Invariance as the Mark of the Psychological Reality of Language

Devitt articulates and defends what he calls the ‘linguistic conception’ of generative linguistics, where this position stands in contrast to the prevailing ‘psychologistic conception’ of Chomsky and generative linguists generally. I shall argue that the very idea of anti-psychologism vis-à-vis generative linguistics is premised upon a misunderstanding, viz., the thought that there are linguistic phenomena as such, which a linguistic theory may target directly, with psychological phenomena being targeted only indirectly. This thought is incorrect, for the ontology of a theory is ultimately what is invariant over and essential to the explanations the theory affords. In this light, linguistic theory is about psychological phenomena because the psychological states of speaker-hearers are the invariances of linguistic explanation, and there are no such invariances that involve externalia. What ultimately counts as psychological itself is partly determined by the very kind of explanations our best theories offer. In a nutshell, the explanations of generative theories neither entail nor presuppose an external linguistic reality, but do presuppose and entail a system of internal mind/brain states the theories seek to characterise.