The emic and the etic in perceptual dialectology

The study of language regard treats the thoughts, beliefs, attitudes, and actions of nonlinguists as central to our understanding of linguistic variation (Preston 2010b). Focusing on the evaluation of linguistic forms by speakers themselves allows for clearer insight into the cognitive processes that motivate linguistic choices, and the study of language regard permits an analysis of the attitudes and ideologies that come to bear on identity constructions and encourages analysts to delve deeper into this confluence of macro- and micro-level phenomena. Perceptual dialectology (cf. Preston 1989, 1999a; Long & Preston 2002; Cramer & Montgomery 2016), a subfield within the broader field of language regard, employs many different methods to capture these processes of identification in this crossroads of what Pike (1954) would call the emic and the etic. In this chapter, I explore how perceptual dialectology research produces emic and etic descriptions of nonlinguists’ perceptions, and through an exploration of how perceptual dialectologists have manipulated the emic/etic coin, I reveal how the study of language regard has been bolstered by obtaining both perspectives, as such investigations into the knowledge of nonlinguists living in their dialect landscapes “are important to the most basic questions of language variation and change, in some cases even providing explanations for otherwise puzzling events” (Preston 2010a:88). And while it will be shown that perceptual dialectology has not neglected these two perspectives, I also discuss some of the ways in which previous attempts to address both sides of the emic/etic coin were disjointed. In particular, early perceptual dialectology studies (e.g., Preston 1989) maintained a division between the emic representations of the dialect landscape created through the use of individual mental maps and the etic representations acquired through the rating of individual states in terms of degree of difference, pleasantness, and correctness. Preston (1999b) saw limitations in the top-down approach of giving participants what the researcher believed to be the important categories, and he instead had one set of participants write down the words they would use to describe a region, and with each word’s oppositional partner (e.g., friendly/unfriendly, educated/uneducated), another set of participants rated the states on those characteristics using a Likert scale. Such a technique still values the emic perspective, in that local residents were the ones who came up with the social descriptors, but it still allows for researcher control in terms of generalizability.