Asymmetries in the acquisition of definite and indefinite NPs

Definite noun phrases are typically associated with established discourse referents familiar to speaker and hearer, while indefinite noun phrases are used to introduce new ones. The literature shows that children overuse the up to the age of 5 when referring to a new discourse entity, where the target grammar requires an indefinite. We investigate if this problem in production is accompanied by problems in comprehension, and so if there are any asymmetries between production and comprehension. Experiment 1 tested 25 learners of English (mean age 4;0) and Experiment 2 tested 19 learners (mean age 4;6). In both studies the preschoolers took a production as well as a comprehension task. The method for the comprehension tasks was different in each study: truth-value-judgment in Experiment 1 and a new method which we call the referent-selection paradigm. The production tasks involved structured elicitation. We find two asymmetries. First, while children indeed overuse the, their interpretation of definites is on target, and so it seems that comprehension precedes production, which is unsurprising. Second, with indefinite noun phrases children make no errors in production, but their interpretation is overly liberal. The latter asymmetry with production preceding comprehension is a surprising pattern. We explain it in terms of immature pragmatics, arguing that children fail to draw a scalar implicature, and we model our analysis in bi-directional Optimality Theory.