No abstract

German pronouns (er(MALE/masculine), sie(FEMALE/feminine)) that refer to a person are determined by the biological gender (MALE/FEMALE) and/or syntactic gender (masculine/feminine) of the person. Pronouns (er(masculine), sie(feminine)) that refer to a thing are determined by the syntactic gender of this thing (Garten [garden]masculine, Tasche [hand-bag]feminine). The study aimed to investigate whether semantic integration, syntactic integration, or both are involved in establishing co-reference between pronoun and subject/antecedent in sentences. Here we focused on two event-related potential components: the SPS/P600, related to syntactic violation and reanalysis, and the N400 component, related to semantic integration problems. In one condition, a person was introduced as antecedent and later referred to by a pronoun, which either agreed in biological/syntactic gender or not (biological/syntactic gender violation). In a second condition, a thing was introduced as antecedent and the corresponding pronoun either agreed in syntactic gender or not (syntactic gender violation). Results at critical pronouns showed a P600 effect for incongruent compared with congruent pronouns in both conditions with a centro-parietal maximum. This effect was larger for the person compared to the thing condition. We interpreted this finding as reflecting a syntactic integration process that can be influenced by conceptual/semantic and syntactic information of the antecedent type. Furthermore, at the word following the pronoun, we observed an N400 for the thing but not for the person condition. We suggest, supported by the results of a control experiment, that this effect reflects continuous integration processes for things, whereas for persons the integration seems to be finished at pronoun position.

No abstract

The subject and the object in a transitive clause should be distinguishable. We discuss five different ways in which distinguishability can be obtained: (i) case morphology, (ii) agreement, (iii) selectional criteria of the verb, (iv) word order, and (v) prominence, in particular animacy. We argue that these different sources of information can be viewed as violable and potentially conflicting constraints that play an important role in the incremental process of interpretation. We propose a new model of incremental optimization of interpretation and argue that we can thus account for the findings of several on-line studies in which distinguishability was manipulated.