel, Lord and Taylor Middle-ranking: Macy’s Herald Square. 34th St. and Sixth Ave, near the garment district, along with Ciinibels and Saks-34th St., other middle—range stores in price and i'n'estige. Lei-vest~i‘;inking: Klein I Union Square, 14th SI. and Broadway, not far from the Lower East Side. The advertising and price policies of the stores are very clearly stratified Perhaps no other element of class behavior is so sharply differentiated in New York City as that of the newspaper which people read; many surveys have shown that the Daily News is the paper read first and foremost by working-class people, while the New York Times draws its readership from the middle class.3 These two newspapers were examined for the advertising copy in October 24—27, 1962: Saks and Macy‘s advertised in the New York Times, where Kleins was represented only by a very small item: in the News, however, Saks does not appear at all, while both Macy's and Kleins are heavy advertisers. No. of pages of advertising October 24—27, 1962 NY Times Dotty News Saks 2 0 Macy’s 6 15 Klein 1/1 10 We may also consider the prices of the goods advertised during those four days. Since Saks usually does not list prices, we can only compare 1.)rit:es for all three stores on one item: women‘s coats, Saks: $90.00, Macy’s: $79.95, Kleins: $23.00. On four items, we can compare Kleins and Macy’s: 3. This statement is fully confirmed by answers to a question on newspaper reader— ship in the Mobilization for Youth Survey of the Lower East Side. The readership of the Daily News and Daily Mirror {now defunct) on the one hand, and the New York “Times and Herald Tribune (now defunct] on the other hand, is almost comple- mentary in distribution by social class. The Social Stratification of (r) 47 Macy’s 5. Klein dresses $14.95 $ 5.00 girls' coats 16.99 12.00 stockings 0.89 0.45 men's suits 49.95—84.95 26.00-66.00 The emphasis on prices is also different. Saks either does not mention prices, or buries the figure in small type at the foot of the page. Macy‘s features the prices in large type, but often adds the slogan, “You get more than low prices.“ Kleins. on the other hand. is often content to let the prices speak for themselves. The form of the prices is also different: Saks gives prices in round figures. such as $120; Macy‘s always shows a few cents off the dollar: 349.95: Kleins usually prices its goods in round numbers, and adds the retail price which is always much higher, and shown in Macy's style: "$23.00. marked down from $49.95." The physical plant of the stores also serves to differentiate them. Saks is the most spacious, especially on the upper floors, with the least amount of goods displayed. Many of the floors are carpeted, and on some of them, a receptionist is stationed to greet the cus- tomers. Kleins, at the other extreme, is a maze of annexes. sloping concrete floors, low ceilings; it has the maximum amount of goods displayed at the least possible expense. The principal stratifying effect upon the employees is the prestige of the store, and the working conditions. Wages do not stratify the employees in the same order. On the contrary, there is every indica- tion that high-prestige stores such as Saks pay lower wages than Macy’s. Saks is a nonunion store, and the general wage structure is not a matter of public record. However, conversations with a number of men and women who have worked in New York department stores, including Saks and Macy’s, show general agreement on the direction of the wage differential.‘l Some of the incidents reflect a 4. Macy's sales employees are represented by a strong labor union, while Saks is not unionized. One former Macy's employee considered it a matter of common knowledge that Saks wages were lower than Macy's. and that the prestige of the store helped to maintain its nonunion position. Bonuses and other increments are said to enter into the picture. It appears that it is more difficult for a young girl to get a job at Saks than at Macy's. Thus Saks has more leeway in hiring policies, and the tendency of the store officials to select girls who speak in a certain way will play a part in the Stratification of language, as well as the adjustment made by the employees to their situation. Both influences converge to produce stratification. as, SOCIOLINGUISTIC PATTERNS willingness of sales people to accept much lower wages from the store with greater prestige. The executives of the prestige stores pay a great deal of attention to employee relations, and take many unusual measures to ensure that the sales people feel that they share in the general prestige of the store-'3 One of the Lower East Side informants Who worked at Saks was chiefly impressed with the fact that she could buy Saks clothes at a 25 percent discount. A similar concession from a lower-prestige store would have been of little interest to her. I From the point of view of Macy’s employees, a job in Kleins is well below the horizon. W’orking conditions and wages are generally considered to be worse, and the prestige of Kleins is very low indeed. As we will see, the ethnic composition of the store employees reflects these dilfercnccs quite accurately. I A socioeconomic index which ranked New Yorkers on occupation would show the employees of the three stores at the same level; an income scale would probably find Macy’s employees somewhat higher than the others; education is the only objective scale which might differentiate the groups in the same order as the prestige of the stores, though there is no evidence on this point. However, the working conditions of sales jobs in the three stores stratify them In the order: Saks. Macy‘s, Kleins; the prestige of the stores leads to a social evaluation of these jobs in the same order. Thus the two aspects of social stratification—diflerentiation and evaluationeare to be seen in the relations of the three stores and their employees. The normal approach to a survey of department store employees requires that one enumerate the sales people of each store, draw random samples in each store. make appointments to speak with each employee at home, interview the respondents, then segregate the native New Yorkers, analyze and resatnple the nonrespondents, and So on. This is an expensive and