Conventional Representation of Natural Language

Noel Burton-Roberts

 Abstract

In generative linguistic theory, the question of how the sub-personal (the innate, natural Faculty of Language, or Universal Grammar) enters into personal action (thought, and speech in particular languages) has not been satisfactorily answered. In the most recent version―Minimalism―it becomes acute, giving rise to pervasive tensions in the theory.

 While Minimalism constitutes a radical departure from traditional linguistic assumptions, it nevertheless retains the traditional assumption that the language faculty is a syntactic system operating over a Saussurian vocabulary―a vocabulary consisting of sound-meaning pairs. Chomsky refers to this as ‘the double-interface property of language’. On the basis of this traditional assumption, the structured expressions of natural language are supposed to be ‘realised’ (instantiated), as sound, in acts of utterance (speech).

 I will argue that it is this “double-interface” assumption that gives rise to the tensions, and to the personal/subpersonal problem in the linguistic context. It is worth noting that the assumed relation of “realisation” is completely unexplained.

I will outline a new hypothesis about the relation between

(a) the personal acts and external sounds of speech in a diversity of particular, acquired, conventional languages

and

(b) the expressions of mind-internal, innate, universal/invariant, sub-personal, natural language.

This (Representational) hypothesis rejects both the notion of ‘realisation’ and the double-interface assumption that, to explain the relation between (a) and (b), it is necessary to postulate entities (expressions) having properties pertaining to both (a) and (b). Instead, the acts and sounds of speech have exclusively phonetic properties and are conceived of as standing in a relation of conventional representation to the (exclusively syntactico-semantic) generated expressions of natural language. Speaker-hearers do not produce or hear or parse those expressions―they produce/hear/parse only phonetic representations of them. I outline implications of this idea for the architecture of the language faculty, for the distinction and relation between what is innate and what acquired and for the personal/subpersonal relation in the linguistic context.