Effects of age on the acquisition of agreement inflection

Grammaticality judgement tasks show that second language learners who started during childhood are significantly more accurate on judging inflection than learners who started after puberty [Johnson, J., & Newport, E. (1989). Cognitive Psychology, 21, 60-99; Johnson, J., & Newport, E. (1991). Cognition, 39, 215-258; McDonald, J. (2000). Applied Psycholinguistics, 21, 395-423]. Production data confirmthat inflection is a bottleneck in adult language acquisition, and that they differ from child learners in this respect [Lardiere, D. (1998). Second Language Research, 14, 359-375; Prevost, P. (2003). Studies in Second Language Acquisition, 25, 65-97; Prevost, P.,& White, L. (2000). Second Language Research, 16(2), 103-133]. Although the observations suggest that the acquisition of inflection is influenced by age, there is no study that focuses on this particular issue nor is there an articulated explanation available for the observed age-related difference. In this contribution, we compare child L2 learners of Dutch to child L1 and adult L2 learners of Dutch in order to investigate effects of age on the acquisition of verbal and adjectival inflection. We hypothesize that adult agreement paradigms differ from child agreement paradigms, the reason being that adult learners cannot rely on syntactic cues, whereas children make reliable use of syntax in building paradigms. By effect, adult learners end up with non-targetlike small paradigms that contain underspecified suffixes. We focus on the types of errors in the three learner groups (child L1, child L2 and adult L2). Our empirical basis consists of results obtained in a series of production experiments.