Sociolinguistic Variables

24.901 Sociolinguistics and Sound Change 0. Gender differences in spoken language Japanese • Lexical items: boku (masc. first person) vs. watakushi (neutral) • Grammatical particles: da yo (masc. blunt) wa yo (fem. Soft) • Voice quality: breathy voice (fem.) Sapir (Male and female speech forms in Yana 1928) • Most words have two forms: male and female • Female derived by reduction (devoicing) of final syllable • Male used when male addresses another male, female otherwise Vaupe_s region of Amazon (Jackson 1983) • Eastern Tukanoan • Linguistic exogamy: man must take wife who speaks a different language • Individual’s identity marked by language of father • When woman moves to husband’s community she continues to use her father’s language while husband (and children) use father’s language • Extensive multilingualism: Barasana, Desano, Tatuyo, …. 1. Quantitative sociolinguistics (Labov) • major finding: ordered heterogeneity: a particular grammatical variable (e.g. rhotic vs. nonrhotic) increases/decreases monotonically along some social dimension • variation by style (formal---casual conversation) • variation by age (may reflect sound change in progress) • variation by social class and gender • variation: stable vs. changing • implication: Structural-Generative abstraction of ideal speaker in homogeneous speech community overlooks an important aspect of linguistic competence: people alter their grammatical output in response to different social contexts and to signal their identity or affiliation to particular social groups 2. William Labov (b. 1927, BA Harvard '48, Ph.D. Columbia '64) 1 • • • • • • founder and chief exponent of quantitative sociolinguistics student of Uriel Weinreich (Columbia, U Penn since 1970) large scale quantitative analysis of the vernacular speech: NYC, Philadelphia, other American dialects editor Language Variation and Change director of Dialect Survey of North America (2005) author of Principles of Linguistic Change I, II, III 3. Sociolinguistic interview; stylistic variants • Minimal pair • Word list • Reading style • Careful speech • Casual (non-monitored) speech Figure 4.2 • • • • • NYC nonrhotic Rhotic pronunciation restored as prestige norm after WWII (change from above) 81 adults of Lower East Side (1962) systematic correlation with attention to speech and social class lower middle class increases in more formal setting: linguistic insecurity 4. Labov’s Department Store Study (1966) fig. 4.3, 4.4 • 264 individuals encountered in two afternoons in 1964; anonymous • Saks, Macy’s, S. Klein • Floorwalker, sales clerk, sweeper • Item from 4th Fl. WL: Where’s X(e.g. shoes)? [fourth floor] Excuse me. [Fourth Floor] Results • Stratification by store (social class): 4.3 • Stratification by phonological context (preconsonantal vs. word-final) • Stratification by style (emphatic speech more rhotic) • Stratification by occupation (2.4 in p. 56) • Stratification by age 5. Study replicated by Fowler (1986) • Same results but overall more rhotic (sound change in progress) • The next generation increments the change • Alternative explanation of age grading is not plausible • Age grading: speakers become more conservative as they get older • Philadelphia [aw] fronting and raising also shows this [3.7, 3.6] • Observations of different generations in same family: 4.9 a,b: raising of æ vowel • Recordings of one speaker at two separated points: Jenny Rosetti (Philadelphia) at ages 68 (1973) and 85 (1990): vowel system largely stable 2 • But Queen’s Annual Christmas Message: 1950’s to present (Harrington (2006): some effect of changes in RP (raising of lax I in kit, happy) 6. Classification of sound changes: a. shift: p -> h in Japanese; V -> V: / ____ [+voiced] b. mergers: a and o merge (father and hot); t & d merge to flap, etc. Herzog’s Principle: at a dialect boundary mergers expand: whale vs. wail Garde: mergers are irreversible by normal sound change (from below) A sound change from above may lead to hypercorrection: generalization of change to to novel nonetymological lexical items intrusive [r] of Eastern Mass is commonly interpreted as reflex of hypercorrection: the spa[r] is (spa or spar) Hock (1986:428) border soda opera Missoura Southern sody opry Missouri c. split: umlaut in German (original phonetic context lost) d. chain shifts: one shift elicits another related one drag chain: one sound moves into the space formerly occupied by another: English Great Vowel Shift where [i:] diphthongizes and centralizes to [aj], [e:] shifts to former position of [i:] push chain: one sound encroaches on another, which shifts to remain distinct (more difficult to document) 7. Sound (more generally linguistic) change • some variation is stable (e.g. t/d deletion of wes(t) documented since 16th century) • while other kinds are dynamic with the gradual replacement of one category by another over time as in nonrhotic > rhotic. • mystery: how does the child discern which variation is stable and which is to be carried forward (assuming that children are the primary vehicles of change from below)? 