Alissa Melinger & Christian Dobel
Max-Planck Institute for Psycholinguistics, Nijmegen
Syntactic or structural priming refers to the phenomenon of using a particular syntactic structure given prior exposure to the same structure. This behavior has been observed when speakers hear, read or write sentences (overview: Pickering & Branigan, 1999). Syntactic priming in the production literature has only been observed when the prime is a sentence or sentence fragment. However, recent studies from the sentence comprehension literature suggest that a single verb presented in isolation may be sufficient to drive syntactic priming (Trueswell & Kim, 1998). In the production literature, however, it has been claimed that morphosyntactic features associated with lemma representations are only retrieved when they are needed for the relevant construction. Specifically, subcategorization information is retrieved when producing a sentence which requires this information but not when naming a verb. If this claim is true, priming in sentence production should only be observed in the context of full sentences or sentence fragments and not in the context of single word primes.
To investigate these issues we conducted a series of primed picture description experiments in which we attempted to bias speakers towards producing picture descriptions with a particular syntactic structure based on the characteristics of the verb prime. Dutch native speakers saw ditransitive verbs that are restricted either to a prepositional (dative) or double object construction, e.g., "transfer" (PP only) vs. "refuse" (DO only). Following the presentation of the verb, participants are asked to describe a drawing depicting a three-participant event. Across experiments we varied the degree to which participants processed the prime by changing the secondary task (e.g., recall vs. naming). In the first experiment, participants additionally kept track of the prime words for a secondary recall test that was interleaved throughout the experiment. In the second experiment, we replaced the secondary recall test with a prime naming task. In the third experiment, participants used the prime verb in a sentence.
We found that a single word is sufficient to prime a particular sentence frame. Furthermore, our results show that the more actively the participant processes the verb, the stronger the priming effects become. The results of these three studies are discussed in light of the current claims about the locus of syntactic priming and the representation and access of morphosyntactic information.