Semantic vs. syntactic subject: a comprehension test based on control constructions.

The test presented here is made of 32 + 32 complex sentences that are submitted with two different tasks: one to prove comprehension of the event denoted by their embedded clause and the other to prove the corret interpretation of this clause, whose null subject is syntactically controlled. The tasks have been administered to four stabilized aphasic patients of medium severity to see whether their comprehension could be considered asyntactic with respect to the specific Control module of Universal Grammar (Chomsky, 1981).