The speech of two fluent and three non-fluent aphasic patients was analysed using the quantitative method devised by Saffran, Berndt and Schwartz et al. (Brain and Language37, 440–479, 1989) [6]. The differences between agrammatic and non-agrammatic aphasic speech were broadly replicated, in that the non-agrammatic non-fluent patients were impaired only on measures of syntactic complexity and
... [Show full abstract] well-formedness. The fluent patients' sentence production was indistinguishable from that of the non-fluent non-agrammatics, except with respect to fluency itself. All five patients were retested 1–2 years after the first sample. The two patients who had improved in the quantitative analysis showed a similar improvement in picture naming. For one patient at least, an increased availability of low-frequency nouns appeared to be underlying this improvement.