Social priming or not? Investigating generalizability via individual variation

Abstract

Although the effects of social priming-where social information about a speaker influences linguistic decisions-have long been accepted (e.g., Niedzielski, 1999; Strand, 1999; Hay et al., 2006), recent studies call attention to a lack of generalizability across contexts (e.g., Lawrence, 2015; Chang, 2017; Walker et al., 2019). The current study finds that social priming, in the case of /a℧/-raising in Michigan and Canadian English, is primarily observed for individuals who scored high on both stereotype awareness and cognitive perspective-taking, suggesting that aspects of both the sociolinguistic context and participant sampling can impact when and why social priming effects do not occur.