Accordingly, he sought to transfer all those processes of com bining, discriminating, judging, which Locke had spoken of as faculties of the mind, to the perceptions themselves, and to regard these so-called faculties as on the one hand modes of grouping among the ideas and on the other as particular ways of having ideas, the differences of which are due to variations of an indefinable element of feeling, giving rise to what he vaguely called "belief." All those references to external existents, which perceptions seem to carry with them, are to be ascribed to this mysterious factor of "belief,"—mysterious because it obviously implies that very distinction between mind and its perceptions which Hume was refusing to recognize.
Since the constituents of experience are discrete units, it fol lowed that the relations between them must be wholly external to the units themselves, and can be only relations of order, the order in which the units come together and succeed one another. From this point of view, it is difficult to see why Hume should have repeated the familiar distinction between two types of knowledge,—relations of ideas and matters of fact,—because one would have supposed that the only possible type of knowledge would have been the latter. And, in truth, Hume does not admit any fundamental difference. Necessity in thought signifies for him non-contradiction.
Accordingly, a relation between ideas must be either an actual occurrence, and then its non-occurrence would involve no contra diction, the characteristic, namely, of matters of fact, or it must be restricted to the content of the single idea, that is to say, the proposition expressing it must be purely analytical. Although in the well-known passage of the Inquiry, he instances mathe matical propositions as coming under the head of relations of ideas, he did not really regard them either as analytical or as necessary and universal. They have, he was ready to assert, no other basis than experience ; it is just as impossible to extract a mathematical proposition, a proposition implying a relation, from the content of a single idea, as it is to extract from any one perception taken by itself the idea of its cause or effect. And in reference to the idea of cause, the only ground of explanation which experience can furnish is that of custom or habit. Im pressions come to us in certain orders of sequence or coexistence, and this natural conjunction, frequently recurring, gives rise to a particular feeling, which occasions the belief that they are neces sarily connected. Similarly with reference to our ideas of perma nent things and of the individual self. The former is due to the
resemblance of recurring impressions which we take to be im pressions of the same object, although, in truth, each is an inde pendent fact. The latter arises from and is based upon the ease and rapidity with which each new impression or idea is conjoined with the train of ideas already present.
The value of Hume's work is largely dependent upon the thoroughness with which he attempted to work out a theory of knowledge from the strictly empirical point of view. Given only isolated mental states, he has done all that it is possible to do in attempting to show how the knowledge which we appear to have of things and their relations comes about. And in the end he was constrained to confess that the attempt had been unsuc cessful, and names with penetrating sagacity the exact reason of the failure. His effort terminated in a position of thorough scepticism. It may be that the beliefs to which experience leads correspond to fact ; but, if they do, there are no possible means by which we can logically show that such is the case.
c. The Critical empirical and rationalistic theo ries of knowledge, while diverging widely in fundamental prin ciples, had been brought to a stand before similar problems. The rationalistic method had terminated in a complete severance of thought from things. Throughout the movement from Descartes to Leibniz it is discernible that the final conception of perfect knowledge was destined to become that of a mere system of iso lated notions, each possessed no doubt of inner relations, but deprived of any significance as interpretative of real fact. On the other hand, the empirical method in its developed form as presented by Hume, found itself in the difficulty of offering, in terms of its main conception, any explanation of knowledge at all. Each isolated idea might either be called real or regarded as indicative of what was real, as by Hume and Locke respec tively, but in either case the isolation seemed to make all relation of ideas, all systematic insight into existent reality, impossible.
Indeed, the empirical theory easily led to a final view as to the kinds of knowledge which even in words is almost identical with that reached by rationalism. Each isolated idea, seeing that it exists, has at least the formal mark of inner consistency. There appear to be always possible with regard to it those judgments which do not go beyond mere consistency,—analytical judg ments. But any judgments whose import is relation among real facts must be pronounced to be impossible, or be designated by other terms than that of knowledge.