Categories, stereotypes, and the linguistic perception of sexuality

Abstract

A B S T R A C TThis article examines how social stereotypes influence listeners' perceptions of indexical language. Building on recent developments in linguistics and social psychology, I investigate the extent to which stereotypical attitudes and beliefs about categories of speakers serve to enable the association of linguistic features with particular social meanings while simultaneously blocking others. My arguments are based on an analysis of listener perceptions of the intersecting categories of gender, sexuality, and social class among men in the UK. Using a modified matched-guise paradigm to test three category-relevant variables (mean pitch, spectral characteristics of /s/, and TH-fronting), I demonstrate how the perception of social meaning is governed by a combination of both attitudinal and cognitive factors. This finding is important because it illustrates the listener-dependent nature of sociolinguistic perception. Moreover, it also provides further empirical support for an understanding of social meaning as an emergent property of language-in-use. (speech perception, attitudes and stereotypes, sexuality, phonetic variation)*

Perception in a Variable but Structured World: The Case of Speech Perception

“…This choice is based on feasibility: our analysis depends on having sufficient data to estimate cue distributions for multiple talkers, within groups defined by socio-indexical variables. Sociolinguistics increasingly recognizes that the meanings of socio-indexical variables are dynamically constructed and not necessarily static within a single individual (either producer or perceiver; Eckert, 2012a;Podesva et al, 2001;Podesva, 2007;Foulkes & Hay, 2015;Levon, 2014). The techniques we introduce here can-in principle-be applied to more dynamic and contextually conditioned variables, but they present unique and interesting challenges that are beyond the scope of the current paper.…”

“…The subjective nature of these expectations is particularly important, because they may deviate in substantial ways from the actual nature of cross-talker variation. For instance, arbitrary aspects of linguistic variation can become "enregistered", where listeners consider particular variants stereotypical of social groups, whether they are in fact differentially associated with them or not (see, e.g., Eckert, 1989;Eckert, 2012b;Podesva, Roberts, & Campbell-Kibler, 2001;Niedzielski, 1999;Levon, 2014). highlights areas where more data is needed.…”

“…Rather, some variants can become disproportionately salient or enregistered, can further vary based on context (Eckert, 2012a;Podesva et al, 2001;Podesva, 2007;Foulkes & Hay, 2015;Levon, 2014;. These deviations between objective informativity and subjective salience remain to be explained and specified in more detail.…”

“…One example comes from Levon (2014). He finds that when UK listeners hear a male talker with high /s/ spectral center of gravity (COG), they infer that the talker is a gay man.…”

Matched guise effects can be robust to speech style

“…Evidence for the context-sensitivity of sociolinguistic meaning is not limited to ING. For example, recent work on the perception of sexuality in both the United Kingdom (Levon, 2014) and Denmark (Pharao et al, 2014) indicates that various linguistic correlates of perceived sexuality are contingent on other cues in the acoustic signal.…”