ABSTRACT
It is relatively easy to quantify performance on tests of language. A more difficult task is translating a pattern of performance into conclusions regarding the linguistic abilities/mechanisms that underlie that performance. For example, poor performance can be caused by many things: deviant or missing knowledge, parsing difficulty, memory overload, and so on. In the adult processing literature, it is taken for granted that adults who are unable to interpret tripleembedded clauses or who misinterpret relative clauses of some types in speeded tasks do not do so because of impaired grammatical knowledge. However, when some population of interest has a cognitive or linguistic disorder, they cannot be granted the same benefit of the doubt because the integrity of their linguistic knowledge may itself be the object of study.
