Wendy Best, Astrid Schröder, Ruth Herbert

• This paper presents a study of PH, a woman with aphasia, who shows a robust impairment in naming pictures of non-living relative to living things. Un-timed investigations of feature knowledge show similar performance across categories suggesting that, as in previous studies, the category effect may be arising at a post-semantic level. However, her performance on a timed feature verification task was slower for non-living than living things (relative to matched controls), in line with her naming. This suggests that the source of PH's category deficit is in fact semantic and that thorough investigation is necessary before claiming a post-semantic category specific deficit in word production. Finally, the results of an intervention study, which apparently eliminated the effect of semantic category on PH's naming, are reported.