Semantic relation priming is not constituent-specific: Evidence from ERP

Abstract

It is humankinds unique wisdom to compose a limited number of words together through specific rules to convey endless information. Researchers have found that this composition process also plays a vital role in the comprehension of compounds. The specific manifestation is relation priming; that is, the previously used relation will promote subsequent word processing using the same relation. This priming phenomenon is bound to morpheme repetition (modifier or head). This study combines a self-paced priming paradigm with electrophysiological technology to explore whether relation priming will occur without sharing morphemes and its time course. We found that relation priming can occur independently of morpheme-repetition, which shows an independent representation of relation information. And it has been activated at a very early stage (about 200ms). As the word processing progresses, this activation gradually strengthens, indicating that the relations role is slowly increasing in the process of compound word recognition. It may first be used as a kind of context information to help determine the constituent morphemes meaning. After the meaning access of the constituent morphemes, they begin to play a role in the semantic composition process. This study uses electrophysiological technology to precisely describe the representation of relation and its time course for the first time. Which gives us a deeper understanding of the relation priming process, and at the same time, shed light on the meaning construction process of compounds.