time-consuming procedure. but for most purposes there is no short cut which will give accurate and reliable results in this case, a simpler method which relies upon the 5. A former Macy's emt'iloycc Iold me of an incidt-znt that occurred shortly before tiltristmas several years ago. As she was shopping in Lord and 'i‘aylor's. she saw the presidenl of the company making the rounds of every aisle and shaking-hands wrlh uverv employee. When she told her fellow employees at Macy‘s about this scene, the most common remark was. "How else do you get someone to work lor that kind ol money?” One can say that not only do the employees of higher—status stores borrow prestige. from their employer—it is also deliberately loaned to them. The Social Stratification of (r) 49 extreme generality of the linguistic behavior of the subjects was used to gather a very limited type of data. This method is dependent upon the systematic sampling of casual and anonymous speech events. Applied in a poorly defined environment, such a method is open to many biases and it would be difficult to say what population-had been studied. In this case, our population is well defined as the sales people {or more generally, any employee whose speech might be heard by a customer) in three specific stores at a specific time. The result will be a view of the role that speech would play in the overall social imprint of the employees upon the customer. It is surprising that this simple and economical approach achieves results with a high degree of consistency and regularity, and allows us to test the original hypothesis in a number of subtle ways. The Method The application of the study of casual and anonymous speech events to the department-store situation was relatively simple. The interviewer approached the informant in the role of a customer asking for directions to a particular department. The department was one which was located on the fourth floor. When the interviewer asked. “Excuse me, where are the women's shoes?” the answer would normally be, “Fourth floor.” The interviewer then leaned forward and said, "Excuse me?” He would usually then obtain another utterance, “Fourth floor,” spoken in careful style under emphatic stress.8 The interviewer would then move along the aisle of the store to a point immediately beyond the informant’s view, and make a writ— ten note of the data. The following independent variables were included: the store floor within the store7 sex age (estimated inlunits of five years) (1'. The interviewer in all cases was myself. i was dressed in middle—class style, with jacket. white shirt and tie, and used my normal pronunciation as a college‘educated native of New jersey (r-pronouncing). 7. Notes were also made on the department in which the employee was located, but the numbers for individual departments are not large enough to allow comparison. '30 SOCIOLtNC-UiSTIC PATTERNS occupation (floorwaikor, sales, cashier, stockboy) I'LICE) ‘ foreign or regional accent, if any The dependent variable is the use of (r) in four occurrences: casual: fourth floor emphatic: fourth lion}; Thus we have preconsonantal and final position, in both casual and emphatic styles of speech. in addition, all other uses of (r) by the informant were noted, from remarks overheard or contained in the interview. For each plainly constricted value of the variable, (r-l] was entered; for unconstricted schwa, lengthened vowel, or no representation, (r—O) was entered. Doubttul cases or partial con- striction were symbolized d and were not used in the final tabulation. Also noted were instances of affricatcs or stops used in the word fourth tor the filial consonant, and any other examples of nonstand- ard tth) variants used by the speaker. ‘ y This method of interviewing was applied in each male on the floor as many times as possible before the spacing of the informants lJECEiI'IIB‘SO close that it was noticed that the same question had been asked before. Each floor of the store was investigated in the same way. On the fourth floor, the form of the question was necessarily different: “Excuse me, what floor is this?" ‘ ‘ 1 Following this method, 68 interviews were obtained in Saks, 125 in Macy‘s, and 71 in Kleins. Total interviewmg time for the 264 subjects was approximately 6.5 hours. I I At this point, We might consider the nature of these 264 interviews in more general terms. They were speech events which had entirely different social significance for the two participants. As far as the informant was concerned, the exchange was a normal salesman- customer interaction, almost below the level of conscrous attention, in which relations of the speakers were so casual and anonymous that they may hardly have been said to have met. This tenuous relationship was the minimuin intrusion upon the behavior of the subject: language and the use of language never appeared at all. From the point of view of the interviewer, the exchange was a systematic elicitation of the exact forms required, in the desued context, the desired order, and with the desired contrast of style. The Social Stratification of (r) 51 Overall Stratification of (r) The results of the study showed clear and consistent stratification of (r) in the three stores. in Fig. 2.1, the use of (r) by employees of Saks, Macy’s and Kleins is compared by means of a bar graph. Since Saks Macy's 32 3] V S. Klein I g A 4 TIIIII; N= 68 125 71 Fig. 2.1. Overall stratification of (r) by store. Shaded area = tit. all (r—tt); unshaded area : at. some (M); "a no {r—l) not shown. N : total number of cases. the data for most informants consist of only [our items, we will not use a continuous numerical index for (r), but rather divide all in- formants into three categories. all (r-1): those whose records show only (r—l} and no (r-U] some (r—l}: those whose records show at least one [r»1) and one tP-O) no (r—l): those whose records show only [r-O} From Fig. 2.1 we see'that a total of 62 percent of Saks employees. 