To illustrate here by a familiar case, when we speak of things that are solely matters of personal preference, such as the pleasure of a sleigh-ride, the taste of olives, or the comfort of a given room, and when we only try to tell how these things appear to us, then plainly our judgments, if sincere, cannot be in error. As these things are to us, so they are. We are their measure. To doubt our truthfulness in these cases is to doubt after the fashion of the student who wondered whether the star that the astronomers call Uranus may not be something else after all, and not really Uranus. Surely science does not progress very far or run into great danger of error so long as it employs itself in discovering such occult mysteries as the names of the stars. But our present question is, How do judgments that can be and that are erroneous differ in nature from these that cannot be erroneous? If astronomers would be equally right in case they should agree to call Uranus Humpty Dumpty, why are not all judgments equally favored? Since the judgment chooses its own object, and has it only in so far as it chooses it, how can it be in that partial relation to its object which is implied in the supposition of an erroneous assertion?
Yet again, to illustrate the difficulty in another aspect, we can note that not only is error impossible about the perfectly well-known, but that error is equally impossible, save in the form of direct self-contradiction, about what is absolutely unknown. Spite of the religious awe of some people in presence of the Unknowable, it is safe to say, somewhat
