On the basis of Hindi-English Interlanguage data, Bhatt and Hancin-Bhatt (2002) advance the Structural Minimality hypothesis in which CP-domain categories are not licensed in early second language (L2) acquisition, and claim that this leads early L2 learners to misconstrue PPs in the immediate vicinity of the complementizer THAT introducing an embedded clause, as in (1):
(1) a. PP-COMP: Peter said in the car that he saw a dog.
b. COMP-PP: Peter said that in the car he saw a dog.
We show, however, that the Structural Minimality thesis is conceptually weak and the argumentation seeking to establish it flawed in three areas: experimental design, reporting of results, and interpretation.
Using data from Garcia (1998), we report asymmetries--in the construal of PPs immediately preceding the complementizer (1a) versus PPs immediately following the complementizer (1b)--that show that knowledge of the CP-domain is necessarily implicated in early L2 English. We argue that these asymmetries are rooted in the performance system, and we present additional supporting evidence from English-French Interlanguage that is, moreover, highly revealing of the precise, universal sentence processing mechanisms involved. Bhatt and Hancin-Bhatt's (2002) proposal, in contrast, simply fails to explain any of these asymmetries.