only some of them are also good themselves; as an example of those which are not, I have already mentioned inanimate beautiful objects. As examples of the second class of things to which he certainly ascribes more value than they possess, we may take pleasure and knowledge. This error is even more grave than the last, since it compels him to ascribe value to things which are not merely indifferent but positively bad; and it follows not from his definition of “good” but from a principle which I have not yet mentioned. This principle is that no one thing is better than another, unless it contains a greater number of good or a less number of bad parts. In this book, indeed, Brentano is even inclined to maintain that all judgments of intrinsic superiority are purely analytic—that “better” merely means “having a greater number of good parts.” But the translator tells us, in the Appendix, that he has now definitely abandoned this view: he now recognizes “that it is by no means evident from analysis that one good plus another is preferable to each of these goods taken singly” (p. 122, note). We are not told, however, that Brentano has also abandoned the views with which we are now concerned—the views that, as a matter of fact, not only are two good things always better than one of them, but also that no one thing is better than another unless it contains more good parts. From the second of these principles it would follow that all good things, which are not composed of good parts, are equally good; and also that any quantity of pleasure is one such good thing. For it is certainly true that a greater pleasure in a very beautiful thing is sometimes better than a less pleasure in the same; and, this being so, it follows from Brentano’s principle, that what differentiates the former from the latter—namely the excess of pleasure—is good in itself. And similarly it is certain that one state may be better than another, where it only differs from that other in containing more knowledge; and hence it would also follow that some knowledge, at all events, was good in itself. But the principle from which these results follow is certainly false: that it is an error, and a grave one, may be shown by taking a case in which it would prove to be good what is in fact positively bad in a high degree. For Brentano is bound to hold that a very great pleasure in what is wholly bad, but not very bad, is not merely better than a less pleasure in the same, but positively good. He must even, if the bad thing be purely imaginary, pronounce it to be better than a less pleasure in an imaginary thing which is wholly good. In fact,
