Long-term structural priming involves a memory-based mechanism

Abstract Speakers adapt their syntactic preferences based on syntactic experience. However, it is not clear what cognitive mechanism underlies such adaptation. While error-based mechanisms suggest that syntactic adaptation depends only on the relative frequency of syntactic structures, memory-based mechanisms suggest that both frequency and recency of syntactic structures matter in syntactic adaptation. To distinguish between these two mechanisms, I manipulated the order of passive and active primes in two syntactic priming experiments, presenting passive primes either before active primes (active-recent condition) or after them (passive-recent condition), while controlling for frequency. The results showed that the magnitude of priming was numerically greater in the passive-recent condition than in the active-recent condition in Experiment 1, and significantly greater in Experiment 2. These results provide novel evidence that syntactic adaptation involves a memory-based mechanism.