Recruiting comparative crosslinguistic evidence to address competing accounts of agrammatic aphasia

Several hypotheses have been advanced whose aim has been to provide a descr iptive generalization of comprehension patterns in agrammatic aphasia in te rms of current linguistic theory, most notably, the Trace-Deletion Hypothes is. The basic insight of these syntactic accounts of aphasia is that chains are disrupted. In this paper, we seek to confront the Trace-Deletion Hypot hesis (TDH) and one of its variants, the Double-Dependency Hypothesis (DDH) , with discriminating, crosslinguistic data. We adduce evidence that on rai sing constructions both hypotheses are able to derive Spanish agrammatic da ta correctly. However, neither the TDH nor the DDH are able to account for above-chance performance on SV or VS truncated passives. Finally, only the DDH explains the observed data on passive constructions in which a postverb al subject follows the by phrase (V-by phrase-S). The VS word order data ar e the critical cases because focusing simply on English would not allow the se structures to be tested and, in the case of the V-by phrase-S passive, b oth hypotheses make different predictions. While the data on raising constr uctions extend the range of both the TDH and the DDH, the VS data suggest t hat modifications are required. (C) 1999 Academic Press.