categories of science: its metaphysical disproof must be the demonstration of the adequacy of such scientific categories. In the words of Mr Shadworth Hodgson: "Either liberty is true, and then the categories are insufficient; or the categories are sufficient, and then liberty is a delusion." Such a determination of the sufficiency or insufficiency of scientific categories is the business of philosophy, as universal critic. A negative, as well as a positive, vindication of freedom is therefore possible the former by the condemnation of the categories of science as insufficient, the latter by the provision of higher and sufficient categories for its explanation. Even if such higher categories should not be forthcoming, and we should find ourselves unable to formulate a theory of freedom, or to categorise the moral life, we might still vindicate its possibility.

That the problem of freedom is ultimately a metaphysical one, is indicated by the fact that all deterministic theories base themselves, either explicitly or implicitly, upon a definite metaphysic. The denial of individual freedom is, for instance, the obvious corollary of such a pantheistic metaphysic as Spinoza's. Human personality being resolved into the all-comprehending divine Nature, from the necessity of which all things, without exception, follow, man's conception of his freedom, and of his resulting importance as an imperium in imperio, is explained away as an illusion of his ignorance, destined to disappear in an "adequate" knowledge of the universe. The consequence is strictly logical. If I am not a person, but merely an aspect or expression of the universe or God, I cannot be free. The life of the universe is mine also: freedom can be predicated, in such a system, of God alone, and even of him in no moral sense. Materialism, again, carries with it the same ethical consequence. If matter is everything, and spirit merely its last and most complex manifestation, once more freedom is an illusion. Freedom means spiritual independence; and if spirit is

the mere product of matter, its life cannot in the end escape the bondage of material law. The evolutional metaphysic, whether of the biological or of the mechanical type, also obviously binds its adherents to the denial of freedom. Moral life is interpreted either as a series of adjustments of the individual to his environment, or as a series of balancings of equilibrium. In neither case is room left for freedom, or a new beginning.

In such cases as those just indicated, the connection of the interpretation of human life with the general metaphysical theory is obvious enough. The connection, though not less obvious, has not been so generally remarked, in the case of the 'psychological' theory of determinism. This theory has been chiefly studied in the form given to it by Mill, and in that form the parallel between the metaphysical sensationalism and the ethical determinism is easily detected. The theory was originally stated, however, by Hume, and its logical dependence upon his philosophical empiricism or sensationalism is no less evident. If I am resolvable into the series of my conscious states; if I am merely the bundle or mass of sensations and appetites, desires, affections, and passions which constitute my experience; if, in short, my existence is entirely phenomenal,—then the phenomena which are 'me' can be accounted for, or refunded into their antecedents, like any other phenomena which are animals or things.

Here, then, emerges the sole possibility of a metaphysical vindication of freedom-namely, in another than the Humian, empirical, or 'psychological' account of the moral person or self. The nature of the self is a metaphysical question, and must be investigated as such; it is not to be taken for granted on the empirical or sensationalistic side. There is another alternative account, the transcendental or idealistic-namely, that the self, so far from being equivalent to the sum of its particular experiences or feelings, is their permanent subject and presupposition.

Thus the central problem of morality is seen to be, like the central problem of knowledge, the nature and function of the self. We have to choose between an empirical and a transcendental solution of both problems. If, on the one hand, the self is resolvable into its phenomenal states, if these exhaust its nature, the case for freedom is lost these states determine, and are determined by, one another in the unbroken nexus of antecedent and consequent. If, on the other hand, such a resolution of the self into its successive experiences is impossible, if moral experience presupposes at each stage the presence and operation of a permanent self, the case for freedom is made good.