8. Two kinds of sound change: • changes from above: introduced by dominant social class, often with full public awareness, show up in careful speech: NYC rhoticization • changes from below: initially below level of consciousness, typically lower or middle classes initiate: Northern Cities Shift • NCS: [æ] raises and tenses, [a] of cot fronts, [_] lowers and unrounds; [_] backs encroaching on [_] which backs and rounds • more detailed studies like Eckert's show group identification rather than social class per se is better predictor (cf. solidarity of values, identification (Sturtevant 1947) 8. Belten High (Eckert) 3 • • • Jocks: middle class values, achieve goals by conforming to normative institutions established by adults Burnouts: resist adult authority, achieve goals by escape from adult control, rely on local resources For a number of sound changes in the NCS burnouts are in lead: eg. Extreme backing of wedge (bus -> boss): Jocks .43 • Burnouts .59 but social class gives way to gender: 13.7 9. Labov sees gender as most important social determinant in sound change Finding 1: For stable variables women show a lower rate of stigmatized variants and a higher rate of prestige variants • -ing ≈ -in; former used by women more in New England, NYC, Norwich, Australia, Ottawa, … • stop and affricate forms of English interdentals used less by women: NYC, North Carolina, Belfast • Quebecois Fr, Chilean Sp., Taiwanese studies report similar findings Finding 2: in changes from above women adopt the prestige form more than men • rhotic NYC, QF retraction of /a/ in –ation Finding 3: For changes from below (the proto-typical, unconscious sound change) woman are typically in the lead • • • Gauchet (1905) Charmey (Sw Fr) women one generation ahead of men in diphthongization of a: and e: NYC raising of ah and oh NCS early changes of æh, o, and oh (Wolfram, Eckert) Speculation: • young women and female adolescents more fashion/style-conscious than males • extending a linguistic variable is comparable to other fashion statements • they spend more time with young children, who set their variables accordingly ------- 4 Social Class 6-8: Lower middle 9: Upper middle (20-29 yrs) 100 9: Upper middle 90 3-5: Working class % constricted [r] 80 0-2: Lower class 70 60 50 40 30 20 10 0 Casual speech Careful speech Reading style Word lists Minimal pairs Figure 4.2 Social stratification of (r) in New York City Image by MIT OpenCourseWare. 1962 80 60 0 < [r] < 100% [r] 100% 40 20 0 % using constricted [r] % using constricted [r] 80 60 40 20 0 Saks 1962 Macy's 1962 S. Klein 1962 Store Figure 4.3a Overall stratification of (r), 1962 1986 Saks 1986 Macy's 1986 May's 1986 Store Figure 4.3b Overall stratification of (r), 1986 Image by MIT OpenCourseWare. 1962 % using all [r] 80 60 40 20 0 fourth floor FOURTH FLOOR Saks % using all [r] Macys S. Klein 1986 80 60 40 20 0 fourth floor FOURTH FLOOR FIGURE 4.4: Differentiation of (r) by style in 1962 and 1986 Image by MIT OpenCourseWare. 5 Floorwalkers Sales 8 35 46 Stockboys 18 N= 13 14 105 7 Fig. 2.4 Stratification of (r) by occupational groups in Macy's. Shaded area = % all (r-1); unshaded area = % some (r-1); % no (r-1) not shown. N = total number of cases. Image by MIT OpenCourseWare. 2400 Group means Regression on age Multiple regression 2300 2200 2100 F2 2000 1900 1800 1700 1600 aw 8-14 15-19 20-29 30-39 40-49 50-59 60-69 70-79 Age Figure 3.7 Distribution of F2 (aw) in apparent time by group means, simple regression and multiple regression Image by MIT OpenCourseWare. 7 F2 400 2600 2400 2200 2000 1800 1600 iyC 450 oiyF 500 eyC oeyF oh owC owF e 700 o aw P < .01 ahr ^ ay0 750 800 ohr oyo u F1 650 P < .10 1000 uwC uwF hN 600 850 1200 i 550 800 1400 P < .05 P < .001 o ayVo 900 Figure 3.6 Movement of Philadelphia vowels in apparent time. O = mean values for 116 speakers in the neighborhood study. Vectors connect values for groups 25 years older and younger than the mean. ___F = free vowel; ___C = checked vowels; ___0 = before voiceless finals Image by MIT OpenCourseWare. F2 300 2400 2200 2000 1800 1600 hN 400 500 Other h 1000 800 600 O uw i ey ow nd nd d ng oy fl g fl g 700 l s l st ch l stf nd l kl s f ns h d 900 1200 iy F1 600 800 1400 ln- lnk t oh e k bl k k o k ch b ah bl k t 1000 Figure 4.9a Vowel system of James Adamo, 55, Detroit Image by MIT OpenCourseWare. 8 F2 300 2000 1800 1600 1400 hN iy 400 1200 Other h 1000 800 O n ey uw nd ndz n s i 500 d F1 h th 600 g g k e d r d pt ow th gr bd st sht 600 t sk t kt 700 t t oh o d pt kt 800 Figure 4.9b Vowel system of Chris Adamo, 13, Detroit Image by MIT OpenCourseWare. Male Jocks Female Jocks Percent advanced tokens 70 Male Burnouts Female Burnouts 60 50 40 30 20 10 0 Q o oh e ^ Figure 13.7 Relations of gender and social class in the development of the Northern Cities Shift in Belten High (from Eckert 1989b) Image by MIT OpenCourseWare. 9