Religion as a Sociolinguistic Variable

When considering variables that are rarely coded for in sociolinguistic studies, we discovered that general demographic studies [e.g., the census, Pew Research studies] have only recently realized that the question sets provided for demographic information rarely permit coding of religion, or when they do, provide only coarse-grained coding; yet recent studies in sociolinguistics and social anthropology have demonstrated that fine-grained distinctions in religious identification are necessary to account for sociolinguistic variation. This paper reviews the information from both sociolinguistic studies and demographic studies which should be considered when developing a protocol for analysis of speech variation. The paper also points out that the variation is often due to network effects from the ‘community of practice’, but at least some of the variation can be traced to ideological positions or choice of referee, both of which appear to influence language use.