4 The Knower and the Known (1927)

We are accustomed to think of Realism and Idealism as conflicting views about knowledge. In this paper I shall be concerned partly to bring forward arguments in support of the realist view of knowledge, but even more to indicate what I take to be important consequences, in regard not only to mind and knowledge but to philosophy in general, of accepting that position. While it may be conceded to Professor Montague that “the point at issue between realism and idealism should not be confused with the [point] at issue between empiricism and rationalism”,note in that the former has specially to do with knowledge while the latter has not, there are reasons, which I think conclusive, for holding that a realist can only be an empiricist. The question of the nature of relations is at any rate one issue between rationalists and empiricists, and, as the authors of The New Realism have shown, the basis of a realistic theory of knowledge can only be a certain theory of relations; which enables us to draw definite conclusions from the contention that knowledge is a relation. Thus, according to Professor Montague, “Realism holds that things known may continue to exist unaltered when they are not known, or that things may pass in and out of the cognitive relation without prejudice to their reality, or that the existence of a thing is not correlated with or dependent upon the fact that anybody experiences it, perceives it, conceives it, or is in any way aware of it.”1 And Professor Marvin makes the theory of relations here indicated still more explicit. “In the proposition ‘the term a is in the relation R to the term b’, aR in no degree constitutes b, nor does Rb constitute a, nor does R constitute either a or b.”1 Knowledge being taken as a relation, it is thus asserted that, when I know this paper, “I know” in no way constitutes this paper, nor does “know this paper” in any way constitute me, nor does “know” in any way constitute either me or this paper.

The view that knowledge is a relation implies that knower and known are two different things or that, in knowledge, the knower is not the known. It is indeed admitted on any view that there is a distinction between them, and if knowledge were not then called a relation, it would only be because relations were held to be comparatively unreal; but those who would say this would say the same about distinctions. The realist is thus found to be maintaining that distinctions are absolutely real. According to the opposing view, distinctions are “distinctions within identities”, and any relation is a “form of identity”. If a thing

  ― 28 ―
is really related in a certain way, the relation in question belongs to its “nature”, and since that to which it is related is thus not essentially separate from it, a certain “identity of nature” holds between the two. But this theory of natures or essences is precisely rationalism, and the realist, in denying that aRb asserts or implies any identity between a and b, is taking up an empiricist position.

He does not, of course, deny that there is a certain “identity” in the case, viz., the identity of aRb as a given situation or state of affairs. When it is argued that the distinction between subject and object must be “a distinction within an identity”, the reason alleged is that there is no subject without an object and no object without a subject. But this merely means that any “subject-object” relation has two terms; it could not for a moment show that knower and known are not two different things, or that anything is to be regarded as in itself either a subject or an object. Hume points out that the facts that it takes a man to be a husband and that every husband has a wife do not imply that every man has a wife; similarly, to assert that it takes a mind to know or a thing to be known does not imply that every mind knows or every thing is known. But of more importance here is the fact that while we speak of a certain marriage, and agree that there is no husband without a wife and no wife without a husband, we find husband and wife in any marriage to be two different persons. The one is not the other. Those who argue that knower and known are in some way identical because they are in a certain relation, have also to maintain that any two different things are in some way identical, since any difference is a certain difference or since “A is different from B” is a certain state of affairs. So that when we say that A is not B, we are somehow also saying that A is B and B is A. On this basis discourse would be impossible.

It is thus seen to be logically necessary to hold that, in knowledge, “the knower is not the known”. It follows that to tell us what a man knows is not to give a description of that man (to state some character or quality which he has), any more than to tell us who knows it is to give a description of a thing. The fact that a man does know certain things may enable us to infer that he is a man of a certain character, but this inference would not be possible unless we had previously come to believe that only persons of that character knew these things, i.e., unless we had had previous opportunities of observing that character independently. But again, since knowledge is a relation, to tell us that a man knows is not to give a description of him, any more than to say that a thing is known is to say what sort of thing it is. We may believe that only beings of a certain sort do know, but that depends on our having recognised their character independently of their knowing. And, in general, in saying of any two related things that they are distinct, we must suppose each to have some character, or certain qualities, of its own. We must distinguish complete statements like “X is a man” from incomplete statements like “X is a husband”. The latter is, of course, used roughly to convey the fact that X has those characteristics which will be found in the first

  ― 29 ―
term whenever there is a true statement of the form “X is the husband of Y”. But those characteristics are understood to be discoverable by observing X alone, while we could not in that way find out what was meant by his being a husband.

