From Academic Kids
Representationalism, or the representational theory of perception, is a philosophical doctrine that in any act of perception, the immediate (direct) object of perception is a sense-datum that represents an external object, which is the mediate (indirect) object of perception.
Two 17th century philosophers, René Descartes, and John Locke most prominently advocated this theory. The term they used was not "sense-datum" but "idea." This article does not discuss any differences in meaning that these terms might have. "Idea" as used in the theory of perception is a technical term, meaning roughly the same thing as sense-datum. Representationalism is one of the key assumptions of cognitivism in psychology.
Representationalism asserts that sense-data represent external objects -- physical objects, properties, and events. But this immediately raises a question: How well do sense-data represent external objects, properties, and events? At least sometimes, they do not represent them at all well. It is often the case that our perceptions do not correlate at all well with physical reality and this aspect of representationalism has led to psychologists questioning such things as police identity parades.
Dreams and imaginings can be considered representations in a way analogous to perceptions, perhaps, as recent fMRI studies have shown, using similar areas of the brain.
A problem with representationalism is that if it assumes that something in the brain, described as a homunculus, is viewing the perception, this suggests that some physical effect or phenomenon other than simple data flow and information processing must be involved in perception. This was not an issue for the rationalist philosophers such as Descartes, since dualism held that there is indeed a 'homunculus' in the form of the mind. For those who doubt dualism, explaining precisely what it is that sees the representation is problematic. But if Representationalism is thought of as an explanation of how we indeed see, then it falls foul of the homunculus fallacy.
A further difficulty is that, since we only have knowledge of the represetnations of our perceptions, how is it possible to show that they resemble in any significant way the objects to which they are supposed to correspond?
A final difficulty arises when attempting to explain reference from a representational viewpoint. If I say "I see the Eiffel Tower" at a time when I am indeed looking at the Eiffel Tower, to what does the term "Eiffel Tower" refer? One might wish to say it refers to the Eiffel Tower, but in the representational account we do not really see the tower, presumably the reference is to our sense experience. But this would mean that when I refer to the Eiffel Tower, I am referring to my sense experience; but when you refer to the Tower, you are referring to your sense experience. Therefore when we each refer to the Eiffel Tower, we are not referring to the same thing - an apparent absurdity.