Broca's aphasia is associated with a single pattern of comprehension performance: a reply.

Over many years now, we have provided evidence that the cortical area associated with Broca’s aphasia sustains operations necessary for the analysis of syntactic constructions that contain displaced constituents. A massive body of empirical data— from comprehension, real-time processing, and grammaticality judgment—shows that Broca’s area supports mechanisms involved in the computation of transformational relations (cf. Zurif, 1995; Grodzinsky, 2000). This evidence has been challenged and debated—most recently in the pages of this journal. Thus in 1999 we published the results of a survey of comprehension scores for Broca’s patients conventionally selected via the Boston Diagnostic Aphasia Exam (Goodglass & Kaplan, 1972) and other like instruments (Grodzinsky, Piñango, Zurif, & Drai, 1999; GPZD henceforth). We included all the relevant data published between 1980 and 1996 of which we were aware. The picture that emerged from these data was very clear: The Broca’s patients as a group performed significantly above chance level in their comprehension of structures without displaced constituents—actives, subject relatives, and subject-clefts; in contrast, they were at chance level in their comprehension of transformationally derived structures (those with displaced phrases), i.e., on passives, object relatives, and object-clefts. For one particular instance of this generalization—the active/passive contrast—there exists a considerable body of experimental evidence, obtained at different times and in different laboratories. It is on this nar-