Response Retrieval in Active and Passive Sentences

Retrieval latencies for active and passive sentence examples are compared. The argument is proposed that the sentences are searched by means of a strategy that permits direct entry into specific word groupings. Both syntactic structure and pausing behaviour are considered as indices of the groupings employed. It is concluded that neither descriptive source can on its own fully account for the obtained latency profiles, although the structural description derived from pause location provided a better prediction than constituent analysis. The implications of this conclusion are discussed.