This correlation, however, is not a real, objectively founded connection. We know, for instance, that in the thing commonly called by the name of "gold," a certain density, a certain colour, a certain specific weight, etc., appear regularly in conjunction ; but the necessity of this conjunction, the "ground" for the com bination of the various qualities in this, and no other, mode, is not understood. Even if there be such a ground in reality, it must be undiscoverable and unintelligible. (Locke, Essay on the Human Understanding.) The subjective nature of substance is emphasized even more by Hume.
The concept of substance, according to Hume (Treatise on Human Nature) , is not reducible to any sense impression ; it has, therefore, no positive significance for knowledge, but is merely a product of the imaginative power which combines what is fre quently together, into one idea and gives it one name. This criti cism of the concept of substance refers to the realm of external as well as of internal experience. The supposition that a material thing as such persists even when not perceived is denied ob jective significance; and the concept of "soul-substance" is re jected and the "I" declared a mere "bundle of perceptions." Kant.—Kant's Critique of Pure Reason takes up the results of Locke and Hume ; but it draws from them an entirely dif ferent epistemological conclusion. For Kant, too, substance is a pure concept of understanding; a form of connection established by pure thought. But this form of connection cannot be reduced to the empirical rules of habit and association ; it has universal and necessary validity; it is a synthesis a priori. The validity of this synthesis consists in the fact that it is the basis of experience itself. Without the idea of a Something persisting in the stream of appearances, it would be impossible to establish and to make in telligible that order of phenomena which we conceive under the name of "experience." The proof, the "transcendental deduction" of the concept of substance is, accordingly, presented by exhibiting it as a con stitutive condition of any possible experience. This implies, on the one hand, that it is indispensable for all experience, for all scien tific cognition of nature; on the other hand, that it is designed for use in experience only, that its validity does not refer to but to phenomena. The principle of sub stance, according to Kant, belongs to the "analogies of expe rience," i.e., to those principles on the basis of which alone it is possible to set a fixed, objective time-relation between phenomena and to conceive them as "nature" coherent in itself and ordered according to universal laws. Such an objectivation of time would be impossible without the category of substance. "The persistent is the substratum of the empirical notion of time itself, and it is on its basis alone that all temporal determination is possible."
For change does not affect time itself, but only its contents, only the concrete events which we conceive as following each other in time. Time itself does not change, but all change occurs in it as the "constant corollary of all Being of phenomena." However, the strictly "empty" time does not constitute any possible object of perception ; in order to think of time as constant, as "duration," we must, therefore, presuppose a persistent element within ap pearance itself and oppose it to all that is merely changeable.
The concept of something persistent is, accordingly, "the condi tion of the possibility of all synthetic unity of perceptions, i.e., of experience; and in proportion to this persistent element, all being and all change in time can be viewed only as a modus of the existence of that which remains and persists." (Kritik der reinen V ernunft, 2nd. ed. ; p. 225 seq.) Recent Epistemology.—The contrast between Hume's and Kant's conceptions of substance, reappears in the recent episte mology. The further pursuit of the road of Hume leads to em piricism and positivism ; the further pursuit of the road of Kant, to logical idealism. For positivist epistemology as developed in Avenarius' Kritik der reinen Erfahrung (1888-9o) and in the writings of Ernst Mach (Beitrage zur Analyse der Empfindungen, ; Die Mechanik in ihrer Entwicklung, 1883; Prinzipien der iirmelehre, 1896), the concept of substance has an essentially biological significance. It serves the "economy of thought" in asmuch as it is a means of correlating a multiplicity of experiences, and giving them one name.
But the unity of substance is a merely nominal unity. What we call material substance, is only "a relatively constant sum of sen sations of touch and light associated with sensations of space and time." This constancy can never be considered absolute, only relative, so that the identity of the "thing" is a mere function.
Critical idealism essentially agrees with this result, but dif fers from empiricism and positivism in the explanation which it gives. It, too, emphasizes the fact that, the further the scientific cognition of nature progresses, the more the concepts of things are replaced by concepts of relations. The concept of substance resolves into the concept of function. But the concept of function is not considered as the expression of a mere "togetherness" of experienced facts, but as a genuine achievement of thought. It is the original form of connection as such, and of experience itself. The concept of substance is to be replaced ultimately by that of invariance. (E. Cassirer, Substance and Function, 1923.) (E.Cr.)