ABSTRACT
It is in the context of investigating patterns of semantic loss that categoryspecific impairments in AD were first reported. Chertkow and Bub (1990a) investigated the semantic knowledge of a group of ten AD subjects on items from a variety of taxonomic categories by means of probe questions requiring yes-no answers. Although the overall accuracy rate was 87.3%, they noted that the subjects were considerably worse for the categories of animals, fruits and vegetables (mean accuracy = 68.2%) than for non-biological categories such as vehicles, furniture, or clothing. Since then, several researchers have specifically tested whether the semantic loss in AD is homogeneous across semantic categories, and the deficit most frequently reported involves poorer semantic knowledge about biological objects than about nonbiological objects. More recently, interest has also turned to investigating whether there are imbalances in loss of knowledge with regards object words versus action words, and abstract versus concrete knowledge. We will first consider in detail the biological-non-biological data and briefly consider the other contrasts as they are relevant for the discussion.
