• Last week we looked at social factors that
apply to entire regional groups: isolation
and contact. • While it is traditional for speech
communities to be identified regionally,
they can in fact be identified according to
different criteria. For example:
– social networks
– ethnicity • When these criteria are taken into
consideration, every regional community
can be seen as being heterogeneous,
consisting of multiple sub-varieties. • “…where there are social distinctions, we
can expect to find them reflected in
linguistic distinctions.” (p.457) • The study of the connection between social
differences and linguistic differences is
known as variationist sociolinguistics.
• An important methodological tool in this
work is the identification of sociolinguistic
variables: alternative ways of saying the
same thing. • e.g. brother, brudder, bruvver, bro’er
• Each possible realization is a variant of the
variable (‘th’) • Relation between variables and variants is
analagous to relation between phonemes
and allophones. • Just as phonologists study the contexts and
rules that determine the choice of allophone,
variationists study the factors that underlie
the choice of variants. These factors include
e.g. position of ‘th’ in a word
e.g. speaker’s age • Unlike phonological rules, which are
usually assumed to be categorical, the
factors governing variants are probabilistic
(i.e. more or less likely to apply) • Variation can involve features at all levels
– lexicals Phonetic variation
• In some New York City dialects, alveolar
consonants are produced with contact
between the tongue tip and the upper teeth
(i.e. they are dento-alveolar), while in
standard dialects, the alveolars are not
dental. (Gerfen 2002)
• This is phonetic variation: at the phonemic
level, there is really no difference between NY
English and standard English. Both have the exact
same set of alveolar consonant PHONEMES. Phonological variation
• In some dialects of American English there
is a difference between the vowel in the
word "caught" and the vowel in the word
"cot”: these are a minimal pair. In other
dialects difference has been neutralized, i.e.
lost. (Gerfen 2002)
• This is phonemic variation: these dialects
differ in terms of their underlying phonemic
inventory. Morphological variation
• In northern England and Southern Wales, the -s
suffix is used as a general present tense marker. In
many other dialects of English, -s is reserved for
marking the present tense in third person singular
forms only. (Gerfen 2002)
– I likes him.
– We walks all the time.
• Appalachian English has a number of past tense
forms that are non-standard. "Et" for "ate", "hEt"
for "heated". These are all examples of
morphological variation. (Gerfen 2002) Syntactic variation
• In the U.S., some southern dialects use the
word "done" used as (perfect) auxiliary, as
in "she done already told you" or "I done
finished a while ago." In standard American
English, this isn't the case. (Gerfen 2002)
• Double modals (combinations of
auxiliaries) are also common across parts of
the South: "I might could do it" or "They
useta could do it" or "He might would if you
asked him nice enough." (Gerfen 2002) Lexical variation
• Lexical items can differ in their meanings
• In some cases, different lexical items are used to express
the same meaning (e.g. couch vs. sofa vs. chesterfield)
• In other cases, the same lexical items (or expressions) can
have different meanings (e.g. dresser refers to a kitchen
cupboard in British English, but to bedroom furniture for
storing clothes in North American English; to knock
somebody up means to get them pregnant in North
American English, but to wake them up in British
English). Social network
• A general factor that has been used to
explain patterns of variation: the cluster of
social relationships an individual has
(family, friends, neighbourhood,
employment, religious affiliations, etc.).
• The study of these clusters is called social
network analysis. www.notesolution.com • The basic idea: the more often you talk to a
certain group, the more likely you are to
share the same speech patterns.
• Speakers are given individual network
scores based on the kinds of ties they have
and the density of the networks they’re
involved in. • A dense network is one in which there are a
a large number of interconnected speakers.
• This is in contrast with multiplex networks,
which are those where relationships are
formed with several clusters or kinds of
relationships (so there is less
interconnectedness in the group). • Dense social networks enforce community
• Speakers most closely integrated into a
dense network show the highest rates of use
of the variants associated with that network. • Speakers least integrated into the network
are the most likely to introduce new forms.
• Linguistic behaviour in dense networks is
constrained in much the same way that
other kinds of behaviour are constrained in
closely-knit communities. Gender
• Gender is one of the most widely reported
factors in linguistic variation.
• Both cross-linguistically and cross-
culturally, the speech of men and women
tends to differ. • In some cases these differences are subtle and
speakers may be unaware of them (e.g English),
but research shows differences in, for example,
rates of usage of certain terms (e.g. women>colour
terms, men>sports terms) and also variable usage
of certain grammatical forms.
• Such differences involve rates of use of shared
forms, rather than differences in the forms
themselves. • In other cases the differences are
• Gros Ventre (Montana): women pronounce some
words differently from men.
• Koasati (Louisiana): verb forms differ depending on
sex of speaker (table slide) www.notesolution.com • Japanese • In speech communities where gender differences
are strongly encoded in the language, one usually
encounters gender-exclusivity: strong social
prohibitions against using forms associated with
the opposite sex. • Some frequently made observations:
– (1) When variation is stable (no change in
progress), men use more non-standard forms
than women. (And they also claim to use even
more of them than they actually do).
• covert prestige (e.g. ‘street cred’) – (2) When there is a change in progress, women
use more of the novel forms than men. This is
especially true with change ‘from above’.
• linguistic insecurity?
• differences may simply reflect different gender
expectations regarding toughness, etc. Also, women
more often do jobs that require standard speech (e.g.
teachers). • Women have been observed to use more
prestige variants than men. There are a
number theories as to why.
– theory 1: the difference results from differences
in social network density.
– theory 2: attributes this to women’s role as
child caregivers. Small children are more likely
to hear innovative forms from women than
from men, so the change happens because it is
transmitted to the novel generation. • Preconceptions about gender roles have
often coloured claims that are made about
language and gender.
• For example, it has been claimed in the past
that women are more likely than men to use
verbal hedges (e.g. perhaps, you know, sort
• This was claimed to reflect unassertiveness
of women • But a study of you know showed that it has
more than one use, and these are
differentiated by intonation patterns.
– rising intonation you knów: uncertainty (11a)
– falling intonation you knòw: confidence (11b)
• The study found that women are more likely
than men to use you know to express
confidence. • The relationship between language and
gender is complex.
• Some consider gender to be an independent
dimension along which speech varieties can
• Others consider gender differences to
simply be a special case of the more general
sensitivity to social power. Class
• Speech varieties that are strongly associated
with a particular socioeconomic class
provide a valuable window on both
variation and social structure (and mobility). • In many communities, certain variants are
more closely associated with upper classes,
and these carry prestige.
– e.g. these things vs. dem tings • Upper classes, in turn, often adopt norms
associated with prestige dialects from
outside the community, e.g. features of
British English in Canadian English, or
European French in Quebecois French. • It isn’t straightforward to assign individuals
to a particular class. It certainly isn’t as
simple as knowing a person’s income.
• Other factors:
– education, housing, occupational prestige,
family background, etc.
– occupational prestige has been found to be one
of the most significant indicators of class • Labov’s famous fourth floor study: