Speech Science Forum - Teasing apart social and cognitive factors in sociolinguistic perception

Speaker: Dr Erez Levon, Queen Mary University of London

That stereotypes affect sociolinguistic processing is by now uncontroversial. Research over the past fifteen years has demonstrated that listeners are attuned to a variety of socio-indexical cues both within and external to the speech signal, and that these cues influence how linguistic features are perceived and evaluated (e.g., Strand 2000; Hay, Warren & Drager 2006; Hay & Drager 2010; Campbell-Kibler 2011). In this talk, I consider how this attested stereotype effect interacts with broader psychological principles governing person perception, including selective attention and category activation/inhibition. I review some recent research on stereotypes in sociolinguistics that has looked specifically at how listeners respond to stereotypically incompatible cues in the speech signal (e.g., Campbell-Kibler 2009; Pharao et al. 2014; Levon 2014). I argue that the results of these studies point to two equally plausible interpretations: one in which social stereotypes serve to “block” the emergence of an incompatible category, and the other in which this apparent “blocking” is an incidental consequence of broader economical processing constraints.

I go on to describe a negative priming experiment (Neill 1977; Tipper 1985) designed to distinguish between these possible interpretations. The experiment is comprised of two related tasks: a priming task and a probe task. In the priming task, listeners rate a speaker for a given social characteristic (age, sexuality, social class) in the course of a standard matched-guise test. The purpose of this task is to prime respondents for the relevant social category. The subsequent probe task is then an adaptation of the Generalized Phoneme Monitoring (GPM) paradigm (Frauenfelder & Segui 1989), in which listeners identify occurrences of TH-fronting (i.e., the labiodental realisation of the interdental fricatives in English) in the same extracts used in the matched-guise test. Given the strong stereotypical association between TH-fronting and working-class speech in Britain (e.g., Kerswill 2003; Stuart-Smith & Timmins 2006), the central question investigated is the extent to which having been primed for a given category in the first task (e.g., “working class” or “gay”) facilitates or inhibits variant monitoring in the probe task. If facilitation/inhibition is found, this would provide strong evidence in favour of the active lateral inhibition of stereotypically-inconsistent traits in perception (e.g., Macrae, Bodenhausen & Milne 1995; Dijksterhuis & van Knippenberg 1996) and thus the notion that social stereotypes can “block” the emergence of incompatible categories, even in the presence of relevant linguistic cues. In the talk, I present current findings from this ongoing work, and discuss the ramifications that these findings have for our understanding of the relationship between social and cognitive factors in sociolinguistic perception.

Time: 1pm, Thursday 1st December

Venue: Room 118, Chandler House, 2 Wakefield Street

The Speech Science Forum (SSF) is a joint seminar series at UCL organised by the Department of Speech, Hearing and Phonetic Sciences (SHaPS) and the Institute of Cognitive Neuroscience (ICN).