51 percent of Macy's, and 20 percent of Kleins used all or some (r-l}. The stratification is even sharper for the percentages of all (r-l}. As the hypothesis predicted, the groups are ranked by their differential use of (r-l] in the same order as their stratification by extralinguistic factors. Next, we may wish to examine the distribution of (r) in each of the four standard positions. Fig. 2.2 shows this type of display. vvhere once again, the stores are differentiated in the same order, and for each position. There is a considerable difference between Macy‘s and Kleins at each position. but the difference between Macy’s and Saks varies. in emphatic pronunciation of the final (r), Macy's employees come very close to the mark set by Saks. It would seem that r-pro- nunciation is the norm at which a majority of Macy employees aim, 52 SOClOLlNGUISTlC PATTERNS 4th floor 4th floor {it} ()3 {14 [)0 m Fig.2.2 Percentage of all (r-tt] by store for four positions.{S = Saks, M = Macy‘s, K = Kleins.) yet not the one they use most often. In Saks, we see a shift between casual and emphatic pronunciation. but it is much less marked. In other words. Saks employees have more security in a linguistic sense)‘ The fact that the figures for (r-l) at Kleins are low should not obscure the fact that Kleins employees also participate in the same pattern of stylistic: variation of (r) as the other stores. The percentage of r-pronunciation rises at Kleins from 5 to 18 percent as the context becomes more emphatic: a much greater rise in percentage than in the other stores, and a more regular increase as well. It will be important to bear in mind that this attitude—that (r4) is the most appropriate pronunciation for emphatic speech—is shared by at least some speakers in all three stores. Table 2.1 shows the data in detail, with the number of instances obtained for each of the four positions of (r), for each store. It may be noted that the number of occurrences in the second pronunciation of tour is considerably reduced, primarily as a result. of some speaker-3' tendency to answer a second time, "Fourth." t}. The extreme style shifting of the second-highest status group appears throughout the New York City pattern, and is associated with an extreme sensitivity to the norms of an exterior reference group (see Ch. 5 in this volume). In Table 5.1 on p. 133 the data on the index of Linguistic Insecurity is given, which is the number of items in which a speaker distinguishes between his own pronunciation of a word and the correct pronunciation. The second—highest group has the highest scores on this index. We Find parallel phenomena in Shuy, Wolfram. and Riley 1967, Wolfram 1969, and Levine and Crockett 1966, who found in their study of Hillsboro, North Carolina that the second-highest group on the basis of education showed the most extreme stylistic shift ul' (1‘). The Sociai Stratification of {r} 53 I TABLE 2.1. DETAILED DISTRIBUTION OF (r) BY STORE AND WORD POSITION Saks Macy’s 5. Klein Casual Emphatic Casuai Em h ' WT" ' p atic Casuai Em that c (r) 4th floor 4th floor 4th floor 4th floor 4th floor 4thJr flodr (r—1) 1? 31 16 21 33 48 13 31 3 5 6 7 3—0} 39 18 24 12 81 62 48 20 63 59 40 33 4 5 4 4 o 3 1 0 No 1 1 3 3 data" 8 14 24 31 11 12 63 74 22 28 _ _ H __ 4 6 Total no. 53 as as as 125 125 125125 F1- ; 5 3? "The "no data" category for Macy‘s shows relat' ' I I Iverr high values under the em hatic cate o I . r . This discrepancy :5 due to the fact that the procedure for requesting repetitionpwas not Sisal“:- ardrzecl in the investigation of the ground floor at Macy’s, and values for emphatic response were not regularly obtained The effects of this loss a ‘ . re checked 11 plate responses are compared. I Tabla 22‘ Whefe only com— Since the numbers in the fourth position are somewhat smaller than the second. it might be suspected that those who use [r] in Saks and Macy’s tend to give fuller responses, thus giving rise to a spurious impression of increase in (r) values in those positions. We can check this 'point by comparing only those who gave a complete response Theu' responses can be symbolized by a four-digit number. repre: senting the pronunciation in each of the four osit' ' ions (see Table 2.2). p mspecmely Thus we see that the pattern of differential ranking in the use of TABLE 2.2. DISTRIBUTION OF (r1 FOR COMPLETE RESPONSES ‘36 of total responses in tr) Saks Macy’s 5, Kieth All (r-1) 1 1 1 1 24 22 6 Some (r-1} 01 1 1 46 37 12 0 0 1 1 0 1 O 1 etc. No (r-1) 0 o o a 30 41 32 100 E E N = 33 48 34 SOClOLINCUlSTlC PATTERNS 54 sponses, and omission ed in this subgroup of complete re t a factor in this (1‘) is preserv “floor” by some respondents was no of the final pattern. The Effect of Other Independent Variables tratification of the stores, may explain the regular pattern of r—pronunciation seen above. or this effect may be the contribution of a particular group in the population, rather than the behavior of the sales people as a whole. The other inde— pendent variables recorded in the interviews enable us to check such possii'iilitics. Other factors, besides the s anc There are many more black employees in the Kleins sample than in Macy’s, and more in Macy‘s than in Saks. Table 2.3 shows the percentages of black informants and their responses. When We compare these figures with those of Fig. 2.1, for the entire population, it is evident that the presence of many black informants will con- tribute to a lower use of (r-1). The black subjects at Macy’s used less (r-i) than the white informants, though only to a slight extent; the black subjects at Kleins were considerably more biased in the r—less direction. The higher percentage of black sales people in the lower-ranking stores is consistent with the general pattern of social stratification. since in general, black workers have been assigned less desirable iobs. Therefore the contribution of black speakers to the overall pattern is consistent with the hypothesis. TABLE 2.3. DlSTRlBUTlON OF (r) FOR BLACK EMPLOYEES it": of responses in E_ [.r) Saks Macy‘s 5. Klein All {r-'l) 50 '12 0 Some (r-‘lJ 0 35 6 No Lr-'l} so 53 94 '100 Wt] 111) N z 2 1? ‘l8 9:; of black 14 25 informants: 03 The Social Stratification of (r) 55 Occupation t There are other differences in the populations of the stores Th ypes of occupations among the employees who are acce 'bl e customers are quite different. In Macy's, the employees wshSI e to 1ntervilewed could be identified as iloorwalkers (by red a d0 “he're carnations), sales people, cashiers, stockboys. and elevator or1 