The Linguistic Variable: Fact and Fantasy

Abstract

REGARDLESS OF DESCRIPTIVE TRADITION, virtually all studies of fluctuating linguistic variants operate with some notion of the LINGUISTIC VARIABLE. For example, a traditional, qualitative dialect study that sets forth variants of final unstressed vowels of items like sofa and china (see Kurath and McDavid 1961) as a choice between [a], [3], and [i] operates with some underlying notion of the linguistic variablejust as surely as a sophisticated, probabilisticbased model for displaying variation in the deletion of final /t/ and /d/ in consonant clusters (Guy 1980). While all studies of linguistic variation operate on some notion of the linguistic variable, the underlying definition of a linguistic variable is often implicit rather than explicit. In fact, since the original formulation of the linguistic variable by Labov (1966a, 1966b), there have been few discussions of how this construct is defined. Most studies of linguistic variation are more preoccupied with working out the descriptive details of fluctuation than they are with the specification of parameters that bind fluctuating forms. Several decades into the tradition, the field remains plagued by some fundamental questions about the linguistic variable. Furthermore, we will see that the basic definition of the variable has changed in subtle but important ways over time. Notwithstanding the discussion of the levels of language to which this construct might apply (Lavendera 1978, Labov 1978, Romaine 1982), we are still left with lingering questions about the fundamental nature of this construct for any level of language. What exactly is the linguistic variable anyhow? Is it a construct that has independent linguistic validity? Is it a specialized construct uniquely sociolinguistic? Does it have independently motivated sociolinguistic validity, or is it simply an operational heuristic that provides us with a manipulative starting point in the examination of linguistic and social covariation? The fundamental nature of such a construct is hardly trivial if we are to take the notion of systematic variation seriously, and variationists' current concentration on developing more sophisticated methodological models for displaying data will not make it go away. In the following discussion, I shall examine some of the types of linguistic variation subsumed under the rubric of the linguistic variable, as well as the way in which this construct has developed within the variationist tradition. I do not propose to answer all of the fundamental questions surrounding this intriguing construct, but at least I hope to open the notion up for serious, reflective discussion.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 65 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.