Fusing class and network in community-level variationist sociolinguistics Robin Dodsworth North Carolina State University Mary Kohn University of North Carolina Kelly Abrams North Carolina State University NWAV 39 November 6, 2010
Fusing class and network in community-level variationist sociolinguistics
Robin Dodsworth North Carolina State University
Mary Kohn University of North Carolina
Kelly Abrams North Carolina State University
NWAV 39November 6, 2010
Class and network
� Large-scale urban class effects in sociolinguistics are probably (also) network effects.
Class and network
� Large-scale urban class effects in sociolinguistics are probably (also) network effects.
� For example, the high density of working class networks can help preserve vernacular variants (e.g., Johnstone et al. 2006).
Class and network: Milroy & Milroy 1992
Class and network
� In the attempt to explain class effects, it is difficult to separate network effects from “class identity” effects because they usually operate in the same direction.
Class and network
� In the attempt to explain class effects, it is difficult to separate network effects from “class identity” effects because they usually operate in the same direction.
� But what happens when class identity and large-scale network effects represent competing forces on linguistic variation?
RALEIGH
Population: 394,774(City of Raleigh Department of City Planning. Raleigh Population Estimate, July 2010.)
Southern Vowel Shift
(Labov 1991, Labov et al. 2006)
/e/
/ /ɛ
/æ/
/o/
/u/
/ɔ/
/a/
/i/
/ɪ/
� Southern vowels are being reallocated (Trudgill 1986) to the working class in Raleigh as the result of migration from outside the South.
� The migration is motivated partly by a nearby tech industry hub established in 1960.
Metros Ranked by Rate of Population Growth, 1990-2000 (U.S. Census 2000)
Rank Metro Percent Growth1. Las Vegas, NV--AZ 83.33%
2. Naples, FL 65.27%
3. Yuma, AZ 49.70%
4. McAllen--Edinburg--Mission, TX 48.47%
5. Austin--San Marcos, TX 47.69%
6. Fayetteville--Springdale--Rogers, AR 47.52%
7. Boise City, ID 46.14%
8. Phoenix--Mesa, AZ 45.27%
9. Laredo, TX 44.94%
10. Provo--Orem, UT 39.81%
11. Atlanta, GA 38.93%
12. Raleigh--Durham--Chapel Hill, NC 38.85%
70. San Antonio, TX 20.20%
Metros Ranked by Rate of Population Growth, 1990-2000 (U.S. Census 2000)
Rank Metro Percent Growth1. Las Vegas, NV--AZ 83.33%
2. Naples, FL 65.27%
3. Yuma, AZ 49.70%
4. McAllen--Edinburg--Mission, TX 48.47%
5. Austin--San Marcos, TX 47.69%
6. Fayetteville--Springdale--Rogers, AR 47.52%
7. Boise City, ID 46.14%
8. Phoenix--Mesa, AZ 45.27%
9. Laredo, TX 44.94%
10. Provo--Orem, UT 39.81%
11. Atlanta, GA 38.93%
12. Raleigh--Durham--Chapel Hill, NC 38.85%
70. San Antonio, TX 20.20%
Inside/outside the beltline
Inside/outside the beltline: Consciousness
� “We draw a very big distinction in Raleigh, inside and outside the beltline. … If you say ‘Oh, she’s ITB’, that means she’s from inside the beltine.”
� “There’s always the talk of you know inside the beltline and outside the beltline.”
Inside/outside the beltline: (current) Class meaning
� “The implication when people say “I’m from inside the beltline”, is that you’re super wealthy, because you can afford this inside the beltline property. … And probably kinda stuck up.”
� “[In high school] there’s an inside the beltline clique, and then there’s multiple outside the beltline cliques. I feel like class to class there’s always like one crew that’s, everybody’s from inside the beltline, everybody’s parents are loaded. … They associate with each other and that’s it.”
Inside/outside the beltline: Cultural meaning
� “[Old Raleigh people] are gonna be I guess not so showy. … They’ve got this commonality, it’s something quiet. … They tend not to drive too erratically.”