eri-i‘lolte Ln hS-alés, the cashiers are not accessible to the customer Iivorkifis‘ e in ‘ the sales counters, and stockboys are not seen The w k' 3 operation of the store goes on behindthe scenes, and does n t ‘Or mg upon the customer’s notice. On the other hand, at Kleins Oallncifutlt'ie ptpillptllogegiifzizImcgobtéffieieratipg on the same level: it is difficult is _ H I I I I n as es people, mane era, a f l‘letdebagain, the extralinguistic stratification g1” the IsltcfirsdsoiffZi: czfzuestiigbljegttlfie objefrévatlons in the course of the interview We 1 ‘ ‘ese‘ i erences are not responsible for 't l . part of the stratlfication of (r. For the st ' d eaSt 'a would be desirable to show th)at the stratiflzgfoigfit £033,152]: 12:21:32,111}: . at]. [Ork‘ "lllte Sales “CIneIl' settlng aslde leale opcupations besides selling itself, the black and Puerto Rican em p oyees, and all those with a foreign accent? there are still a t t l of 141 informants to study. 0 a Fig. 2.3 shows the percents ges of [r—l) used b the ' ' galeszworTr‘ifin of the three stores, with the same time oflgihvpfhtilf: 1g. . . e stratification is essentiall ' ' ' y the same in direction Epéljme. thoulgh SOH’letflthat smaller in magnitude. The greatly reduiiicd ns samp e sti s ows by far the lowest f is ahead of Macy‘s in this res use 0 [ll-1L and saks . pect. We can therefor the stratification of r is ' e CUHCIUde that the sample ( ] a process which affects every section of 9. l ' ' and 0121;223:211: asla Xhole, 1? informants With distinct foreign accents were found na 0 aracteristics which were cl 1 ‘ T _ I ear y not of New York C‘ ‘ ’ willjtilfeoifilgn language speakers in Saks had French. or other western Europeahtii mng use in Kleins had Iewish and other eastern European accents Thergcfvli'trS, - E r . . . :EEEEBESIriFEO Kazan employees in the Klems sample, one in Macy's none in Saks Th an an 1.94 women. Men showed the following small differences from worried: in percentages of{r-1] usage: men women all {M} 22 30 some {pl} 22 17 no (r—i) 57 54 ‘36 SOCIOLINGUISTIC PATTERNS Saks Macy’s 3 i S. Klein 7 / 26 /A 4 Illllll N = 49 c5 23" Fig. 2.3. Stratification of {r} by store for native New York white sales women. Shaded area = “it all {r—‘lt; unshaded area 2 % some (r-l): ‘it. no (r-l) not shown. N = total number of cases. We can now turn the heterogeneous nature of the Macy’s sample to advantage. Fig. 2.4 shows the stratification of according to occupational groups in Macy‘s: in line with our initlal hypothesis, this is much sharper than the stratification of the employees in general. The total per-Centage of thosa who use all or some (r-l} 15 almost the same for the floorwalkers and the sales people but a much higher percentage of floor-walkers consistently use [r-l]. Another interesting comparison may he made at Saks, where there is a great discrepancy between the ground floor and the upper floors. The ground floor of Saks looks very much llke Macy s: many crowded counters, salesgirls leaning over the counters. almost elbow to elbow, and a great deal of merchandise displayed. ‘But the upper floors of Saks are far more spacious; there are long tastes of empty carpeting, and on the floors devoted to high fashion, there are models who (.tispiay the individual garments to the customers. Receptiomsts Floorwalkers Sales y Stockbovs #4 N : I05 7 Fig. 2.4. Stratification of [r] by occup:-nienal groups in Mag-3y s. Shaded area : ‘.’r’u all {r-1}; unshadcd area = ‘it. some {M}: it. no {r-‘lJ not shown. :-_ total number of cases. The Social Stratification of (r) 57 are stationed at strategic points to screen out the casual spectators from the serious buyers. It would seem logical then, to compare the ground floor of Saks with the upper floors. By the hypothesis. we should find a differential use of (r-l). Table 2.4 shows that this is the case. TABLE 2.4. DISTRIBUTION OF tr) BY FLOOR IN SAKS (r) Ground floor Upper floors ‘tt'. all (H) 23 34 Li; some [r-'I) 23 40 at, no (r-i) 54 26 100 100 N z 30 38 In the course of the interview, information was also collected on the [th) variable, particularly as it occurred in the word fourth. This is one of the major variables used in the Study of social stratification in New York [Labov 1986a) and elsewhere {Wolfram 1969: Anshen 1969). The most strongly stigmatized variant is the use of the step [t] in fourth, through, think, etc. The percentage of speakers who used stops in this position was fully in accord with the other measures of social stratification which we have seen: Saks 00% Macy’s 04 S. Klein 15 Thus the hypothesis has received a number of semi-independent confirmations. Considering the econorny with which the information was obtained, the survey appears to yield rich results. It is true that we do not know a great deal about the informants that we would like to know: their birthplace, language history. education, partici- pation in New York culture, and so on. Nevertheless, the regularities of the underlying pattern are strong enough to overcome this lack of precision in the selection and identification of informants. Differentiation by Age of the Informants The age of the informants was estimated within fivesyear intervals, and these figures cannot be considered reliable for any but the simplest kind of comparison. However, it should be possible to break 58 SOCIOLINGUISTlC PATTERNS down the age groups into three units, and detect any overall direction of change. If, as we have indicated, {r—l) is one of the chief characteristics of a new prestige pattern which is being superimposed upon the native New York City pattern, we would expect to see a rise in r—pronunciation among the younger sales people. The overall distri- bution by age shows no evidence of change, however in Table 2.5: TABLE 2.5. DISTRIBUTION OF (r) BY ESTIMATED ACE Age groups {r} 'lr-L‘lU 35-50 55-70 all (r-‘IJ 24 20 so It. some [r- I} 2"I 28 22 ‘fi, no (r—‘I} 55 52 58 This lack of direction is surprising, in the light of other evidence that the use of [r—l) as a prestige variant is increasing among younger people in New York City. There is clearcut evidence for the absence of (r-fi.) in New York City in the 1930’s {Kurath and McDavid 1951) and a subsequent increase in the records of Hubbell {1950} and Bronstein (1962). When we examine the distributions for the indi- vidual stores, we find that the even distribution through age levels disappears, Fig. 2.5 shows that the expected inverse correlation with age appears in Saks, but not in Macy‘s or Kleins. Instead, Macy’s shows the re verse direction at a lower level, with older subjects using more (r-‘lj, and Kleins no particular correlation with age. This com- plex pattern is even more puzzling, and one is tempted to dismiss it as the absence of any pattern. But although the numbers of the subgroups may appear to be small, they are larger than many of the subgroups used in the discussions of previous pages, and as we wdl see, it is not possible to discount the results. The conundrum represented by Fig. 2.5 is one of the most sig— nificant results of the procedures that have been followed to this point. Where all other findings confirm the original hypothesis, a singIc result which does not tit the expected pattern may turn our attention in new and profitable directions. From the data in the department store survey alone, it was not possible to account for The Social Stratification of (r) 59 Saks Age: iS-SU : ‘J 38 54 26 S. KIein S. Klein 19 4 VIIIIIJ 35—50 55—?0 "—= 20 26 22 Fig. 2.5. Stratification of {r} by store and age level. Shaded area = “it: all (r-I): unshaded area = “it: some (r-l}: ‘ii. nc [r-1] not shown. N = total number of case . Fig. 2.5 except in speculative terms. In the original report on the department store survey, written shortly after the work was com- pleted, we commented: How can we account for the differences between Saks and Macy's? 1 think we can say this: the shift from the influence of the New England prestige pattern [r-less) to the Midwestern prestige pattern {r—ful) is felt most comn pletely at Saks. The younger people at Saks are under the influence of the r—pronouncing pattern, and the older ones are not. At Macy‘s, there is less sensitivity to the effect among a large number of younger speakers who are completely immersed in the New York City linguistic tradition. The stockboys, the young salesgirls, are not as yet fully aware of the prestige attached to r-pronunciation. On the other hand, the older people at Macy's tend to adopt this pronunciation: very few of them rely upon the older pattern of prestige pronunciation which supports the r-less tendency of older Saks sales people. This is a rather complicated argument, which would certainly have to be tested very thoroughly by longer interviews in both stores before it could be accepted. SOCIOLINCUlSTIC PATTERNS 60 n of Fig. 2.5 offered a considerable challenge anation, but one possibility that always had to be considered was that it was the product of th; marilyfi:pr:rr;1c:3 of error inherent in rapid and anonymous surveys. it: be: necessary explain the results of the department store survey it up I m mm to look ahead to [he results ot the systematic interviewrpgdp ng the discussed in Chs. 3a7. When the results of the maior y2 5 was Lower East Side were analysed, it became clear that 1g. ‘ [.Labw not an artifact of the method but reflected real social patting?) to the 196t‘iaz342 ff). The Lovver East Side data rnost compare ed 1 S department store study are the distribution pf (r) by age’anb pkazf in Style i'l—«the relatively careful speech which is the 1151;? 181;) To the individual interview [see Ch. 3 tor the dC‘llnI-ltlltgilio 16:; lower Saks, Macy’s, Kleins, we can compare upperr not or?“ WhiCh middle class, and working class as a whole. The age tzogzg 30-39 are most comparable to the department store ranges arte u thPré and 40-. (Since the department store estimateslare quite; rt); Fg , 2 6. is would be no gain in trying to match the figures exac y. g. . The complex patter for interpretation and expl 40- [g 1' _ 4 55 Fig. 2.6. Classification of (r) by age and class on the Lower East Side: in style l3, careful speech. then the age and class display for the LOWer Easifiikde $8;;€J(;)g§;t;s$ comparable to Fig. 2.5. Again. we see that the 1g es Bakers use: shows the inverse correlation of [r—l] wrth age: youngetr :gower level, more {r—il); the second-highest status group shows [Lia £1853 mum and the reverse correlation with age; and the wor rnglth a Eg at a still lower level with no particular correlation wrt dges. have. This is a very striking confirmation. since the two 3 u 1 ' ‘ ’ rve quite complementary sources of error. The Lower East Side su y The Social Stratification of (r) 6‘! was a secondary random sample, based on a Mobilization for Youth survey, with complete demographic information on each informant. The interviews were tape-recorded, and a great deal of data on (r) was obtained from each speaker in a wide variety of styles. On the other hand, the department-store study involved a much greater likelihood of error on a number of counts: the small amount of data per informant. the method of notation. the absence of tape recording and reliance on short-term memory, the method of sampling, the estimation of age of the informant, and the lack of background data on the informants. Most of these sources of error are inherent in the method. To compensate for them, we had the uniformity of the interview procedure, the location of the informants in their primary role as employees, the larger number of cases within a single cell, the simplicity of the data, and above all the absence of the biasing effect of the formal linguistic interview. The Lower East Side survey was weak in just those areas where the department-store study was strong, and strong where it was weak. The methodological differ- ences are summed up in the table below. Lower East Side survey Department-store study LES > DS sampling random informants available at specific locations short term memory and notes recording of data tape-recorded demographic data complete minimal: by inspection and inference amount of data large small stylistic range wide narrow US > LES size of sample moderate large location home, alone at work, with others social context interview request for information effect of observation maximal minimal total time per subject {location and interview) 4-8 hours 5 minutes The convergence of the Lower East Side survey and the depart menI—store survey therefore represents the ideal