:

6. The transcendental solution. That the latter, and not the former, is the true statement of the case, has, I think, been finally proved by the transcendental analysis of experience. It is still possible, of course, to rest in the scientific or psychological view of moral activity; one may not be prepared to adopt the transcendental standpoint, and may fall back upon the psychological or empirical view, as more in accordance with commonsense. Moral, like intellectual scepticism, and even agnosticism, are still, even after Kant and Hegel, intelligible attitudes of thought. But, unless it is shown that the scientific or psychological is also the final and adequate, or metaphysical, view; unless, that is, the whole self is resolved into its several states or its experience,-freedom is not disproved. Now such an empirical resolution of the self is as impossible in the moral as in the intellectual sphere; the phenomenal or empirical view, when offered as a metaphysic, is at once seen to be abstract and inadequate. To understand or think out the moral, equally with the intellectual life, we must regard the former as, like the latter, the product of the activity of the self. That activity is the heart and centre of the process, from which alone its real nature is recognised. Neither the moral nor the intellectual man can be re

It implies him; for, as

solved into his experience. experience, it is not a mere series or sum of states, but the gathering up of these in the continuous and single life of an identical self. If determinism is to be established, all the elements of the action must be known and observed as its phenomenal factors; but the source of the action cannot be thus phenomenalised. Determinism gives a mere dissection or anatomy of the action. Under its analysis, the living whole of the action itself is dissolved into its dead elements; the constitutive synthetic principle of the ethical life is absent. That principle is the self, or moral personality, to which the action must be referred if we would see it as a whole and from within. Motive, circumstances, temperament, character-the several parts of the determinist whole-all imply such an activity of the self, if they are to enter as living factors into the moral situation. And the self which is shown to be the source of this original and formative activity is thereby proved to be free. The self cannot be snared, any more than the spider, in the web of its own weaving.

The transcendental proof is essentially the same in the case of the moral and the intellectual life. It is the necessary complement, in either case, of the empirical or psychological view. For the previous question of metaphysics or first philosophy' is: How is experience itself possible? Experience, not being self-explanatory, requires to be explained. The empirical or psychological self is not ultimate, but only phenomenal; we must therefore ask: What is the self which manifests itself in these phenomena or states, and what is the rationale of its selfmanifestation ? The transcendental answer is, that the entire process of experience is a process of self-activity. The psychologist is concerned only with the empirical process; his business is to establish the true causal connections between the antecedent and consequent phenomena. But if, in an intellectual reference, it can be shown that the presupposition of knowledge is a constant

activity on the part of the self in the synthesis of the presentational data, that, without a unifying self, the ordered unity of experience would be impossible, it is no less evident that, without a similar synthetic activity on the part of a single central rational self, the unity of moral experience would also be impossible.1 The self weaves the web of its own experience, intellectual and moral. Out of wants, out of animal promptings, out of the provocations of sensibility, the self, by an activity of appropriation, constitutes motives or ends of its own activity. The entire process of motivation takes place within the circle of its being, and is conducted by itself. To press the psychological or empirical view, and to insist that the scientific interpretation of the moral life is the ultimate and sufficient interpretation of it, is to rest in a superficial view when a deeper view is possible and necessary. The empirical or phenomenal self may be regarded as the mere sum of motive- forces, of tendencies and counter-tendencies, whose resultant describes its life. But when we ask what a motive is, we find that it is nothing apart from the self; it is mine, I have made it. I am not merely the subject of tendencies, or the permanent deposit of tendency. I am the theatre of the entire process; it goes on within me.

Hence the well-marked limits of psychological explanation. The life of man, which is in its essence a personal life, is regarded by psychology as an impersonal stream of thought, a series of phenomenal states of consciousness. But metaphysics must correct the abstractness of psychology, as it corrects the abstractness of science generally, and must re- view the moral life from its personal centre-from the standpoint of that selfhood which, as unifying principle, is not to be phenomenalised, because, without its constant operation, there would be

1 The parallel between the intellectual and the moral activity of the self is strikingly enforced by Green, Prolegomena to Ethics, bk. ii., and by Professor Laurie, in his companion volumes, Metaphysica and Ethica.