Optimal and Non-Optimal Interpretations in the Acquisition of Dutch Past Tenses

An aspectual comprehension experiment with Dutch learners reveals an asymmetry in their understanding of the Present Perfect on the one hand versus the Imperfective Past and Periphrastic Past Progressive on the other. Dutch 2 and 3-year-olds have acquired the completion entailment of the Perfect, but do not properly restrict their interpretation of the Imperfective Past and Periphrastic Past Progressive tenses. With the latter children accepted both completion and ongoing situations, in contrast to adults, who favored only ongoing situations. I argue that adults take both speaker and hearer perspectives in comprehension into account. Such bidirectional reasoning about tense forms and their aspectual meanings leads to the selection of the best form/meaning associations and the blocking of less good associations; blocking applies when there exists a better form to express a certain meaning. Children, however, cannot do bidirectional optimization, and hence do not block out non-optimal form/meaning pairs. I show how this results in asymmetry: an overly liberal interpretation of the Imperfective and Progressive tense forms, but target-like understanding of the Perfect. De Hoop and Krämer (2005/2006) recently modeled the hypothesis that children do not apply both hearer and speaker perspectives in the framework of bidirectional Optimality Theory (OT). When they need to assign meaning to a certain form, children take a pure hearer perspective and when they need to produce a form to express some meaning, they take a pure speaker perspective. De Hoop and Krämer use this model to explain children’s non-target-like interpretations of indefinite subjects and objects. Hendriks and Spenader (2005/2006) use it to explain an asymmetry in the comprehension and production of reflexives and pronouns. Bidirectional OT is advanced as a model that is well-suited to explain such asymmetries. In section 2 I discuss the aspectual semantics of three different past tenses in Dutch, and raise some acquisition questions. Section 3 introduces a bidirectional OT analysis of these aspectual form/meaning pairs and spells out the predictions of the hypothesis that children do not apply bidirectional reasoning. Section 4 presents an experiment on children’s (and adults’) interpretation of the three Dutch past tenses. In section 5 I draw some conclusions.