| About | Program | Registration | Travel | Call for Abstracts | Previous Symposia |
Qualitative age effects in second language learning
Robert DeKeyser, University of Maryland
Most research on age effects in L2 learning documents quantitative differences as a function of age; sometimes there is an attempt at explaining the quantitative patterns by referring to various possible maturational causes or, on the other hand, to attribute them to various non-maturational confounds. Very little research so far has tested hypotheses predicting specific qualitative differences that follow from specific maturational views.In this talk, I will first give an overview of various previous attempts at testing relatively broad concepts of the 'critical period'. On the maturational side of the debate these include hypotheses about changing access to UG, a changing role of memory vs. analysis, changing memory resources, and a gradual shift in reliance on implicit vs. explicit learning processes; on the non-maturational side the most prominent examples are changing quantity/quality of input, different amounts of practice, different amounts of instruction, and different socialization patterns.
I will show, through an analysis of the existing literature, that the non-maturational views do not predict specific patterns matched by the data. I will also show that various maturational views all lead to specific hypotheses that mesh well with the data in at least some studies, so that there is not only descriptive and explanatory, but even some predictive adequacy for several of these views. In particular, I will assess how well the available data match the predictions of the less-is-more hypothesis and those of the implicit/explicit shift hypothesis, the latter both through an analysis of the differential role played by individual difference variables at different ages and through an analysis of the differential impact of age effects on different elements of the L2, even within domains such as phonology, morphology or syntax.
Besides drawing on the published literature, I will also use my own unpublished data from native speakers of various languages learning English or Hebrew as L2.