摘要：Evidence from language-unimpaired populations shows that sub-lexical units such as suffixes are routinely parsed from their carrier words at an early stage and that semantic information may be available at the lexical level. Patients with agrammatic Broca’s aphasia show lexical processing deficits, i.e. deficits in lexical access and/or lexical integration, but the exact nature of these deficits is yet unknown. This study aimed at addressing two questions regarding how Broca’s aphasics process written language. First, is morphological parsing in Broca’s aphasia subject to the same rules as in normal processing, i.e., are inflected and derived forms parsed in the same manner when matched for suffix and stem length and for frequency? Second, is lexical-semantic information available early on during lexical access or during lexical integration? We addressed these questions in two realtime computerized tasks by comparing Broca’s aphasics and normal controls in Greek, a language that allows perfect matching of inflected and derived words for length and frequency at both the stem and the whole-word levels.
