The rest of the paper will be spent expositing and, when necessary, critiquing, some of the literature on second language processing. I will pay particular attention to the recent Shallow Structure Hypothesis (SSH), due to Clahsen and Felser. In contrast to squabbles about the vague notion of access, the SSH provides a concrete, empirically backed interpretation of second language processing. The theory has the potential to shed substantive light on the workings of second language. It goes to show, by contrast, that the access question can at best be considered a secondary concern.
[...] The article substantiates the existence of language specific module, and thus lends legitimacy to the access question. Might we then expect the access question to revolve around evidence from cognitive neuroscience? We might, but we would be mistaken: if the work of Epstein et al. (1996) is any indication, output data are the guiding lights in the access debate. Obviously, this should raise some concerns. Moro et al's article, for example, raises the following issues: can non-language specific domains of the brain be used to learn foreign languages? [...]
[...] The results from filler-gap dependency tests starkly illustrates this point: whereas consideration of output data and grammaticality judgments would indicate access to UG in this respect, a detailed look at speaker processing goes to show that fundamentally different operations are at work Conclusion On reflection, perhaps I should amend my initial statements: this paper does not quite show that the access question is substantively void, or that it is unanswerable. What this paper does make clear is that the access question is a broad one. [...]
[...] The upshot is that even if the access question is construed in such a way as to allow a straightforward answer, it appears as if this answer will be nothing more than the tautological fulfillment of some definition; e.g., neuroscientific data proves that first and second language representation overlap in part, and therefore second language learners have partial access to UG. This is immensely unsatisfying: what we should really be interested in is the details behind the differences, and how and why they manifest themselves as they do. [...]
[...] She goes on to state, with regard to semantically incorrect sentences, that the problems that both native speakers and second language speakers had with semantic integration processing were essentially the same. With regard to semantics in general, she notes that ERP differences were only quantitative,” indicating that during semantic processing the two groups exhibited similar brain processes, but to different extents. This contrasts sharply with the nature of syntactic processing, wherein “second-language learners did not process syntactic category information and its integration into the existing phrase structure in the same way as native listeners in other words, second language learners showed ERPs which were different in kind, and not just in degree, from native speakers. [...]
[...] Of course, this is an enormous project, and perhaps one that encompasses nearly all aspects of second language research; but the access question is a broad and multifaceted, and thus one should expect that its answer requires a similarly broad set of facts. It should be clear by now that the way to avoid the overly vague question of UG access while still addressing the status of UG in an SLA context is to focus on more specific questions. Of course, each of these specific questions and their respective answers are just pieces in a much larger puzzle; one such piece is second language processing, an issue to which I will now turn my attention Processing David Birdsong offers a concise statement of what might be involved in a discussion of the differences between L1 and L2 processing. [...]
Online readingwith our online reader
Content validatedby our reading committee