Variable analysis

The quantitative study of variables has been a vital analytical tool in the development of the discipline of sociolinguistics since its emergence from other related fields in the mid twentieth century. The work of William Labov (1966, 1972) was revolutionary in its methodological rigour and its ability to draw on large samples of data in order to better understand the relationships between social categories and language variation. Subsequent scholars of variationist sociolinguistics (Milroy and Milroy 1978, Eckert 1989, Kiesling 2004) built on these methods by integrating greater degrees of speaker agency into the social categories under examination, resulting in an increased focus on locally-constructed identity categories. Recently, Hawkey (2018) advanced the use of attitudinal data as potential quantitative correlates of language variation, uniting the agentive focus of recent variationist scholarship with quantifiable attitudinal findings. The complementarity of language attitudes and variable analysis has long been overlooked, and this chapter seeks to uncover how variable analysis can enrich our knowledge of attitudes.

Since Rosenberg and Hovland’s (1960) early tripartite division of attitudes into affective, conative and cognitive components, social psychological studies have attempted to elucidate the links between attitudes and behaviours, accepting the behavioural element as integral to the attitude itself. By using attitudinal findings as independent variables in the study of linguistic data, we are first able to rigorously ascertain the links between attitudes and practices. When this is then viewed in conjunction with broader qualitative data (also obtainable through approaches already common in attitudes studies), we can better understand how our knowledge of language behaviours (i.e. variable language usage) is directly linked to attitudes and ideologies.

This work will focus on Catalan as spoken in southern France (in the region of Northern Catalonia), thus also offering insights into the specific attitudinal dimensions of a situation of language obsolescence. This study examines phonetic, morphosyntactic and lexical variables that were elicited by means of a wordlist translation task. Participants were provided with a list of words and phrases in French, and asked to give translations into Catalan. Participants were of a broad age range (28 to 97) and came from two settings: the urban centre of Perpignan and the rural village of Serdinya, some sixty kilometres away.