Let us now examine one of these radical positions, say, the sceptical one first. According to this, since error is a fact, we may as well treat every perception as erroneous, only undetected so far and so accepted as true. But if we accept error as a fact we must also accept truth as such, for an error cannot be known as one unless we know something as true. When one says one mistook a rope for a snake one must (...) know the rope truly, otherwise there can be no mistake. One may think erroneously, "I am apparently seeing a rope but it may be that I am only imagining it, just as I imagined the snake; I may be mistaking something else for a rope." But then one must also see that as the rope is thus doubted, the unreality of the snake must also be doubted, for it is against the rope as real that the snake becomes unreal. And now one would have to say, "Perhaps I am imagining this rope which might really be the snake I thought I imagine." But in this state of oscillation one cannot pronounce any judgment at all, and so cannot assert anything about the truth and falsity of any perception. The original perception proving false leads one, through analogy, to doubt the latter perception. But if there is a doubt regarding the falsity of the original perception, there cannot be a legitimate doubt regarding the validity of the later one. More generally we may say that if one doubts one's perceptions of certain objects on the ground that some perceptions regarding these had proved false in the past, this doubt logically involves the possibility of those perceptions, rejected as false being true. So that though one may psychologically take up the attitude of doubt with respect to every perception, this attitude has no rational backing. If, however, one doubts certain objects of perception because some other perceptions e.g., snakes) proved false, one has only a weak analogy and is, moreover, forced to doubt all perceptions, including those which proved certain perceptions false, thereby starting the original doubt. But then the reasoning given above which will dispel this doubt will have to be applied. So that this doubt cannot be rationally justified. The argument that since certain perceptions have proved false therefore every one of them may prove so is, after all, a very weak one. A perception is false only when it is falsified by a true one, and we cannot doubt a perception unless it is so falsified. (shrink)