Friday, February 1, 2008
Dinsmore 2006
I was re-reading Dinsmore’s paper now and I think there is a very serious flaw in it: having found a 1.25 effect size he assumes that “UG does not fully operate in adult/adolescent L2 learning” (p. 80). However, when he talks about 3 types of access to UG (p. 57) he presumably includes the Full Transfer Full Access Hypothesis (Schwartz & Sprouse 1994, 1996, among many others) into the last group which he calls “UG is partially available” which completely goes against the tenets of the Full Access Full Transfer Hypothesis. Of course, there would be a significant statistical difference in the results obtained from studies with L1 and L2 learners. The problem is that the proponents of the Full Access Full Transfer Hypothesis do NOT think it’s because of access/no access to UG; it’s just because in case of L2 acquisition we have epistemologically different competence (what Schwartz 1987 calls ‘encyclopedic knowledge’). So Dinsmore did a great job but he totally misinterpreted the results.
Subscribe to:
Post Comments (Atom)
1 comment:
Alex-- Sorry, I said I would comment on your posting, but it's taken me a while...
Maybe Dinsmore is being a bit loose with the terminology, and combining under "no full access" the "full" notion and the "direct" notion. Sprouse and Schwartz's hypothesis is for full access but also full transfer, which means that UG is accessed indirectly through the L1.
In the end, I think the finding sticks and the interpretation is sound... but very limited. The mean difference in NS and NNS performance is large enough that we can see something is going on and UG is certainly not fully/directly available to NNS for all areas of UG. But that is not saying much, is it :-)
Thanks for your careful reading of the meta-analysis and your sharing of comments with the class!
Post a Comment