Recent "representational deficit" approaches to SLA (e.g. Hawkins 2000, 2003, Hawkins and Chan 1997, Hawkins and Liszka 2003, Tsimpli 2003) attribute L2 morphophonological variability or error to a deficit in the selection of parameterized formal features. According to the representational deficit approach, parameterized formal features (perhaps restricted to uninterpretable ones) which are present in the L2 but not selected in the learner's L1 are hypothesized to be unacquirable due to critical period effects. This paper focuses not on the selection but rather on the assembly of features in SLA. Drawing on detailed case-study data from Patty in the areas of number, definiteness, wh-movement, pronominal casemarking and subject-raising (implicating the presence of uninterpretable features in the grammar), the author argues that accounting for morphological variability by appealing to the parametric (non-)selection of features is too simplistic. Instead, the ways in which grammatical features are combined and conditioned in morphological exponence plausibly affects their overt realization in SLA. The author discusses the sorts of learning problems presented by the different ways in which primitive features are clustered in different languages (specifically, in the L1 vs. the L2), and argues that there is a kind of morphological competence that is distinct from performance, and distinct from (narrow) syntactic representation.
Proceedings of the 7th Generative Approaches to Second Language Acquisition Conference (GASLA 2004)
edited by Laurent Dekydtspotter, Rex A. Sprouse, and Audrey Liljestrand
Table of contents