Ling 320 mcgill

socioling 1

chomskyan view of language

theoretical, concerned w perfect speakers and homogenous behavior (NOT REALITY)

central role of theoretical ling

find descriptive and explanatory theory for language

competence vs performance

ability vs execution. chomsky → general nature of any variation is that average person has errors. what he cares about is the underlying theoretical universal grammar

labovian view of ling

chomsky — theory of opinion of lang, not lang itself. new idea: if we observe ppl speaking empirically, might lead to diff conclusions. ppls intuitions are often wrong (not going to admit ingrammaticality), and every lang shows varuation — none are homogenous. THEORY MUST INCLUDE VARIATION, competence/performance distinction is all LIESSS

overt/covert prestige

talking changes based on style/formality. overt — formal, covert — informal.

communicative competence

who uses which forms in which contexts — innate knowledge. competence is abstract, language is performance

orderly heterogeneity

systematic variation based on social/identity factors (race, sex, class), can only be revealed thru systematic observatuon of vernacular in representative sample. ppl agree is what is standard and lean more toward tat in formal/careful speech

observers paradox

intrusion of observation on behavior (hawthorne) → ppl change behavior when youre looking at them

vernacular

unselfconcious, natural environment speech

micro-socioling

society influeces language (Labov), age/class/gender

macro-socioling

lang influences society

variety of a lang

set of linguistic elements w common distribution

all varieties sum up to a lang

dialect

regional variety (in english, has non-standard connotation). pronunciation, grammar, lexicon

dialect differences

result of lang change and expansion of speech communities. alive langs never stop changing. overtime diffs → seperate langs

lang vs dialect

mutual intelligibility (usually). sentiment (sometimes — mandarin). scalar degrees of understanding.

Assessing status of a lang

standardization, autonomy, vitality, ethnic identity (historocity), reduction, mixture/purity, norms

SAVERMN

standardization

is it codified in graammar/literature/dictionary?

vitality

stability, gain/loss of speakers, domans

historocity

association w ethnic identity/culture

autonomy

varieties, relation to other langs (eg Canadian eng related to American and British)

reduction

of status, resources, social/economic uses

mixture/purity

italian — purely latin descent. english — hybrid germanic/french

de facto norms

popular attitudes, proper usage, good/poor variation

dialect continuum

small movements, small changes → larger distance larger changes, less intelligible

isoglosses / isogloss bundle

lines on a map demarcating where pronunciation / word use changes. geographical bounds (eg mountains). transition zone — ppl use both

bundle — many lines in same territory. dialect boundary

accent

standard dialect w only changes in pronunciation

sociolects

socially divided dialect → eg Queens english, tied to high social class/education, not geog

baltimore recordings

LC is shown w divergence from standard eng, shorter paragraphs grounded in daily life, dropping t/d, aint/double negatives. MC — more abstract, very standard w accent.

contextual variation/speech styles

formal vs informal, everyone has 1 dialect but multiple styles and can shift

produce 1, perceive many. some ppl have more than 1. every speech event occurs in matrix (eg business meeting in mtl french vs halifax eng at home)

vocational registers

communities of practice w specialized vocab eg military

standard form

specific variety w a special status

speech communities

population that shares a variety of some language

Labov: speech communities share subjective norms ab language use/correctness

Milroys: BELFAST. based on social networks, dense & multiplex (WC, everyone knows eachother) vs loose & simplex (MC, you know two ppl they dont know eachother)

Eckert: (jocks vs burnouts) communities of practice, ppl united by shared interests

community model

excludes L2 speakers. did you go to primary/secondary in a place? parents local? Montreal is difficult to implement bounds bc ppl are L2 of either language just by chance basically

nested varieties

eng → NA eng → Can eng → western can eng → BC eng → Van eng → MC → F → young → vernacular

specificity depends on the goal of the study

dialectology

19th cen industrial revolution. ppl didnt need to farm cities. disappearance of dialects

lexiography - dictionaries, lists of words

geography - atlases, maps of regional words/features

More worried about where diffs are than why

sociophonetic dialectology

acoustic phonetic analysis. out vs oot. meeting point btwn physics and ling. vocalic chain shifts (need to retain differentiation so one vowel moves other takes place and chain keeps going

sociolinguist vs dialectologist

socio → transition zones. why is there variation, social profile of who displays certain trait

dialect → simply finding where changed are or aren’t. lines, maps, transition zones/

acoustic phonetic analysis quick summary

f1 = height

f2 = backness

use formants to compare dialects

standard transcript = accent, nonstandard = dialect

northern cities shift is opposite of southern shift

vowel chart

can specificy w acoustic data (precise measurements) or vague placement

social and stylistic variation

indicators — intergroup, below level of awareness, not style

markers — intra/intergroup style variation, subconscious

stereotypes — extreme social marking, social/conscious awareness, exaggerated/obsolete, known outside community

