For different reasons, speakers re-use recently used or heard linguistic options whenever they can, a tendency which is referred to as ‘persistence’ in the present paper. The phenomenon has been largely neglected in extant corpus-based, variationist research, and no standard methodology for dealing with the phenomenon is available. By analyzing three well-known alternations (analytic vs. synthetic comparatives, particle placement, and future marker choice) in several spoken corpora of English, this paper demonstrates that factoring in persistence increases the researcher’s ability to account for linguistic variation. It is also shown that persistence itself is subject to several determinants, such as textual distance between two successive choice contexts in discourse, or turn-taking. In conclusion, I argue that persistence is a factor which deserves empirical attention, and that its existence has consequences for both linguistic theory and practice.
Stored only when both papers are in the corpus — 286 citations exist upstream.