Abstract
In this study we investigated how monolingual and bilingual learners of German between five and ten years of age resolve anaphoric pronouns. We employed a visual world design to measure what parts of a visual stimulus participants attend to while presented with short spoken texts, in which a sentence beginning with an anaphoric pronoun follows a sentence which includes two referents. In order to shed light on pre- and post-pronominal processing we manipulated (1.) ‘gender’ and (2.) syntactic function/position in the preceding sentence of the referents (in the following the latter will be referred to as, ‘antecedent type’). Results suggest that the reliable cue ‘gender’ is acquired earlier than the probabilistic cue ‘antecedent type’, which is in line with previous findings in the field. However, our findings show that in monolingual and bilingual children the reliability of the gender cue steadily increases even up to age 9. With respect to ‘antecedent type’, as a probabilistic cue in German anaphoric pronoun resolution, our results suggest differences between acquisition types. Monolinguals use this cue during pre-pronominal processing at an early age, and only at later ages during post-pronominal processing, together with the gender cue. Bilinguals, on the other hand, do not use the cue at an early age, and when they do at age seven, they use it during post-pronominal processing. Only the oldest bilingual participants use ‘antecedent type’ information for pre-pronominal processing. Taken together, our study shows that all factors manipulated in our experiment (age, acquisition type, cue type) have to be considered when explaining how anaphoric pronoun resolution is acquired by language learners.
