direct realism the theory that perceiving is epistemically direct, unmediated by conscious or unconscious inference. Direct realism is distinguished, on the one hand, from indirect, or representative, realism, the view that perceptual awareness of material objects is mediated by an awareness of sensory representations, and, on the other hand, from forms of phenomenalism that identify material objects with states of mind. It might be thought that direct realism is incompatible with causal theories of perception. Such theories invoke causal chains leading from objects perceived (causes) to perceptual states of perceivers (effects). Since effects must be distinct from causes, the relation between an instance of perceiving and an object perceived, it would seem, cannot be direct. This, however, confuses epistemic directness with causal directness. A direct realist need only be committed to the former. In perceiving a tomato to be red, the content of my perceptual awareness is the tomato’s being red. I enter this state as a result of a complex causal process, perhaps. But my perception may be direct in the sense that it is unmediated by an awareness of a representational sensory state from which I am led to an awareness of the tomato. Perceptual error, and more particularly, hallucinations and illusions, are usually thought to pose special difficulties for direct realists. My hallucinating a red tomato, for instance, is not my being directly aware of a red tomato, since I may hallucinate the tomato even when none is present. Perhaps, then, my hallucinating a red tomato is partly a matter of my being directly aware of a round, red sensory representation. And if my awareness in this case is indistinguishable from my perception of an actual red tomato, why not suppose that I am aware of a sensory representation in the veridical case as well? A direct realist may respond by denying that hallucinations are in fact indistinguishable from veridical perceivings or by calling into question the claim that, if sensory representations are required to explain hallucinations, they need be postulated in the veridical case. See also PER- CEPTION , PHENOMENALIS. J.F.H.