Another instance of the way in which Mr. Sidgwick evolves an incongruity which he considers fundamental, out of what I should have thought he would see is a {241} defective expression, I will give in his own words. Speaking of a certain view of mine, he says:― “He tells us that ‘logic . . . contemplates in its propositions certain connexions predicated, which are necessarily involved with certain other connexions given: _regarding all these connexions as existing in the non-ego_—not, it may be, under the form in which we know them, but in some form.’ But in § 473, where Mr. Spencer illustrates by a diagram his ‘Transfigured Realism,’ the view seems to be this: although we cannot say that the real non-ego resembles our notion of it in ‘its elements, relations, or laws,’ we can say that ‘a change in the objective reality causes in the subjective state a change exactly answering to it—so answering as to _constitute a cognition of it_.’ Here the ‘something beyond consciousness’ is no longer said to be unknown, as its effect in consciousness ‘constitutes a cognition of it.’” This apparent inconsistency, marked by the italics, would not have existed if, instead of “a cognition of it,” I had said, as I ought to have said, “_what we call_ a cognition of it”—that is, a relative cognition as distinguished from an absolute cognition. In ordinary language we speak of as cognitions, those connexions in thought which so guide us in our dealings with things, that actual experience verifies ideal anticipation: marking off, by opposed words, those connexions in thought which _mis_-guide us. The difference between accepting a cognition as relatively true and accepting it as absolutely true, will be clearly shown by an illustration. There is no direct resemblance whatever between the sizes, forms, colours, and arrangements, of the figures in an account-book, and the moneys or goods, debts or credits, represented by them; and yet the forms and arrangements of the written symbols, are such as answer in a perfectly-exact way to stocks of various commodities and to various kinds of transactions. Hence we say, figuratively, that the account-book will “tell us” all about these stocks and transactions. Similarly, the diagram Mr. Sidgwick refers to, suggests a way in which symbols, registered in us by objects, may have forms and arrangements wholly unlike their objective causes and the _nexus_ among those causes, while yet they are so related as to guide us correctly in our transactions {242} with those objective causes, and, _in that sense_, constitute cognitions of them; though they no more constitute cognitions in the absolute sense, than do the guiding symbols in the account-book constitute cognitions of the things to which they refer. So repeatedly is this view implied throughout the _Principles of Psychology_, that I am surprised to find a laxity of expression raising the suspicion that I entertain any other.