As this study purposes to be constructive no less than sceptical, and as the claim to unite construction with scepticism savours somewhat of paradox, I think it will be well to offer what preliminary explanation is possible by outlining the general character of the view of experience which the text aims at establishing.

To consider, first, the sceptical aspect of my thesis. My starting-point is Bradley's epistemology. I state and defend the central tenet of that epistemology, and endeavour to show, at the same time, that the implications of the Bradleian Absolute in the way of metaphysical scepticism are a great deal more far-reaching than Bradley is himself ready to admit. But although the argument from the nature and demands of knowledge is my starting-point, what I am especially concerned to maintain is that the essential result of the epistemological argument (i.e. that the ultimate nature of reality is 'beyond knowledge') is supported in the most striking way by considerations drawn from aspects of experience other than the cognitive. In pursuance of this task, I take up in turn three basic forms of our experience - the experience of 'self-activity,' 'moral' experience, and 'religious' experience - and I try to show that in the case of each the experiencing subject is compelled by the very nature of the experience to make affirmations which imply the conviction that reality is 'beyond knowledge.' If these affirmations are really ineradicable from these forms of experience, and if (as I expressly argue in regard to moral experience and the experience of self-activity) these forms of experience are really basic in our nature, then we seem entitled to say that the assertion of the 'supra-rational' character of reality is an assertion which 'our nature' obliges us to make.

Metaphysical scepticism is thus, in my view, the converging point of a variety of independent lines of thought. We are guided to it by the evidence at once of cognitive experience, of the experience of self-activity, of moral experience, and of religious experience. The path of epis-temology is but one of many paths all leading to the same terminus.

So much for the 'sceptical' aspect. But what room does this leave for 'construction'? None, it is obvious enough, for a constructive philosophy which aims at knowing the ultimate nature of things. But for a constructive philosophy with less exalted, though still, as I believe, vitally important aims, there is room. The third chapter, that on 'Noumenal and Phenomenal Truth,' may be taken as an attempt to bridge the gulf between scepticism and construction. When once we realise (it is there urged) the full implications of the fact that 'Noumenal' Truth, Truth in its ideal or ultimate form, can have absolutely no positive or concrete significance for the finite intellect, the concept of 'Phenomenal' Truth, Truth in that degraded, but by intellect not positively tran-scendible, form of itself in which it does have concrete significance for the finite intellect, takes on a real importance. For this 'Phenomenal' Truth has its own standards, and we, as finite intellects, cannot afford to view with indifference the measure in which our thinking attains to, or falls short of, these standards. It is with the endeavour after 'Phenomenal' Truth only that constructive philosophy can, as it seems to me, concern itself; and its highest achievement lies in the articulation of what I have called 'final phenomenal truths.' A proposition has 'final phenomenal truth' if it be such that it is manifestly insusceptible of revision or modification under the conditions of finite experience: if it be 'intellectually incorrigible' (to use Bradley's phrase), though not (for this it could not be without becoming a 'noumenal' truth) 'intellectually satisfying.' It is possible, I believe, to arrive at a number of these propositions which, while not fulfilling the full demands of 'knowledge,' are yet such as cannot be significantly questioned under the conditions of finite experience. To this category belong, it is contended, the 'ineradicable affirmations' already alluded to (the explicit formulation of which, however, it would be premature to attempt here). And while these 'final phenomenal truths' must always remain, in one sense, 'illusions,' they are certainly not to be classed with 'mere' illusions. There is obviously a vast difference in status between illusions corrigible by fuller and better thinking, and illusions which belong (as these do) to the abiding framework of human experience. The term 'illusion,' as applied to the latter class, can retain little of its customary dyslogistic significance. It will seem to retain progressively less, I should hold, in proportion as we achieve the reorientation of attitude to ' man's place in the cosmos' which the supra-rationalist premises make necessary.

This bald outline must suffice by way of indicating the central purport of my thesis. I am only too acutely conscious that that thesis involves (as will already be in part apparent) subscription to a goodly number of tenets now commonly deemed unrespectable. The temper of modern philosophy is not congenial to the reintroduction of an Absolute - and the prefixing of the qualification 'supra-rational' will probably seem to aggravate rather than to extenuate the crime. It would be futile to attempt to anticipate here the justification in the body of the work of these and other heresies, but I should like to say just this with regard to the central 'heresy' that the ultimate nature of reality is 'beyond knowledge.' There is a tendency to stigmatise doctrines of this kind as 'escapist' - as though they stood for a somewhat craven retreat from the admittedly tremendous difficulties of the metaphysical problem. This attitude seems, on the face of it, radically unfair. The mathematician is not suspected of intellectual cowardice if he declines to wrestle with the 'difficulties ' involved in squaring the circle. It is recognised that he has satisfied himself, after due consideration, that the 'difficulties' are just not surmountable, that the problem is in principle insoluble. Ought it not to be accepted, as at least a provisional hypothesis, that the metaphysician who pronounces ultimate reality to be 'unknowable' is in precisely the same case?