� “North Raleigh’s got connotations. … More of the people I knew [inside the beltline] were from Raleigh, or at least lived there for a period of time, not just newbies coming down from the north.”
� “A lot of the kids [inside the beltline] still have thick southern accents. … I’d say that inside the beltline is kinda holding its own as far as just staying more of a southern city.”
� “You know I don’t think one’s better than the other. …It’s just what you’ve grown up with.”
The perspective from inside the beltline
Inside the Beltline• Raleigh natives• Old money• Southern• Dense networks
Outside the Beltline• Migrants, not Raleigh natives• New money, and wider
economic range• Not culturally southern• Loose networks
The perspective from inside the beltline
Inside the Beltline• Raleigh natives• Old money• Southern• Dense networks
Outside the Beltline• Migrants, not Raleigh natives• New money, and wider
economic range• Not culturally southern• Loose networks
Inside The Beltline female, born 1983:
“That’s probably the number one thing I’m worried about, just all the people moving in.”
� From a class identity perspective, and in view of much previous sociolinguistic work, young ITB speakers are expected to retain southern variants to a greater extent than young white collar OTB speakers.
� From a class identity perspective, and in view of much previous sociolinguistic work, young ITB speakers are expected to retain southern variants to a greater extent than young white collar OTB speakers.
� Alternatively, the large-scale class/network interaction in Raleigh predicts little or no distinction between young white collar ITB and OTB speakers.
A broad class/network perspective
� Homophily: the tendency for similar people to interact with one another more than with dissimilar people (McPherson et al. 2001).
� Occupational homophily emerges consistently in studies of (strong) friendship ties.
� The most homophily has often been found at the highest and lowest socioeconomic groups as defined by occupation (e.g., Verbrugge 1977) and sometimes education (Marsden 1988).
Wright (1997): Occupational homophily in industrialized nations
Occupation category
Odds of tie with unskilled worker relative to employer/worker tie
Employer 1.0
Expert manager 1.3
Professional 1.6
Manager 3.2
Skilled employee 3.4
Supervisor 4.9
Class homophily in Raleigh: Dwyer (2010)
Indicator 1
high unevenness: a lot of affluent people would have to move to create even class dispersion in physical space.
Class homophily in Raleigh: Dwyer (2010)
Indicator 1high unevenness: a lot of affluent people would have to move to create even class dispersion in physical space.
Indicator 2low exposure: affluent people don’t come into contact with low-income people very often.
Class homophily in Raleigh: Dwyer (2010)
Indicator 1high unevenness: a lot of affluent people would have to move to create even class dispersion in physical space.
Indicator 2low exposure: affluent people don’t come into contact with low-income people very often.
Indicator 3high centralization: affluent people live near the center of Raleigh, relative to low-income people.
� So the affluent migrants are geographically both peripheral and central.
� This distinguishes Raleigh from many other urban areas, where the affluent are peripheral.
� So the affluent migrants are geographically both peripheral and central.
� This distinguishes Raleigh from many other urban areas, where the affluent are peripheral.
� This means that in Raleigh:o Some of the migrants live inside the beltline.o ITB people are not likely to interact with working class people.o White collar ITB and OTB people are likely to interact.
Therefore, young affluent people Inside The Beltline are likely to acquire roughly the same (mixed) dialect as their affluent Outside The Beltline peers.
Competing forces
Class identity predicts: young ITB speakers retain southern variants to a greater extent than young white collar OTB speakers.
Competing forces
Class identity predicts: young ITB speakers retain southern variants to a greater extent than young white collar OTB speakers.
Class/network interaction predicts: no significant ITB/OTB distinction.