linguistic variables

formal construct of theory, quantitative analysis. not semantic variation but different in social meaning/distribution (sociosymbolic value)

when dialectology → socioling shift in 1960s, reintroduce theoretical ling to socioling

speakers grammars are variable, not categorical

sociosymbolic value

not semantic variation but different in social meaning/distribution

variant

particular word (eg ing, in’)

variable

the thing that changes (eg ing)

frequency of variables conditioned by

social/external (age sex class), stylistic, linguistics/internal (phonetics, syntax)

notation

(variables) [variants]

(s) → [s] or [0]

variable classification

phonetic, phonological, morphological, syntactic, discourse, prosodic, semantic, lexical

discrete variation

clear bounds eg words/phonemes

continuous variation

scalar bounds eg phonetic quality

indep variables in social variation

age (groups, life stages, scale)

sex (binary?)

class(occupation, income, edu, residence)

urban-rural (scalar? based on pop)

ethnicity(categorical?)

social networks/communities of practice (locally defined)

conceptual design

the “what”. speech community, the geographic setting.

dependent variable

linguistic variable of study → define the variants and how they differ

independent variable

internal (ling context) and nonling variables (social/style) expected to influence dependent variable

sampling

who. random → every mem of pop eq chance (expensive, lots of time). judgement: approx equal # of ppl from each grp of interest (most common)

usually more = better but there is a limit

elicitation

how. central challenge is to overcome observers paradox.

rapid and anonymous observation (eg dept stores)

brief interactions w general public. unaware participants, large quantity, can tell generally age sex social class native speaker

limits: small data from each participant, no context, no recording

sociolinguistic interview (labov loved ts) norwich,

awareness and sensitivity manipulated in face to face interviews. range of topics, creates social portrait of person. RECORDED! more and richer data,

limits: small sample size, time consuming, still observers paradoc

Labov protocol for socioling interviews

D’ - minimal pairs (have phonemic contrast of study)

D - word list w/ variables of interest

C - reading style, read story w dependent variables

B - careful speech, direct responses to interview qs, opinions/hobbies/work

A - casual speech, emotional narrative (near death experience?)

participant observation

ethnography basically. stay in community long term, daily interactions, slow but lotsss of data. eg eckert in detroit (jocks/burnouts)

coding in data analysis

establish value for each instance of a specific variable

counting in data analysis

proportion of each variant of variable for each person / group in the sample

measurement

establish value of scalar measure for each instance of dependent variable (can be impressionistic or acoustic measurements)

quantitative analysis

calculating means and standard deviance of scalar measures for each group in sample

hypothesis testing

statistical tests to determine whether grp diffs in proportion are statistically significant. most social diffs vary w/in groups (overlapping distribution).

statistical significance

depends on

N → size of sample (more data = more reliable that a diff is significant)

size of difference → larger diff more meaningful

intra- and inter-group variation → w/in group relative to btwn-group. if groups vary a lot within themselves and the means are different might not mean much

statistical tests

chi-square: significance of 2 proportions, given the size of the difference and N

t-test: btwn means, given size of diff, amt of variation, N

pearson correlation coefficient: significance of correlation btwn 2 scalar values, eg participant age & formants

multivariate analysis: simulataneously determine significance of individual and joint effects (covariation) of more than 1 indep variable on 1 or more dep variables (USE R)

Fisher

1958. socioling first publication. social influences on choice of ling variant. free variation - label, not explanation. SMALL NEW ENG TOWN

Children: 3 tasks. stories ab picture, questionnaire, informal interviews.

findings: more boys used in’ more often than girls. sex has effect on use — women and girls are more standard

overt/covert prestige: split into typical vs model boys

SEC — no diff but small town w/o much variations

style: more ‘in as formality decreases (also, more formal/latin rooted words have full ing)

Labov 1966, 1972 — social stratification of english in NYC

most influential sociolinguistic study.

rapid and anonymous in 3 dept stores

• accomodation theory: workers will speak like the target customers (store choice is proxy for social class)

• mostly locals

• prestige was moving away from r-less speech

• ask where a dept that was on fourth floor was, repeat as if didnt hear

  • 4 conditions: pre-consonantal vs prepausal, stylistic (first response, less careful → more careful second response)

• Results: presence of r is correlated with high-class stores, higher floors, higher status jobs, more present word-finally, and in 2nd response

  • weirdness: expected young ppl @ UMC to use most, but really middle-ages at LMC

sociolinguistic interiews on the LES

• 5 varibales: r, bath/cloth fronting, th → dh

• 70 residence, 0-9 social class, all european-american

• orderly heterogeneuity

• hypercorrection in upper middle class, more correct than upper class (overt prestige)

• interior groups more style switching (insecurity of status??)

