The question of how to properly study ethnicity or race in the advent of complex network studies is widely debated in sociolinguistics. Scholars such as Eckert (2000) and others (Johnstone 2000, Bucholtz & Hall 2005, Fought... [ view full abstract ]
The question of how to properly study ethnicity or race in the advent of complex network studies is widely debated in sociolinguistics. Scholars such as Eckert (2000) and others (Johnstone 2000, Bucholtz & Hall 2005, Fought 2006) argue that the canonical patterns being researched in sociolinguistics underestimate the complexity of the social networks and communities of practice in which they come to play.
Though these objections have been widely discussed and criticized, self-identified ethnicity remains a concept that yields strong results in regression modeling (e.g. Podesva 2016). My paper addresses the issue of ethnicity as a predictor in variationist research with special attention to how this identifying factor overarches the formation of communities of practice. Specifically, I will controvert to what extent ethnicity may be the more significant factor in predicting the outcome of a variable than communities of practice. The variable chosen for elaboration is /s/-retraction, a much studied phonetic phenomenon with (thus far) no ethnic positioning in the US.
This study is based on data collected in a guised memory experiment conducted in Austin in 2016. Partial transcripts were force-aligned with FAVE (Rosenfelder et al. 2011) and /str/-clusters were rated based on comparison to /s/ and /ʃ/ Center-of-Gravity measurements taken in Praat (Boersma & Weenink 2012).
I will juxtapose ethnicity with the membership in a certain social group as a factor in retraction of /s/. This illustrates that even for participants enrolled in the same college classes or the same senior activities program, ethnicity remains a stronger predictor for their distinct pronunciation of /s/ in str-clusters than their membership in these networks or communities of practice. However, ethnicity will not be presented as a simple biologically formed identity factor. Instead, I will discuss in what sense it could be considered a factor in the formation of communities of practice. I argue that the strength of ethnicity as a factor in statistical analyses is not due to over categorization, but rather because ethnicity and race are still defining factors in the process of identity formation and creation of meaningful social groups.
