One last point. The discussion of the last two chapters has made it possible for me to set forth in a very few words my position with regard to the kind of criticism which has been levelled by Bradley against the 'reality' of the finite self (as, for example, in the well-known chapters on the self in Appearance and Reality).

The nerve of Bradley's argument is as follows. The test of the ultimately real is self-consistency. To nothing that fails to possess this character can we assign reality in the full metaphysical sense of the term. Now whatever by its very nature points beyond itself for the understanding of itself is assuredly not self-consistent. But this is so of the concept of finite self-hood. View the self as you may, in whatever aspect or from whatever angle, always you find in the self a necessary reference to something beyond itself which forbids us to find in it the principle of the ultimately real. You may, indeed, in your endeavour to discover the self as an unbroken unity, and thus an adequate principle, descend below the level of relational consciousness altogether to a state of mere feeling. But if you do, you are not dealing with a 'self' at all, but with an experience prior to the very emergence of the distinction of self from not-self. If, on the other hand, you accept the relational level, at which alone self-hood proper can exist, the relation to something not-itself is inexpugnable. Hence the attempt to save the ultimate reality of self-hood is in the end no more successful than the attempt to save the reality of the less dignified concepts which Bradley's dialectic has already traversed. The finite self is, in a word, like everything else save the Absolute, 'appearance' and not 'reality'.

Now with the substance of this argument I am in the fullest agreement. We cannot ascribe ultimate reality to the finite self, and Bradley has, in my view, furnished a sufficient demonstration of this. The immediate sub-relational unity of feeling is not a 'self,' and the self as it exists at the relational level is not a true unity in difference (and thus not ultimately real). But this result does not wear quite the same forbidding significance when it is considered in relation, on the one hand, to what was said (in Chapter III (Noumenal And Phenomenal Truth. Section 1. Absolute Idealism'S Rejection Of Correspondence Notion Of Truth).) on 'final phenomenal truths,' and, on the other hand, to what has more recently been argued as to the directness of our perception of self-activity. Although it is true that we do not have in the self 'ultimate reality' in the full sense, it may also be true that we do have in the self as ultimate a reality as finite mind can hope to apprehend. As we saw earlier, there are certain matters which finite mind just has to accept as final for it, 'human ultimates,' datal or intellectually incorrigible facts, which do not, indeed, satisfy thought's demand for self-explanatory unity in difference, but which yet no conceivable advance in knowledge can either overturn or modify. Of such a character, I submit, is our apprehension of our 'self in so far as pertains to its nature as an active centre of experience. 'Self-knowledge,' in the ordinary and fuller sense of the term, rests, of course, upon a multitude of mediate processes, and is conditioned by the progress of our cognitive contacts with our social and natural environment. And because this kind of self-knowledge is mediate, and in consequence liable to an indefinite amount of revision as the system which inspires the connections expands, the 'self' which is apprehended in this self-knowledge can never be accepted as an ultimate reality even of the phenomenal order. It is quite otherwise, however, with the self-knowledge which pertains to the self only in its formal character as an active centre of experience. This knowledge (if I have been right) is not achieved through mediate processes, but given in direct or immediate awareness. And because this kind of self-knowledge is thus immediate, and in consequence free from the conditions which impose the necessity of subsequent modification, the 'self' which is apprehended in this self-knowledge can, indeed must, be accepted as an ultimate (phenomenal) reality. Thus while we may reasonably doubt the 'reality' of the self of our self-knowledge in respect of any feature whatsoever of its apprehended concreteness, of its 'reality' simply in respect of its character as an active centre there can be no significant doubt.1 In this character of itself the finite self can be absolutely assured of its own ultimate (phenomenal) reality. And bare and formal though this character certainly is, it is, after all, the character of central significance in self-hood, the character which we are chiefly concerned to justify when we contend for the reality of our finite 'self.' Even if our assured knowledge of our self's reality extends no further than this, our knowledge will yet be very far from trivial.

1 I am not suggesting, of course, that we can apprehend bare 'activity' in and by itself. What we apprehend is always a determinate self 'active' in a determinate objective situation. But since the character of activity as apprehended remains precisely identical whatever be the apprehended natures of the determinate self and the determinate situation, the abstraction of the 'perception of activity' from the total experience is a valid, and not a vicious, abstraction.

It is necessary to emphasise, however, that this 'ultimate reality' which we can ascribe to our self qua active centre belongs only to the phenomenal level. We cannot, in the end, ignore the fact that the self qua active centre stands in essential relationship with the self qua possessed of concrete characteristics, and in equally essential relationship with a determinate objective situation, and that these relationships are, for mind, merely de facto. There can be no question of our having arrived, in the apprehension of the self qua active centre, at an 'intellectually satisfying' object, and therefore a 'noumenal' reality. We have arrived only at an 'intellectually incorrigible' object - one which, because not mediately known, can suffer no modification from advancing experience. There still remain essentially askable, though by finite mind essentially unanswerable, questions as to the cosmic significance of that which we apprehend as our 'activity.' And in the absence of answers we cannot ascribe noumenal status to our object. It is, without doubt, a paradoxical situation that the finite mind should thus be able to regard as inadequate pronouncements which it yet sees to be final for itself. But it is a paradox bound up with the essential paradox of man himself, as a creature who partakes at once of finitude and infinitude. In as much as he is finitely conditioned, man is forever precluded from the vision of things sub specie universi: but in as much as he is conscious of his own finitude (and so far infinite), he must forever recognise the deficiencies of his limited vision, and must refuse the title of 'noumenal reality' to the objects even of his most assured insight.

To sum up. There is a significant sense in which we can legitimately hold our finite self to be an ultimate reality. It is only a 'phenomenal' ultimate. But when once we have succeeded in re-orienting our mental attitude in the way demanded by subscription to the doctrine that the ultimately real in the full sense is in principle inaccessible to human reason, and have come to recognise accordingly that we as finite beings must look for our ultimate realities not in the ultimate reality but in whatsoever by the conditions of our nature we are obliged to accept as final for us, the denial of validity to the self's indefeasible certitude of its own 'reality' will be seen to hold good, I think, only in a sense in which it is very nearly unimportant.