...it is very hard to understand: "seeing is irrational, inconsistent, and undependable" (p. 11). ...no single testable theory emerges; Elkins-in this respect, unlike Gombrich-is interested less in proposing a quasi-scientific account of pictures than in providing the basis for a phenomenology of perception. Nothing is likely to be less productive than prolonged discussion by a reviewer who finds the assumptions of an author he admires absolutely implausible. Because I enjoy Crow's energy, originality, and persistence, I can only report my absolute inability to explain or understand why he makes this obviously futile attempt to divide artworks, or art journals, into those that are good because they are politically critical, and those others that, lacking that political meaning, are not good. Some students are taught to feel "free to pick and choose from the past at random," others "to parody tradition"; "most postmodern artists of both kinds . .are perfect products of art teaching in universities" (p. 47). Because students arrive at college with their minds filled with recycled images, survey courses in art history, de Duve argues, should "with a zest of humour" be "deliberately advertised as package deals in cultural tourism" (p. 54).