Lecture: Ruth Maddeaux

Individual cognitive differences as predictors of participation in sound change

Sociolinguistic research is built on analyzing variation among speakers grouped into macro-social categories – e.g., age, gender, socio-economic background (Labov 2001 and many others) – with relatively little attention paid to the individual speaker. But, models of sound change rely on the individual listener/speaker to initiate, adopt and propel change (Ohala 1981; Janda & Joseph 2003). And, psycholinguistic research finds robust correlations between aspects of the individual's cognitive function and linguistic performance. This leads to a question that brings these three facts together: What individual differences in cognitive function play a role in perceiving and producing sociolinguistically evaluated sound change?
As a starting point for investigating the role of individual differences within a group, I employ two cognitive measures: Empathy Quotient (EQ) quantifies our ability to identify another person’s emotions and to respond appropriately, and Systematizing Quotient (SQ) is our ability to construct and analyze rule-based systems (Baron-Cohen 2009). In perceptual tasks, individuals with either low EQ or low SQ scores compensate less for vocalic context when categorizing ambiguous segments (Yu 2013). I examine whether these results extend to production by testing correlations between these measures and the F2 of /u/ in the speech of Toronto English speakers. The vowel /u/ is fronting in apparent time in Toronto (Boberg 2010, 2011). It is at the onset of a change, is highly variable among speakers, and is phonetically conditioned, making it a promising locus for individual differences to emerge.
Statistical analysis reveals that neither EQ nor SQ alone predict a speaker’s degree of frontedness. But, speakers who have a high drive towards both empathy and systematizing are significantly more fronted than any other speakers, most dramatically in the environment that has long favoured fronting (post-coronally). I argue that the cognitive differences are the most apparent in environments in which fronting is occurring as a phonetic, coarticulatory process. In other environments (e.g., post-labially), where fronting is not co-articulatorily motivated and in fact is happening as a result of the loss of the conditioning environment (cf. Janda & Joseph 2003), cognitive effects play a smaller role. Overall, results indicate that, alongside macrosocial group membership and identity, we need also look to individuals’ cognitive profile as predictors of their participation in language change.

Ruth Maddeaux is a PhD candidate in the Department of Linguistics at the University of Toronto. She uses both traditional variationist and experimental methods to investigate sociolinguistic variation and change. Her previous research has included inductive learning of Modern Irish consonant mutation, the link between language processing and visual attention, the nature of the sociolinguistic variable, contact effects in Heritage Polish, and stylistic variation in English pronominal use. Ruth’s Master’s degree (2014) and Bachelor’s degree (2013) were all completed in linguistics at UofT.