solution to the Observer's Paradox (Ch. 8): that our goal is to observe the way people - use language when they are not being observed. All of our methods involve an approximation to this goal: when we approach from two 62 SOCIOLINGUISTIC PATTERNS different directions. and get the same result, we can feel confident that we have reached past the Observer’s Paradox to the structure that exists independently of the analyst. Given the pattern of Fig. 2.5 as a social fact, how can we explain it? The suggestions advanced in our preliminary note seem to be moving in the right direction, but at that time we had not isolated the liypercorret:t pattern of the lower middle class nor identified the crossover pattern characteristic of change in progress. We must draw more material from the later research to solve this problem. Figs. 2.5 and 2.6 are truncated views of the three-dimensional distribution of the new r-pronouncing norm by age. style. and social class. Fig. 2.7 shows two of the stylistic crOss sections from the more 100 Style A 80 of] 20-29 yrs 40 30-39 yrs '- ____ ____J_____.___——- ClassEt I 20 4049 yrs | 50—?5 yrs I ~.—....-__... .I ___u __s E} u ....._ El' _. _...... D .._.. 1:: ._... __ _ __ 0-1 2-5 as M 2—5 6-8 at 2-5 as (M 2-5 as SEC 0-1 2-5 6-8 0-1 2-5 6-8 0-1 2.5 6‘8 '_ 2.5 6-8 Fig. 2.7. Development of class stratification of {r} for casual speech (Style A] and careful speech (Style B} in apparent time. SEC. —_— socioeconomic class scale. The Social Stratification of (r) 63 detailed study of the Lower East Side population. with four sub- divisions by age. The dotted line shows us how the highest status group [Class 9) introduces the new r-pronouncing norm in casual speech. In Style A only upper~middle-class speakers under 40 show anv sizeable amount of (r-1). None of the younger speakers in the other social groups show any response to this norm in Style A, though some effect can be seen in the middle~aged subjects, espe- cially in the second-highest status group {Class 6—8, lower middle class). In Style B. this imitative effect is exaggerated. with the mid— die-aged lower-middle-ciass group coming very close to the upper- middle-class norm. In more formal styles. not shown here, this subgroup shows an even sharper increase in r-pronunciatio’n. going beyond the upper-middle-class norm in the “hypercorreet pattern that has appeared for this group in other studies (see Ch. 5 in this volume; Levine and Crockett 1966; Shuy, Wolfram, and Riley 1987). Fig. 2.7 is not a case of the reversal of the age distribution of (rd); rather it is a one~generation lag in the peak of response to the new norm. The second-highest status group responds to the new norm with a weaker form of imitation in connected speech. with middle- aged speakers adopting the new norm of the younger high-status speakers; Fig. 2.8 shows this schematically. Our studies do not give the exact profile of the use of (r) among younger upper-middle-class speakers, since we did not focus on that age range. In later observa- tions. 1 have met some upper-middle-class youth who use 100 per- cent (r-l), but in most families, (r-l) is still a superposed pronuncia- tion in adolescence and Fig. 2.8 reflects this. If we wish to express the [rn1) distribution in a single function. we can say that it is in- 0 20 40 Age Fig. 2.8. Hypothetical distribution of (r) as an incoming prestige feature. 54 SOCIOLINGUISTIC PATTERNS versely correlated with distance from the highest-status group (taking Class 9 as 1. Classes 6‘8 as 2, Classes 2—5 as 3, and Classes 0—1 as 4]. it is also directly correlated with the formality of style and the amount of attention paid to speech [taking casual speech, style A, as 0, careful speech, style B. as 1, etc). The slope of style shifting is modified by a function which may be called the “Index of Linguistic Insecurity” (1L1), which is maximized for the second-highest status group (see Table 531, p. 133 for the quantitative index). The age dis- tribution must be shown as greatest for the upper middle class at age 20 and at age 40 for the lower middle class. We can formalize these observations by writing (r-‘I) : —a (Class) + b (Style)[ILl) — c|[Class) - 20 -— (Age)| + d The third term is minimized for the upper middle class at age 20, for the lower middle class at age 40. the working class at age 60, etc. Fig. 2.7 supports this semiquantitative expression of a wave effect, which still has a number of unspecified constants. 'l‘here is a considerable difference between the behavior of the highest-status group and the others. The upper middle class develops the use of (r-‘l) early in life—as a variable expression of relative formality to be found at all stylistic levels. For the other groups in New York City, there is no solid basis for (r-l} in the vernacular style of casual speech; for them, (r—l) is a form which requires some attention paid to speech if it is realized at all. As in so many other formal marks of style—shifting, the lewer middle class overdoes the process of correction. This is a process learned late in life. When speakers who are now 40—50 were growing up, the prestige norm was not [r-l) but (r-O]. Before World War II, the New York City schools were dominated by an Anglophile tradition which taught that (r-I} was a provincial feature, an incorrect inversion of the consonant, and that the correct pronunciation of orthographic r in our was (1&0), [ku'], in accordance with “international English".10 '10. See for example Voice ond Speech Problems, :1 text written for New York City schools in 1940 by Raubicheck, Davis. and Carll {1940:336]: There are many people who feel that an effort should be made to make the pronunciation conform to the spelling, and for some strange reason. they are particularly concerned with r. We all pronounce colm, psol'm, almond, know, night, night, and there without worrying . . . Yet people who would not dream of saying kni: or psai'kuladgi insist on attempting to sound the r in words like pu-k or tuna just because an r marks the spot where our ancestors used a trill The Social Stratification of {r} 65 No adjustment in the pronunciation of this consonant was then necessary for New Yorkers who were trying to use the prestige norm—it was only vowel quality which had to