there is no ultimate distinction between mind and matter, if knowing mind and known object are parts of one another and of a spiritual indivisible whole, it is a matter of indifference whether we call such a whole real or true, but if there is a distinction between mind and matter, or if such a distinction be assumed for the purposes of discussion, although we may be convinced that it is invalid in the long run, it will be found convenient to reserve the epithet "true as an attribute of mental acts only, leaving real as an epithet of the thing known or object of those acts.

If

Also it may be held, and, I believe, justifiably held, that there is a Platonic dos of truth. there is such an dos it is non-mental and its existence is a fact. The existence of such an dos might appear at first sight to invalidate the distinction which has just been drawn, but I do not see that there is any inconsistency in saying that the existence of truth is a non-mental fact, although the manifestation of the dos attaches itself only to mental functions. The ειδος of Truth would exist even if there were no minds to know it, but the manifestation of the dos in particulars is dependent upon there being minds to make true judgments. Thirdly it is important to emphasise the distinction between the meaning of truth and the criterion of truth. What truth means is clearly different from how you know a judgment to be true. The correspondence and coherence theories are concerned to find out

what is meant by truth: they are not concerned to discover a criterion by which you may tell what is true and what is not.

These preliminary distinctions having been made, I propose first to consider the correspondence theory of truth.

II.

The correspondence theory is chiefly concerned to establish a meaning for truth, in cases of true judgment and true perception. The meaning of truth is, according to this view, correspondence with fact, and when we call a judgment true what we mean is that there is an external fact with which that judgment corresponds.

In its usual form the theory asserts that a judgment such as "this book is red " is true if there is an external fact namely a red book with which the judgment corresponds, and is false if there is no such fact. Now it is obvious that this view of truth can only be held by an extreme realist. It assumes first that we see the book exactly as it is, secondly that our perceptions are infallible, and thirdly that the book is what it is apart from its relations with all the other objects in the Universe and with the Universe as a whole.

Now Idealist philosophers do not believe any of these things. They believe that we see this book partially and imperfectly and that our fuller knowledge of it is "a revolution in which the book is swept away and determinate connections between

:

determinate universal concepts are substituted and they do not believe that the book can be isolated from its relations and considered as a separately existing entity. It is obvious therefore that the ordinary correspondence view of truth cannot be held by a Monist, or incidentally by a Bergsonian; it is also obvious that directly we modify our extreme Realist view of perception the correspondence theory gets into difficulties, for if the book depends for its being or even for its significance upon its relations with other objects not known, or if we do not see the book in its entirety exactly as it is, we cannot know what the fact is with which our true judgment is to correspond. With regard to the three assumptions mentioned above as being the assumptions upon which the ordinary correspondence view of truth rests, I believe the first to be false and the second and the third to be true.

As regards the third, I have tried to show in the previous essay that objects can be isolated from their relations without falsification of their nature, and that it is possible to know the nature of an object without knowing all its relations. As regards the second, I gave reasons in the first essay for supposing that although our perceptions are selective, partial and incomplete, they do faithfully convey to the mind that part of the selected whole which is in point of fact actually perceived. I also tried to show, however, in that essay that we do not perceive objects

exactly as they are, that is to say, we do not perceive them in their entirety, but that we perceive them only in part, and that we select those aspects of them which form the content of our perceptions according to systems that interest us. I do not therefore believe that the first of the conditions which must be assumed if we are to hold the ordinary correspondence view of truth, namely, that we perceive the red book exactly as it is, is tenable, and for this reason it seems to me desirable to modify the correspondence view as it is ordinarily held.

The modification which it appears to me desirable to make is as follows.

The correspondence which takes place when a true judgment is passed, is formed by a correspondence not between judgment and external fact, but between judgment and perception, or more fully between my perception and the judgment passed by my mind, of which the perception forms the basis or stimulus. The advantage of this modification lies in the fact that both the entities between which the correspondence is to be made out are known. They are both parts of my mind.

44

Now it has frequently been urged against the old copying theory of truth, first advanced by Aristotle, upon which the later correspondence view is based, that it postulates a correspondence between two things of which one is not known.

The difficulty involved is similar to that which

attaches to Representationalism. Either I know directly, the fact with which my true judgment is to correspond or I do not. If I know it directly, what need is there for me to pass a true judgment about it in order to correspond with it? If I do not know it directly, how am I to know that my judgment does correspond with it? Where, that is, we have two entities of which we know (A) but do not know (B), we can never know that (B) corresponds with (A). A true judgment therefore according to the old copying view is either superfluous or unauthenticated.

The view which regards the correspondence as a correspondence between judgment and perception is not open to this objection. I have endeavoured to show in the first chapter that in all perception the mind possesses the power of going out beyond what is actually given, and forming judgments upon our partial and incomplete perceptions. What is before the mind in thinking is not what is actually given in perception but something more than that. That more " is mental construction. If it corresponds with the perception, we have a true judgment: if not, "error" has crept in.

44

The correspondence view in its ordinary form has considerable difficulty in accounting for error. If truth lies in correspondence between a judgment or perception and an external fact, what is the external fact in the case of error? It may be argued that it is just because there is no such fact, because correspondence cannot be made out,