Abstract Ruth Kempson 22 Nov 2017
Language as a Tool for Interaction: An Evolutionary Tale
In this talk I’m going to air an account of language evolution which combines insights from linguistics, psychology, neuroscience, and evolution theory: that languages are systems evolved for interaction. This is a view which conversational analysts have been urging for a long time, and at last the news is spreading through psychology, neuroscience, philosophy (Healey 2007, Anderson, M. 2014, Clark, A. 2016), and also in linguistics via Hawkins, Tomasello, Christiansen & Chater, Simon Kirby et al. But the Dynamic Syntax perspective that I will sketch is not just a usage view that languages are constrained by performance considerations. It is the stronger view that languages are adaptive systems whose properties are evolved mechanisms for inducing interaction.
The talk has four parts. It starts with data from informal conversation and why these pose a challenge for language modelling. Then, first I sketch the Dynamic Syntax view that language as modelled in a grammar is a set of actions for growing information reflecting real time dynamics (Kempson et al 2001, 2016, Gregoromichelaki et al 2011). Second, I show how this perspective fits the embodied cognition account of A.Clark (also M.Anderson), with language a case study for the view that all cognitive processes are anticipatory action processes. Third, I bring in the Multi-Level Selection Hypothesis (Sober & D. Sloan-Wilson 1998, Sloan-Wilson 2007) that groups can be adaptive units in their own right. From this three-fold synergy, we get a glimpse of why and how the language capacity could have gradually evolved; for, contrary to what the orthodoxy dictates, on this account, first language acquisition and language change are forms of language evolution.