Syntactic Structures as Descriptions of Sensorimotor Processes

Alistair Knott

doi:10.5964/bioling.8991

Abstract

In this paper I propose a hypothesis linking elements of a model of theoretical syntax with neural mechanisms in the domain of sensorimotor processing. The syntactic framework I adopt to express this linking hypothesis is Chomsky’s Minimalism: I propose that the language-independent ’Logical Form’ (LF) of a sentence reporting a concrete episode in the world can be interpreted as a detailed description of the sensorimotor processes involved in apprehending that episode. The hypothesis is motivated by a detailed study of one particular episode, in which an agent grasps a target object. There are striking similarities between the LF structure of transitive sentences describing this episode and the structure of the sensorimotor processes through which it is apprehended by an observer. The neural interpretation of Minimalist LF structure allows it to incorporate insights from empiricist accounts of syntax, relating to sentence processing and to the learning of syntactic constructions.

Highlights

Assumption: Two Stages of Language Evolution In Section 7.3 I envisaged a point during human evolution when a collection of ‘interface circuits’ evolved, allowing agents to produce overt behavioural reflexes of their internal sensorimotor representations
If we look first at concrete sentences, we must remember that we are looking in the sensorimotor domain for instantiations of general neural mechanisms, which we expect to find in other more abstract domains
There is a huge amount of compression; my main suggestion is that the basis for this compression is the temporal structure of the sensorimotor processes—its structure as a deictic routine

Summary

Introduction

Assumption: Two Stages of Language Evolution In Section 7.3 I envisaged a point during human evolution when a collection of ‘interface circuits’ evolved, allowing agents to produce overt behavioural reflexes of their internal sensorimotor representations (see Section 7.3). These circuits allow agents to learn a vocabulary of atomic behavioural symbols. These signals still need to be converted into overt motor movements. My suggestion is that the circuit which evolved to coopt the replay mechanism for a communicative purpose transforms the sequence of signals evoked in the premotor medium during replay into an overt sequence of motor movements

Objectives

Discussion

Conclusion