英文摘要 |
A cross-modal lexical decision (CMLD) task was conducted to measure subjects’ reaction times (RTs) at three probe sites simultaneously with their comprehending auditory Mandarin relative clause (RC) sentences. A baseline was first established by asking participants to complete a lexical decision task (LDT) based on a set of visual stimuli, which was then used as visual probes in the CMLD task. Since RTs for the visual targets in conjunction with spoken RC sentences were longer than RTs for the same targets in isolation in the pure LDT task, the RTs for these visual targets under different RC sentences in the CMLD task were hypothesized to index the degree of processing difficulty of different RC sentences. Significant RT differences between subject-extracted RC (SRC) sentences and object-extracted RC (ORC) sentences were found at the main clause (MC) verb: At the post-main-verb site, subjects spent significantly less time in the case of SRC sentences. The subjects’ comprehension performance demonstrated no differences in the interpretation errors between SRC sentences and ORC sentences. However, subjects had a higher percentage of errors in the RC domain than in the main clause (MC) domain and this effect was more obvious in ORC sentences. Our results conform to previous findings that suggest the locus of processing difficulty is at the MC verb, as listeners must complete multiple thematic role assignments at this point. Therefore, a thematic fit between the RC and the MC might play a particular role in spoken RC processing. |