• Older ppl care more ab social prestige → follow younger UMC speakers

Trudgill

1974. Norwich, England, Labovian interviews.ing, glottalizing t, h-dropping. 5 SEC groups in 4 styles. confirmed labov’s orderly heterogeneity, same patterns and correlations → men are more non-standard, women are less secure and drawn towards prestige

Cheshire

1978, Anthrpologist, participant observation. Children in 3 playground groups. Reading, Eng. extending 3rd sing -s to non 3sg persons/numbers. usage attachd to vernacular

index of vernacular culture (covert prestige, peer grp status, toughness, ambition). boys w higher IVC used -s more. less IVC overall in girls, decrease usage in formal styles more than boys.

IVC indicates stable variation, non-standard compedte w standard and correlate to factors.

Wolfram 1969

AAE in Detroit, comparison with UMC whites. [z] absence in 3sg, higher in LWC. [r] less judged

AAE seperate dialect, shows similar stratification and style variaition

gramatical variants have sharp stratification, binary btwn WC and MC

Sankoff and Cedergren 1971, and Vincent 1977

MTL French → L deletion

Jahangiri 1980

vowel assimilation in tehran persian, variation correlated w sex and class exists outside of western societies. ranges for adjacent social groups overlap

Milroys 1978

3WC sectors of belfast (1 cath 2 protestant)

participant observation, friend of friend method to infiltrate existing social networks

structure below SEC: dense multiplex social networks are norm enforcement mechanisms

lowering of ah and deletn of intervocalic /d/

Labov 1972b

AAVE deletes fas(t) care but also wil(d) elephant. variable rule of deletion.

historical linguistics

beowulf → canterbury tales → king james bible. english has changed a whole bunch over time. inspired by philology and observations of changes. morpho and lexical similarieties that couldn’t have been made up. realized we could make systematic rules.

grimm’s law

stops to fricatives. first germanic consonant shift.

second germanic consonant shift

final voiceless stops → voiceless fricatives

tree model hist ling

represent genetic family tree with branches. not completely accurate bc they still interact w eachother after seperation

wave model hist ling

bundles of isoglosses, can see changes spread over time

neogrammarian theory

regular sound changes wo exceptions

issue with neogrammarian theory

some villages in netherlands mus → mys but hus → hus; no phonetic explanation

LABOV → words develop independently, not the whole variable at a time

socioling approach to language change

language change is observable, not all variation has change but all change has variation. no change is sudden and uniform. early and late adopters in every community, competition btwn new and old forms for some time

diachronic trajectory

change in generational increments (changes already happened)

synchonic manifestation of lang change

age variation (abstract, variation is irrelevant)

how to observe lang change

systematic study of language in its social context; compare modern to past

Apparent time construct

synchronic generational differences, bc ppls lang is preseerved after critical period (teens). older speakers represent lang as it was whe they were a child.

can only be confirmed by present-time data

age grading

language can change as adults, ppl sound older as they grow (oppostie of apparent time hypothesis)

Labov 1963

Martha’s vineyard, centralization of dipthongs ai and au. similar to canadian raising. mix of ethnic grps and occupations. reversal in younger gens. mostly middle aged fishermen. revealed SOCIAL MEANING — ppl w more attachment to the island community raised more, younger gens that wanted to leave hadn’t left yet.

real time

actually measuring changes over time.

real/apparent time combo

measuring apical r in mtl french, interviewing multiple generations and then re-interviewing. some adults adopt later

fashion changes / change from above

overt prestige, following UMC speakers, ppl imitate

stereotypical social mechanism

laziness/ignorance of LC speakers

change from below

local identity, covert prestige. ppl diverge from standard to make themselves unique and more united. led by interior groups

interaction of changes from below/above

ppl realize changing from below and “correct” them, happened often durring WWI/II.

eckert 1989

change from below in detroit hs, jocks/burnouts. backing (new process) led by burnouts, raising (older) led by men. the jock/burnout distinction mattered more for girls than boys, soft power, status bound > status conscious

Labov 1990

intersection of sex and social class in philly. changes from below were led by women in interior classes, unless the change had a ‘toughness’ association.

English borrowing

used to be from other langs (latib, low german, fench, etc). now borrowing from eng → geman french, youthful and cooler, mophologized into lang

calque

borrowing of concept, translating loanword

academie francaise is big fan of this

sprachbund

langs that are not necessarily related but share similarities because they coexist closely