Plenary Speakers

We are honoured to have ALS plenary talks this year from:

Professor Gregory Guy (New York University, USA)

Constraints, community, coherence: Do sociolects exist?

Contemporary sociolinguistic scholarship entertains two, somewhat contradictory, models of the social distribution of linguistic variants.  The speech community model emphasizes coherence – shared norms imply that speakers vary together by style, status, etc., while the identity construction model emphasizes individual agency, and asserts that speakers do bricolage, assembling their social personae and social meanings out of the variables at hand, evoking distinctive indexicalities.  The contrast between these two approaches raises several questions. 

One issue is whether whether clusters of speakers who constitute a community coherently use a set of variants defining a sociolinguistic variety or sociolect, or alternatively, whether individuals use idiosyncratic sets of variants.  For example, given that th-stopping, r-vocalization, and raised /oh/ are all socially stratified in New York City, do higher status individuals systematically use low rates of all three variants at once, or do they pick and choose among these variants? 

Another issue is how internal constraints on variation are treated by individuals and communities.  All linguistic variables have linguistic constraints on their places and rates of occurrence, and specific constraint patterns are often taken as defining characteristics of a dialect or sociolect: for example, the specific contexts for /æ/ tensing in New York and Philadelphia.  Are these community-wide patterns adhered to by individuals who are selecting variants for a momentary social-semiotic effect?  Alternatively, is the manipulation of variables for identity construction or stylistic performance limited to varying rates of use rather than contexts?

This paper presents data from several recent studies of English, Spanish, and Portuguese that address these issues by examining the covariation among multiple sociolinguistic variables – both syntactic and phonological – present in the respective communities, and the constraint effects that individuals and communities display.  

Professor Sabine Stoll (University of Zurich, Switzerland)

Syntactic mixing across generations

Understanding areal convergence presupposes understanding the specific situation of bilingualism involved. As already claimed by Gumperz & Wilson (1971) there is an "urgent need for direct investigation of actual mechanisms of linguistic change in their actual settings" (Gumperz & Wilson, 1971). So far, studies on contact-induced change have been restricted to small-scale language samples often based on elicitation or observation of individual utterances. To study language change in situ however, we need naturalistic and spontaneous conversational data in a wide variety of contexts.

In this presentation I propose that such an investigation is best conducted in a large-scale corpus of natural speech including language samples of a large number of different people of different generations in a variety of contexts. I show how collecting and analyzing big data in an endangered context is possible, taking as an example the development of a large annotated corpus of Chintang (Sino-Tibetan, Nepal) naturalistic speech including several hundred hours of speech with over 100 speakers at different ages, totaling over 1 million words ( All Chintang speakers are bilingual in Nepali, but Chintang is still the main language of daily communication. Changes in code-switching behavior across generations are presented and I propose a number of factors that contribute to these changes. Contradicting expectations from the literature we found that the syntactic integration of multi-word insertions of Nepali into Chintang is as likely as the integration of single-word insertions.


Interested in kindred research in language technology, document computing, and music sciences?

On Wednesday 9 December ALS is co-locating with other conferences. ALS attendees are welcome to attend any of the plenaries or sessions of the other conferences at no charge.


Professor Dr Sonja Kotz (University of Manchester): On the importance of timing and rhythm in motor and non-motor behaviour (Australian Music Psychology Society - AMPS conference)

Professor Joe Wolfe (University of New South Wales): Physics of the voice in speech and singing (Australian Music Psychology Society - AMPS conference)

Professor Mark Johnson (Macquarie University): Computational Linguistics: The Previous and the Next Five Decades (Australasian Language Technology Workshop - ALTA)