The pervasive and evasive S-curve of linguistic change
Quentin Feltgen (Institut du Cerveau et de la Moelle Epinière (ICM))
Language is a stable convention, established over a community by way of repeated and multiple interactions between its members. However, within this view, the occurrence of change is somehow problematic. First, if members align towards one another to allow for a convention to be established and maintain, a novelty, being an isolated event to begin with, should not spread over the speakers’ network – an issue known as the Threshold Problem as addressed by Nettle in 1999. Second, the spread of the new variant should occur according to an S-curve; yet, as Blythe & Croft (2012) have shown, this S-curve can only arise thanks to very specific conditions: either the community is highly hierarchical, with change spreading from top to bottom, or the new variant is somehow inherently better than its established competitor. Concerningly, these two conditions seem unfitting to explain the wide array of constantly ongoing linguistic changes. In this talk, we shall offer an alternative view based on the fact that the S-curve is associated with both the lexical and the social diffusion of a language change, and can therefore be the result of some basic, shared features of the cognitive organization of language. The mathematical model that we have developed allows to make sense of disparate empirical findings within a unified, usage-based theoretical frame. It also challenges the traditional relationship between change and variation in predicting that variation could be a stable outcome of language change.
Participation is free, but please register your attendance by sending an email to email@example.com
Room to be added.