Data
� Conversational interviews with 43 white collar native Raleighites
� 15-20 pre-consonantal tokens per vowel, per speaker
� F1 and F2 measured at 25%, 50%, 75%
� Lobanov normalization
� Mixed effects models:� Fixed internal effects: following place, manner, voice, duration � Fixed external effects: sex, birth year, generation, ITB/OTB (for
generation 3)� Random effect: speaker
Generation and birth year
Generation 1: born before 1950-precedes tech industry and the first suburban high school-11 speakers
Generation 2: born 1950 - 1980-concurrent with tech industry rise, precedes the second suburban high school-16 speakers
Generation 3: born after 1980-born after tech industry was well established-16 speakers, 8 ITB, 8 OTB
Results: the Southern Shift reverses in apparent time
/e/
/ /ɛ
/æ/
/ɔ/
/ai/
fronting,*** raising***
backing,** lowering*
backing,*** lowering**
fronting*
/i/ and /I/ do not show apparent time change.
glide fronting**, raising*,diphthongizing via fronting**
ITB vs. OTB
In generation 3, does the Inside The Beltline group retain southern vocalic variants to a greater extent than the Outside The Beltline group?
ITB vs. OTB
In generation 3, does the Inside The Beltline group retain southern vocalic variants to a greater extent than the Outside The Beltline group?
Mostly not.
ITB vs. OTB: distinction 1
/æ/ F2 shows a significant ITB/OTB distinction:
ITB speakers have
fronter /æ/(more southern) (p<.05).
No interaction with following manner.
reversal of the /e/ F1 voicing distinction
The voicing distinction reverses over apparent time.
ITB vs. OTB: distinction 2
The /e/ voicing distinction is greater for OTB than ITB (p=.01).
ITB vs. OTB: distinction 2
The /e/ voicing distinction is greater for OTB than ITB (p=.01).
This makes OTB “less southern” insofar as this voicing distinction reverses the generation 1 voicing distinction.
/ai/ F2 glide
The voicing distinction remains significant for all 3 generations.
ITB vs. OTB: distinction 3
The /ai/ voicing distinction is disappearing for OTB via weakening of the pre-voiceless glide (p<.05).
ITB vs. OTB
� In all other respects, generation 3 ITB speakers pattern with their generation 3 OTB peers, not with older generations.
ITB vs. OTB
� In all other respects, generation 3 ITB speakers pattern with their generation 3 OTB peers, not with older generations.
� This suggests that the network (interactional) consequences of class are stronger predictors of linguistic variation than class identity in this case.
Conclusions
� Class-correlated patterns of linguistic variation are largely motivated by the network consequences of economic (dis)similarity.
Conclusions
� Class-correlated patterns of linguistic variation are largely motivated by the network consequences of economic (dis)similarity.
� The network/interactional consequences of economic (dis)similarity are place-specific to some extent (Dwyer 2010).
Conclusions
� Class-correlated patterns of linguistic variation are largely motivated by the network consequences of economic (dis)similarity.
� The network/interactional consequences of economic (dis)similarity are place-specific to some extent (Dwyer 2010).
� Explaining class patterns in sociolinguistic variation entails having information about: � Class homophily and segregation � Class identity
Works CitedDwyer, Rachel (2010) Poverty, prosperity, and place: The shape of class segregation in the
age of extremes. Social Problems 57, 1: 114-137.Johnstone, Barbara, Jennifer Andrus, and Andrew E. Danielson (2006) Mobility,
indexicality, and the engregisterment of “Pittsburghese”. Journal of English Linguistics 34, 2: 77-104.
Labov, William (1991) The three dialects of English. In P. Eckert (ed.), New Ways of Analyzing Sound Change (1-44). New York: Academic Press.
Labov, William, Sharon Ash, and Charles Boberg (2006) The Atlas of North American English: Phonetics, Phonology, and Sound Change. Berlin: Mouton/de Gruyter.
Marsden, P.V. (1988) Homogeneity in confiding relations. Social Networks 10: 57-76.McPherson, Miller, Lynn Smith-Lovin, and James M. Cook (2001) Birds of a feather:
Homophily in social networks. Annual Review of Sociology 27: 415-444.Milroy, Lesley and James Milroy (1992) Social network and social class: Toward an
integrated sociolinguistic model. Language in Society 21, 1: 1-26.Trudgill, Peter (1986) Dialects in Contact. Oxford: Blackwell.Verbrugge, Lois (1977) The structure of adult friendship choices. Social Forces 56, 2: 576-
597.Wright, Erik Olin (1997) Class Counts: Comparative Studies in Class Analysis. Cambridge:
Cambridge University Press.