be corrected. This r-less norm can be seen in the formal speech of upper-middle-class speakers, over 40, and lower-middle-class speakers over 50. It also appears in subjective-reaction tests (Ch. 6] for older speakers. The lower-middle-class speakers who now shift to (r-1} in formal styles have abandoned their prestige norm and are responding to the form used by the younger high-status speakers that they come into contact with. On the other hand, many upper-middle—class speakers adhere to their original norm, in defiance of the prevailing trend. The pattern which we have observed in the department-store survey is therefore a reflection of the linguistic insecurity of the lower middle class. which has led the older generation to adopt the most recent norm of (r-1} in preference to the older norm. The process of linguistic socialization is slower for lower-middle-class groups who do not. go to college than for upper-middle-class speakers, who begin adjusting to the new norm in the upper class tracks of the academic high schools. For those who do not follow this path, it takes 10 or 20 years to reach maximum sensitivity to the hierarchical organization of formal language in their community. Some Methodological Directions The most important conclusion of the departmenbstore study is that rapid and anonymous studies can be a valuable source of information on the sociolinguistic structure of a speech community. There are a number of directions in which we can extend and improve such methods. While some sources of error are inherent in the method. others can be eliminated with sufficient attention. . More often than not, people do not really say a third sound in a word like po-k but merely say the vowel o: with the tongue tip curled back toward the throat. This type of vowel production is known as “Inversion.” Letitia Reubicheck was the head of the speech program in the New York City schools for many years and exerted a powerful influenCe on the teaching of English there, The norm of "international English" was maintained by William Tilly of Columbia and followed by Raubicheck and many others in the 1930's and 1940's. As far as i know, this norm has lost entirely its dominant position in the school system: a detailed study of its disappearance from the radio networks and the school system in the 1940‘s would tell us a great deal about the mechanism of such shifts in the prestige form. 66 SOCIOLENGUISTIC PATTERNS ‘1n the department-store survey, the approach to sampling might have been more systematic. it would have been preferable to select every nth sales person, or to use some other method that would avoid the bias of selecting the most available subject in a given area. As long as such a method does not interfere with the unobtrusive character of the speech event, it would reduce sampling bias without decreasing efficiency. Another limitation is that the data were not tape—recorded. The transcriber, myself, knew what the object of the test was, and it is always possible that an unconscious bias in tran— scription would lead to some doubtful cases being recorded as (r-i) in Saks, and as [r-D) in Kleins.11 A third limitation is in the method used to elicit emphatic speech. Fig. 2.2 indicates that the effect of stylistic variation may be slight as compared to the internal pho- nological constraint of preconsonantal vs. final position. The total percentages for all three stores bear this out. [n of all [r—I) for each position Casual Emphatic fourth floor fourth floor 23 39 2.4 48 A simple request for repetition has only a limited effect in inducing more formal speech. The use of reading passages, word lists, and minimal pairs in the Lower East Side study gave a wider range of styles. It might be possible to enlarge the stylistic range in rapid and anonymous studies by emphasizing the difficulty in hearing by one technique or another. The sources of error in the department-store study are offset by the comparability of the three subsections, the size of the sample, - and the availability of the population for rechecking. Though the individual speakers cannot be relocated, the representative popula- tion can easily be reexamined for longitudinal studies of change in progress. There are limitations of such a"psoudopane1” as compared to a true panel study of the same individuals; but the advantages in cost and eiiioiency are overwhelming. 1'1. When the phonetic transcriptions were first made. doubtful cases were marked and were not included in the tabulations made later. There is however room as d the decision between {r-Ul and d and between at and (r—l]. [or interviewer bias in . in”. . .. The Social Stratification of (r) 67 With such promising results in hand, it should be possibl.- to refine and improve the methods used, and apply them in a wider ran e of contexts. In large cities it is reasonable to select single large Institutions like department stores, but there is no reason to limit rapid and anonymous surveys to sales people or to institutions of thischaracter. We can turn to any large body of individuals located at fixed “social addresses” and accessible to interaction with the public: policemen, postal clerks, secretaries, ushers, guides bus dr1vers, taxi drivers, street peddlers and demonstrators beggars construction workers, etc. The public groups which are mdst clearl ’ identified tend to be concentrated towards the lower end of the social]! scale, with sales people at the upper end. But we can reach a more general public by considering shoppers, spectators at sports events parades or construction sites, amateur gardeners, park strollers and passersby in general; here the general character of the residential area can serve the same differentiating function as the three depart- ment stores mentioned above. Many profesaionals of relatively hi h social standing are available for public interaction: particularfi teachers, doctors, and lawyers. Such public events as courtroorii trials and public hearings allow us to monitor the Speech of a wide range of socially located and highly differentiated individuals 12 There is in all such methods a bias towards those populations-that are available to public interaction, and against those which are so located as to insure privacy: business and