Resolving number ambiguities during language comprehension

This paper investigates how readers process number ambiguous noun phrases in subject position. A speeded-grammaticality judgment experiment and two self-paced reading experiments were conducted involving number ambiguous subjects in German verb-end clauses. Number preferences for individual nouns were estimated by means of two questionnaire studies and a corpus analysis. The data show garden-path effects for locally ambiguous sentences, suggesting that readers resolved the number ambiguity immediately. Substantial correlations between experimental results and results from the norming studies indicate that the choice of either singular or plural is determined by exposure-based lexical preferences. We discuss the implications of our results for the relationship between lexical and syntactic processing, in particular with regard to the costs incurred by revising earlier lexical decisions. In addition, we address two methodological issues: first, the appropriate way to obtain lexical preferences, and second, the measurement of garden-path effects in verb-end sentences disambiguated by the final word.