Arguing then, as realists, that no thing or quality of a thing is constituted by the thing's relations, we have to assert that nothing is constituted by knowing and nothing by being known. The notion of “that whose nature it is to know” is expressed in the term “consciousness”; the notion of “that whose nature it is to be known” in the term “idea”. Realism is therefore concerned to reject these terms, as involving the attempt to take relations as qualities. If the term “conscious being” merely meant that sort of thing which can know, and “idea” that which can be known, they might be used in incomplete statements similar to “X is a husband”. But we must have some notion of what sorts of things these are, since we could never have supposed that nothing knew something or something knew nothing. Thus we must know what sort of thing a mind is, independently of terms like “consciousness” or “state of consciousness”; and we must be able to describe things independently of their being known or of their being known in some particular way, so that “sensa”, for example, cannot be a proper name for any species of things. A strictly realist theory must dispense with all expressions of these sorts, in order to be consistent with its empirical starting-point and logical basis.


(a) The theory of “ideas” as entities “whose nature it is to be known” (or which are “essentially known”) is most explicitly formulated by Berkeley. To think of what is known as having a nature independent of its being known is, he says, to be guilty of “abstraction”. This error consists in thinking separately of things which cannot exist separately. Thus we cannot truly know any object without knowing “all about” it (its “whole nature”); for if we only knew something about it, we should be separating that something from other somethings which in fact are also about it. In terms of this theory, if a thing really is known, we cannot think of it otherwise than as known, or we should not be thinking of it. “Can there”, asks Berkeley, “be a nicer strain of abstraction than to distinguish the existence of sensible objects from their being perceived, so as to conceive them existing unperceived?”

It is, of course, impossible to maintain a view of this kind consistently, since strictly in accordance with it we could make no statements at all. That Berkeley does not do so is shown when he says, “It is not in my power to frame an idea of a body extended and moved, but I must withal give it some colour or other sensible quality.” He really ought to maintain, in the case of a coloured body which is extended and moved, that its being coloured and its being extended and its being moved and its being a body all mean precisely the same thing; in which case his argument is stultified. But otherwise he is admitting that we can conceive the

  ― 30 ―
thing as having any one of these characters, that we can truly assert something about it without saying “all about” it. He does in fact admit that, though it may be true of an object A that it is given to us by God as a sign of B, we can know A without knowing this fact. It may equally well be that we may know a thing which is known, without knowing that it is known. There is no reason for denying this, if its being known is only something about it. And the alternative, which Berkeley would have to follow, is that, since we can only conceive separately what may exist separately, the separate statement that a thing is known implies that nothing else can be said of it but that it is known. There could not then be a number of different known things; there would simply be the essence “known”. Indeed, since what I know must be “known by me” and there is nothing else that it must be, all that I can ever know is the single essence “known by me”.

The only sort of assertion that we could make in starting from such an essence as “known” would be the identity “The known is known”; and only by means of abstraction (passing from a whole nature to a supposed part of it) could any consequences appear to follow from such a statement. Now Berkeley does start from an identity, stated negatively, viz., “What is perceived cannot be unperceived”; which is merely an expression of the essence “perceived”. But he proceeds from this, as the first quotation shows, to draw the conclusion that what is perceived cannot be conceived to be unperceived. Now the only guarantee of this conclusion is the fact that the thing is perceived; and if this is a guarantee, it must be because the thing is perceived to be perceived. (Here the notion of “idea” emerges, in the form of the “percept”; the conceiving of “concepts” and the sensing of “sensa” are suppositions of the same type.) What is perceived to be perceived cannot be taken (there is no special force in “conceived” here) to be unperceived. The obtaining of the given conclusion from the identity thus depends on the substitution of “perceived to be perceived” for “perceived”. And the plausibility of the conclusion itself depends on ambiguity; it is plausible as meaning that we cannot conceive or suppose that “what is perceived is unperceived”, but not in the required sense that things which are perceived cannot be supposed not to be perceived and must be supposed to be perceived. This cannot be admitted, since the various things that are said to be perceived cannot have their whole nature constituted by being perceived.

The fact, then, that we can make such statements as that red, or something red, is perceived, is sufficient to dispose of Berkeley's theory that what is known must be known as known. It would, on the contrary, be true to say that we know things as independent of being known, since we can only know them as existing and having characters of their own. Berkeley's theory, it should be noted, is not dependent on the use of the term “perception”; it could be maintained in exactly the same way that “whatever is apprehended cannot be unapprehended”, without any reference to modes of apprehension. So that his criticism of Locke, who

  ― 31 ―
had first admitted the “essentially apprehended” and then presumed a further knowledge of independent things, is quite sound. The criticism applicable to both is that we never know “ideas”, but always independent things.

