Modularity in locating wh-gaps

The role of semantic/pragmatic information in gap-filling was examined using a CMLP (cross-modal lexical priming) methodology to test the prediction that a reactivation effect for implausible wh-phrases will be elicited following remind-type verbs. Ss (N = 12 undergraduates) read 24 pairs of sentences under conditions varied by plausibility (plausible vs implausible), probe type (related vs control), & probe site (preverb vs postverb). Mean response times show higher priming facilitation for related probe wh-phrases at postverb sites for both plausible & implausible wh-phrases. Related probes were recognized faster than unrelated probes. Main effects of plausibility & probe site were not observed. The data suggest that gap-filling processes are not due to semantic/pragmatic conditions. Wh-phrases were reactivated (primed) in direct-object position regardless of plausibility. It is argued that modular models of gap-filling are supported by the effect gained using three place predicates such as "remind." The findings obtained through CMLP are compared with those from embedded analytic techniques. 1 Table, 24 References. Adapted from the source document