Test stimuli are rated as less "good" when they follow very good context stimuli than when they are presented alone. This diminution in rating is called hedonic contrast. Contrast is attenuated if the context and the test stimuli are perceived as being in different categories. Because experts use as their basic-level categories what are the subordinate levels for novices, they will categorize when novices do not. Therefore, in the following studies, both experts and novices showed hedonic contrast when attractive context orchids preceded more neutral test orchids. However, only the novices showed hedonic contrast when attractive context irises preceded the test orchids. Novices viewed the irises and the orchids as "flowers" and therefore members of the same category, resulting in contrast. Experts.'however, viewed the irises and the orchids as being in different categories; therefore, hedonic contrast did not occur.