tion made harmless by division, since A is c and D is b. This is the general principle, and I will attempt here to apply it in particular. Let us suppose the reality to be X (a b c d e f g . . . ), and that we are able only to get partial views of this reality. Let us first take such a view as “X (a b) is b.” This (rightly or wrongly) we should probably call a true view. For the content b does plainly belong to the subject; and, further, the appearance also—in other words, the separation of b in the predicate—can partly be explained. For, answering to this separation, we postulate now another adjective in the subject; let us call it β. The “thatness,” the psychical existence of the predicate, which at first was neglected, has now also itself been included in the subject. We may hence write the subject as X (a b β); and in this way we seem to avoid contradiction. Let us go further on the same line, and, having dealt with a truth, pass next to an error. Take the subject once more as X (a b c d e . . . ), and let us now say “X (a b) is d.” This is false, because d is not present in the subject, and so we have a collision. But the collision is resolved if we take the subject, not as mere X (a b), but more widely as X (a b c d). In this case the predicate d becomes applicable. Thus the error consisted in the reference of d to a b; as it might have consisted in like manner in the reference of a b to c, or again of c to d. All of these exist in the subject, and the reality possesses with each both its “what” and its “that.” But not content with a provisional separation of these indissoluble aspects, not satisfied (as in true appearance) to have aα, bβ, and dδ—forms which may typify distinctions that bring no discord into the qualities—we have gone on further into error. We have not only loosened “what” from “that,” and so have made appearance; but we have in each case then bestowed the “what” on a wrong quality within the real subject.
