Picture naming requires early visual analysis, accessing stored structural
knowledge, semantic activation, and lexical retrieval. We tested the effect
of perceptual, lexical, and semantic variables on the performance of aphas
ics in picture naming and assessed prevalence of natural categories vs arti
fact dissociations. Forty-nine aphasics were asked to name CO pictures, fro
m three natural (animals, fruits, and vegetables) and three artificial cate
gories (tools, furniture, and vehicles). For each item visual (drawing comp
lexity, image agreement), semantic (prototypicality, concept familiarity) a
nd lexical variables (word frequency, name agreement) were available. The e
ffect of these variables showed individual differences: altogether, visual
complexity had little influence, whereas lexical and semantic variables wer
e more influential. Name agreement was most important, followed by word fre
quency. On a multiple single case analysis 10 patients (20%) showed a natur
al/artificial category dissociation. Five of the six subjects Faring better
with artifacts were males, and all of four patients faring better with nat
ural categories were females, interpretations of this finding are discussed
. (C) 2001 Academic Press.