It will not repay us, I think, to dwell further upon the Correspondence theory of the criterion of Truth. I pass on to explain what has seemed to me to be, on the whole, the most satisfactory view. This view includes elements both of the Intuitionalist and of the Coherence theory - with Coherence as the dominant element. Coherence, I shall argue, is not the sole test. There are certain well-defined cognitive situations in which an immediate certainty (I am speaking, be it remembered, at the Phenomenological level) would seem to be procurable. But, in the main, Coherence is the test; and since I do not suppose myself able to add much that is of value to the explanations and vindications of this test to be found at large in Idealist writings,1 I shall say little more about it than is required to make clear the manner of its application to the present metaphysic.
What is to be maintained, then, is that the truth of a judgment is to be tested by its capacity for harmonising with all the other judgments which we make about reality. (The necessary reservations just alluded to will be dealt with later.) On the whole, the orthodox Idealist justification of this test does not suffer alteration or diminution of force on account of the different metaphysic that is adopted in this work. The difference lies rather in what it is that is supposed to be thus tested. It is not, for my view, 'Noumenal' Truth that is here being tested. The test of Noumenal Truth has already been considered in Chapter I (The Epistemological Approach To The Supra-Rational Absolute. Section I. Introductory). and found to lie beyond the grasp of intellectual process. The intellect can attach no positive significance, paradoxical as it may appear, to that which would alone satisfy its demands. Here we are to discuss rather the test which is proper to intellect in its positive processes. It is then only Phenomenal Truth of which I am claiming that Coherence is the test - or at any rate the main test. And if this admission should seem seriously to derogate from the importance of the inquiry, it is well to remember that Phenomenal Truth is just another name for that which Truth concretely means for human beings. Philosophy itself (if I am right) can aim no higher than this. The intellectually satisfactory is an impossible ideal. The intellectually incorrigible is the best to which thinking can aspire. In investigating the conditions of intellectual incorrigibility we are investigating the conditions of what may fairly be called final truth for human beings. And this can hardly be esteemed a trivial matter.
1 Among such discussions, chaps, vii. and viii. of Bradley's Essays in Truth and Reality seem to the present writer especially worthy of attention.
This being understood, then, we may proceed to show the manner in which the Coherence test connects with our doctrine. Phenomenal Truth means, we have taken it, the correspondence of our judgments with the objective reality about which we judge. Now if we are to possess any criterion for determining when we have achieved correspondence, it is obvious that we must know something already of the character of this objective reality - while, on the other hand, the something that we know must not be such as to render the process of thinking itself otiose. Do we know anything of this kind? Yes, one thing we do know 'beyond a peradventure.' We do know that the real is 'self-consistent.' We know this, because the presupposition of it is immanent in the nature of all thinking. We cannot question it, because even a significant question assumes it. We may fairly be said then to 'know' that objective reality is non-contradictory, or self-consistent.
Now to be non-contradictory, we saw, means to be a unity of differences in which the differences are in no sense external to the unity. And this character, we also saw, attaches only to a type of whole to which the intellect can legitimately assign no positive meaning. That, however, is beside the point from the Phenomenological point of view. For the intellect does in fact assign a positive meaning to a non-contradictory system, even if it is a meaning which is ultimately indefensible. Reality it rightly pronounces to be self-consistent, and, placing its own interpretation upon the nature of self-consistency, it proceeds, in order to bring about the correspondence of its ideal world with the real world, to endeavour to achieve a fully harmonious body of judgment. The fact that it seeks to secure harmony by uniting differences through a 'ground' which is perforce partially external to the differences - by a method, therefore, which is strictly self-stultifying - matters not at all from our present point of view. What matters is that it does adopt a characteristic method, that it does mean something to itself by 'non-contradiction' or 'self-consistency.' And its 'test of truth' is just what it means by self-consistency. A judgment to be true must unite differences consistently. If it does not, it does not represent ideally the real world, which we know to be self-consistent.
Our operative test of Truth, then, is coherent system; even if, in seeking after such system, we have to interpret it in a way which debars it from ever being truly systematic - a true unity in difference. We may ask now what is to be said of the possibility of completely satisfying the Coherence test, and of thereby assuring ourselves of the correspondence of our ideas with objective reality. The question, however, in a manner answers itself. Coherence is our test, but if coherence were an ideal capable of self-completion we should not have required to reject it as metaphysically unsound. And since I can thus never make my cognitive experience fully answer to my test, I can never be sure that any judgment does represent the aspect of the real order which it tries to represent. On the other hand, by the very character of the test it will be natural to claim greater certainty for our judgments the more fully they are developed along the path which is taken to lead to perfect system; and that means in practice, the more comprehensive and internally harmonious is the unity or system which furnishes the ground of the connection of differences. But complete certainty is plainly an idle dream. I may, for example, be comparatively sure that a certain chemically defined substance acts as a corrosive force upon certain metals. I may have pushed my researches into 'grounding' both deep and wide, and seen that if this judgment is false then a great part of my ideally constructed physical world falls to the ground. But both in depth and in width, both intensively and extensively, it is evident that an indefinite amount of potential data remains unascertained. There is no question then of having here, or anywhere, a fully self-contained system. The insight into an absolutely necessary connection which could come only from a system that is self-complete is forever denied us. We cannot say, therefore, with legitimate assurance that any connection, even if supported by a system which is relatively both expansive and internally harmonious, corresponds with what obtains in the real order. At no stage in finite experience can the Coherence test be applied in a way which excludes grounds for doubt.