SENTENCE PROCESSING AND THE MENTAL REPRESENTATION OF VERBS (LEXICAL ACCESS, VERB ACCESS)

This study seeks to determine the aspects of verb representations that are relevant to sentence processing, and to determine the operating characteristics of the devices that access information pertaining to verbs during the course of sentence comprehension.; The starting point for this inquiry involves two complimentary accounts of verb representation in the lexicon - syntactic subcategorization and argument structure. These accounts are used to address the relation between representational and processing complexity. Verb categories are produced that diverge with respect to syntactic subcategorization complexity and argument structure complexity. Several experiments are then run using a cross-model lexical decision task (CMLD) that taps into the sentence comprehension system in the immediate temporal vicinity of the verb.; The data generated by these experiments are used to make the following points: (1) The relevant processing complexity metric for verbs must be defined in terms of the number of different argument structure arrangements a verb enters into, and not the number of syntactic configurations a verb can dominate as previously hypothesized in the literature; (2) Evidence from morphology and the data obtained in this study show that verbs are organized in the lexicon by their argument structure, and this is presumably why the lexical access device is tuned to argument structure information, and not syntactic subcategorization; (3) All argument structure entries are fair game for lexical access. They are exhaustively accessed and made available for further operations of the sentence processor, like parsing; and (4) Verb access is data-driven, modular, and contextually impenetrable: The access device is oblivious to the structural information contained in the sentence prior to the verb.