Nomi M. Olsthoorn & Gerard A.M. Kempen
Leiden University
The syntactic priming effect is a fact of life, at least for psycholinguists it is. After having processed a sentence in a particular word order, subsequent production or perception of a new sentence in a similar word order is facilitated, response bias being the measure most frequently obtained (however, cf. Smith & Wheeldon's Cognition-2000 paper). The effect is syntactic in nature; semantic, lexical or phonological factors do not play a role. Explanations for the effect are usually in terms of effort reduction. Re-using old material, in this case the grammatical structure of sentences, saves cognitive resources, which allows mental capacity to be concerned with non-syntactic aspects of language processing, lexical retrieval for instance, instead of with syntax. The question we asked ourselves is: does syntax recycling indeed reduce cognitive effort, as measured in reaction time decrease? The answer seems to be yes and no. Voice Onset Times (VOT's) were recorded for cued picture descriptions following either syntactically compatible or incompatible prime sentences. In a series of experiments we find that a priming effect as measured in VOT decrease for compatible prime-target pairs does not obtain when the cued target sentence word order does not allow the participant a free choice of grammatical structure (such as in cued Subject-Verb inversion target responses in Dutch). However, allowing the participant a choice between two alternative and equivalent word orders (such as in the Dutch perfect tense in which auxiliary verb -past participle order is free) does seem to result in a VOT decrease, indicating that syntactic priming is, as are not many other facts of life, just a question of choice.