In a recent controversial article, "Feeling and Thinking: Preferences Need No Inferences" (l980), R. B. Zajonc argues in support of the independence of affect and cognition. Examination of the structure and assumptions of Zajonc's arguments suggests that they do not support the view that affect is non-cognitive. Zajonc appears to leap from the conclusion that people are able to make affective judgments without being able to make certain prior conscious cognitive judgments, to the conclusion that affects have no cognitive aspects, or only very minimal or "primitive" cognitive aspects. Other ways of conceptualizing the cognitive aspects of affect are possible. Zajonc introduces a new ontological category of features, that of "preferenda," in an attempt to account for the features of the stimulus to which people respond in affective reactions. This suggests that Zajonc has here simply re-invented "intentional objects," and that their introduction is not only unnecessary, but may even confuse certain central questions that we may wish to ask about the relevant stimulus features for affective responses. Zajonc's attempts to distinguish affect from cognition on the basis of its being "involuntary" and "primary in phylogeny," are also subject to criticism. (Author/JAC)