It is a common view in cognitive psychology that there is a fundamental difference between what may be called descriptive information, on the one hand, and depictive information, on the other. While the first kind of information is — ideally spoken — non-pictorial and usually equated with the content of a proposition, the second kind of information is pictorial by defmition and accordingly equated with the content of a mental image. Granting the correctness of this distinction, cognitive scientists differ on the role played by mental images in epistemic processes. One faction, represented by the writings of Zenon Pylyshyn, asserts that mental images are merely an epiphenomenon of a more general and abstract processing system, and that they are in principle dispensable since they do not have any truth value.¹ Another faction, represented by the writings of Stephen Kosslyn, favors the opposite view, maintaining that mental images are indispensable since they play, in spite of their lack of truth value, a fundamental role in human thinking.² As Arthur I. Miller has proposed in his book Imagery in Scientific Thought (1986), these two factions may be labeled as the "antiimagist" and the "pro-imagist" view, respectively.³