The variability debate: More statistics, more linguistics

We respond to critical comments and consider alternative statistical and syntactic analyses of our target paper which analyzed comprehension scores of Broca's aphasic patients from multiple sentence types in many languages, and showed that Movement but not Complexity or Mood are factors in the receptive deficit of these patients. Specifically, we do the following: (a) We show how group analyses such as ours are valid and critically important, and then discuss apparent statistical discrepancies between our analysis and others'. (b) We provide new syntactic arguments in support of our decision to categorize passive sentences in German and Dutch as -Movement in the context of Broca's aphasia. These arguments serve to underscore an intriguing correlation between German/Dutch and English: On the one hand, Scope Freezing is found in the former where the latter allows scope ambiguity. On the other hand, Broca's aphasic patients successfully comprehend German/Dutch passive, but fail in English. (c) We reanalyze new data from Dutch and Italian passive, which point to new and potentially interesting cross-linguistic differences. Our current analyses are based on an addition of raw scores from 62 new patients to the existing data base of 69 Broca's aphasic patients. We conclude that while aphasic performance is indeed variable, the group results have distinct statistical and linguistic structure. (c) 2005 Elsevier Inc. All rights reserved.