To recognize something is to respond to it in a way that distinguishes it from other things; to recognize is to tell apart. But differential response cannot be the whole story, for two deeply related reasons. First, what is recognized is always some determinate item, feature, or characteristic of the confronted situation, whereas a given response can equally well be taken as a response to any of several distinct things. Second, recognition, unlike response, is a normative notion: it is possible to misrecognize something, to get it wrong, whereas a response is just whatever response it is to whatever is there. These are related because: only insofar as something determinate is supposed to be recognized, can there be an issue of recognizing it rightly or wrongly; and it is only as that which determines rightness or wrongness that the object of recognition is determinate.-John Haugeland, "Objective Perception" (HT 272)
The Problem of Perceptual RepresentationImagine that John is scouring the Arizona desert for a scorpion, in order to encase it in resin for the aiguillette on his bolo tie. As he looks under a small rock, he sees something that looks like a scorpion, swinging his net down on it. But it is not a scorpion. It is, in fact, a plastic scorpion left there by a mischievous graduate student. In frustration, he sits down on a large rock, thinking he might just give up wearing bolo ties altogether. As he does so, he suddenly feels a sharp pain in his lower regions, realizing that he has found what he was looking for, albeit unintentionally. Consider John's visual experience as he looks under the small rock. It makes sense to assess it as accurate or inaccurate, and it turns out that it is inaccurate. Contrast this with John's pain experience as he sits down on the large rock. There are many things we could say about this experience, but it would not make sense to assess it as accurate or inaccurate.