Key points are not available for this paper at this time.
Abstract In this article we investigate what learners acquire when their input contains inconsistent grammatical morphemes such as those present in pidgins and incipient creoles. In particular, we ask if learners acquire variability veridically or if they change it, making the language more regular as they learn it. In Experiment 1 we taught adult participants an artificial language containing unpredictable variation in 1 grammatical feature. We manipulated the amount of inconsistency and the meaning of the inconsistent item. Postexposure testing showed that participants learned the language, including the variable item, despite the presence of inconsistency. However, their use of variable items reflected their input. Participants exposed to consistent patterns produced consistent patterns, and participants exposed to inconsistency reproduced that inconsistency; they did not make the language more consistent. The meaning of the inconsistent item had no effect. In Experiment 2 we taught adults and 5- to 7-year-old children a similar artificial language. As in Experiment 1, the adults did not regularize the language. However, many children did regularize the language, imposing patterns that were not the same as their input. These results suggest that children and adults do not learn from variable input in the same way. Moreover, they suggest that children may play a unique and important role in creole formation by regularizing grammatical patterns.
Kam et al. (Fri,) studied this question.