social leaders, or those engaged in aesthetic, scholarly. scientific, or illegal activities An of 'these groups can be studied with sufficient ingenuity: soc‘iolin}: guistic research should certainly rise to the challenge to develo rapid and anonymous studies that will escape the limitations opt- convenience. But it should be emphasized that since those who are 12. Hearings of the New York City Board of Education were recorded during the study of New York City. and preliminary analysis of the data shows that the pattern of social and stylistic stratification of (r) can easily be recovered from the wide variet of speakers who appear in these hearings. Courtroom proceedings at the New Y k Court of General Sessions are a natural focus for such studies, but speakers of? lower their voices to the point that spectators cannot hear them clearly. Only a smaell beginning has been made on the systematic study of passersby. Plakins {1969} a preached a wide variety of pedestrians in a Connecticut town with requests is; directions to an incomprehensible place. phrased at three levels of politeness She found systematic differences in mode of response according to dress [as an index of seeioeconomic ' ‘ ‘ ' - ” " to [Dome inqmripeti‘sntIon]: and mode of mqutry, there were no rude responses [huh?] 68 SOCIOLiNGUISTlC PATTERNS most available to public interaction may have the most direct effect upon linguistic change and the sociolinguistic system, the bias thrmigh missing the more extreme and obscure ends of the social spectrum is not as great as it may first appear. Since the department-store survey was carried out in Manhattan, several parallel studies have been made. In Suffolk County, Long island, rapid and anonymous observations of the use of (r) were made by Patricia Allen (“1968). In three stratified stores, 156 cm- ployees were observed. In the highest-status store (Macy‘s), only 27 percent of the subjects used no (r-1); in the intermediate store [Grant City}, 40 percent: and in the low-status store (Floyd’s), 60 percent. We see that the general New York City pattern has moved outward from the city. producing a comparable stratification of (r) in three stores of a somewhat narrower range than those studied in Man- hattan. Our own analysis of the New York City situation shows that rapid and anonymous surveys of this kind cannot be interpreted fully without detailed knowledge of the dialect history of the area, and a more systematic study of the distribution of linguistic variables and subjective norms.” in this case, rapid and anonymous surveys should he considered a supplement or preliminary to other methods, no I. substitutions for them. Yet there are cases where rapid methods can give solutions to problems that have never been circumnavigated 'It-i. Allen's tables resemble the New York City patterns but with one major differ- ence; the number of speakers who use all (r-l} is roughly constant in all three stores: 2? percent in Floyd's, 27 percent in Grant City, 32 percent in Macy's. Examination til‘ the distribution in apparent time showed that this phenolnenon was due to the presence of a bimodal split in the lower-store adults {over 3!) years old}. Eighty percent used no {1'—'l] and 20 percent used a consistent all (r-l}: there were none who varied. On the other hand. at} percent of the adults were showing variable [r] in the two other stores. This [Joints to the presence of an older r-pronouncing vernacular which is now dominated by the r-less New York City pattern (Kurath and McDavid 1951), but survives among working-class speakers. The disengagement of such bimodal patterns is a challenging problem [Levine and Crockett 1965], and certainly requires a more systematic survey. Similar complexity is suggested in the results of rapid and anony— mous survey of stores in Austin, 'I‘exas by M. M. Harris [1969). In this basically r—pronouncing area, the prestige norms among whites appear to be a Weak constricted [I'L with a strongly retroflex consonant gaining ground among younger speakers. But for the few blacks and Mexican-Americans encountered, this strong [r] seems to be the norm aimed at in careful articulation. Although these results are only suggestive, they are the kind of preliminary work which is required to orient a more systematic investigation towards the crucial variables of the sociolinguistic structure of that community. ._-..\_.s;. ._......-_.-......4 was...» __ The Social Stratification of (r) 69 by conventional techniques. We have used observations of the speech of telephone operators to construct a national map of the merger of the low back vowels in hook and howl-r, and the merger ofi and e before nasals in pin and pen. In our recent study of the Puerto Rican speech community in New York City, we utilized such natural experimentation to find out what percentage of those heard speaking Spanish on the street were raised in the United States, and what percentage were born in Puerto Rico (Labov and Pedraza 1971} Future studies of language in its social context should rely more heavily on rapid and anonymous studies, as part of a general program of utilizing unobtrusive measures to control the interactive effect of the observer (Webb et at. 1966). But our rapid and anonymous studies are not passive indices of social use. like observations of wear and tear in public places. They represent a form of nonreactive experi- mentation in which we avoid the bias of the experimental context and the irregular interference of prestige norms but still control the behavior of subjects. We are just beginning to study speech events like asking for directions, isolating the invariant rules which govern them, and on this basis develop the ability to control a large body of socially located public speech in a natural setting. We see rapid and anonymous observations as the most important experimental method in a linguistic program which takes as its primary object the language used by ordinary people in their everyday affairs. 5. iii? 33., mon.o_3m:_m:n wwfimwsm <<FE>Z r>m0< C3753? 9“ 3331533 3m? wEEQmHnEm ...
View Full Document

Page1 / 15

anth 4 reader 12 - 42 SOCIOLINCUISTIC PATTERNS be more...

This preview shows document pages 1 - 15. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online