Descartes’s Dream Argument is a classic example of what philosophy of perception calls the “argument from illusion.” It operates on the claim that, to a perceiver, genuine perceptual experiences cannot be distinguished from illusory ones. Because of this, two very different situations confront the common sense realist. For naive realism, normal cases of perception are a relation between a perceiver and an external object. But with dreams and hallucinations, there is no physical object. Any explanation of perceptual error for the common sense realist relies solely on the perceiver’s mental state.
Because of his desire to secure sense knowledge on a foundation of certitude, Descartes submitted that “ideas” act as intermediaries between a perceiving consciousness and a perceived object. These ideas are mental representations that may or may not be causally connected to a physical object that exists independently of mind. In both cases, whether actually perceiving a tree or dreaming that one perceives it, the person has the idea “tree.” While the connection between idea and external object remains in doubt, the connection between perceiver and idea is clear, distinct, infallible and beyond doubt. The existence and nature of these ideas are independent of the nature of the perceptual state of the perceiver. Descartes’s realism has an advantage over naive realism to the extent that perceptual error occurs in the connection between “idea” and physical object. Without an intermediary “idea,” the naive realist has no explanation.
It is difficult for common sense realism to account for interpretation because it emphasizes the role of the physical object. There is no sign of what, if any, active role the perceiver plays in dealing with these objects. The common sense realist account of perception says nothing about whether different perceivers perceive the same physical object in different ways. An adequate philosophical theory of perception, which explains both perceptual error and interpretation, cannot help but become some sort of critical realism.
 Jonathan Dancy, An Introduction to Contemporary Epistemology (Oxford: Basil Blackwell, 1985), 153.
 Ibid., 151.