(b) Descartes's demonstration that there is something “whose nature it is to know”, or, as he puts it, “whose whole essence consists in thinking” (i.e., “consciousness”), proceeds in a similar fashion to Berkeley's substantiation of “ideas”; in fact, it may be said that Berkeley has simply applied to the known the principle of Descartes's argument about the knower. The latter is complicated by the fact that what guarantees the essential knowingness of a knower is the knower himself; but the same mechanism of essence, identity and ambiguity can be discerned. The assumption is that we cannot suppose ourselves, in knowing, not to know, i.e., we cannot suppose that when we know, we do not know; but it is employed as if it meant that we cannot, in knowing, suppose ourselves not to know. Or, putting the argument positively, we must suppose ourselves, in knowing, to know; hence we must, in knowing, suppose ourselves to know (or, in thinking, think that we think). By means of identity and ambiguity, therefore, Descartes arrives at the conclusion that we always know ourselves as knowing, and never know ourselves as anything else; because we can suppose ourselves, though knowing, not to have that other character. The method, once more, is that what can be conceived separately from a certain thing is not of its essence but is a different thing, while what cannot be conceived separately is of its essence. And the strict consequence would be that no positive (non-identical) assertion could be made, since we could only make it by specifying a distinct part of a “whole nature”.

The view that in knowing we know ourselves knowing, that we know as knowing or consciously know, is thus seen to be as ill-founded as the view that we know things as known. The identity “the known is known” does not imply that it is the same thing to know X and to know that X is known; nor does the identity “I know what I know” imply that I must know that I know it, or know anything about myself at all, in knowing it. Descartes, having taken his knowledge as a subject to be considered, cannot in the same argument doubt that he knows; but a man who knows need not have taken up this position, and might quite well doubt that he knew, or that he doubted, or that there was such a being as himself. The conclusion that a person could not know without knowing his knowing, as we have seen, depends on ambiguity, and the conclusion that he could not know himself without knowing his knowing depends on the assumption that he must know “all about” anything he knows. This theory being logically untenable, there is no ground whatever for supposing that we must know minds as “conscious” or for treating their knowing otherwise than as a relation to other things which is not part of their own “character”. We have no more right to talk of a “conscious state” than of an “on state” or an “above state”. And we may take it as possible that anything which knows may at another time not

  ― 32 ―
know, just as things which are known may at another time not be known.

It is not in the least implied that minds are not known, but only that they are to be known as having certain qualities. “The knower is not the known” has sometimes been taken to mean that the knower is not known; hence the doctrine of the transcendental ego. Alternatively, the distinction is taken to imply that the knower can only be known as knowing, i.e., known in a different way from things which are known as known. But all that is implied is that the relation has order; it is not asymmetrical, but at least it is “non-symmetrical”. When A knows B, B need not know A; and even if B does know A, this is a different state of affairs from A knowing B. Just as, in the relation of parenthood, “the parent is not the child” and yet is always the child of someone else, so, when I know a thing, someone else may know me and he may know my knowing the thing. Only if there are cases of this kind can it be possible for us to talk about “knowledge”. But the person's knowledge of my relation to the thing is distinct from his knowledge of my qualities.

As regards my knowledge of myself, this will have to be accounted for by saying that a certain process in my mind knows another, or knows myself, but without knowing itself. We can only know ourselves, in fact, as certain very familiar objects. And if it is urged that the process which knows does nevertheless belong to myself, the answer must be that what we know consists not of things simply but of states of affairs (or propositions). Suppose, then, that I know that I am angry, the “object” may be roughly expressed by saying that within a certain contour anger is occurring; and the fact that the process which knows it also occurs within the contour is not to the purpose, since we do not require to know “all” that occurs within the contour. That which knows a given occurrence is a different occurrence; it is not my anger which knows my anger. Detailed discussion of how we come to use the term “I” would be out of place here. It is enough to point out that on the realist theory the conception of a mind as a “unity” or indivisible whole cannot be sustained; according to that conception neither I nor anyone else could know anything about my mind.


According to realism, I have argued, we never know “ideas” but always independent things, or rather states of affairs. It seems to me to follow that such expressions as appearances or data, and as concepts, percepts or sensa have no place in realist theory. If, e.g., there is a peculiar way of knowing called “sensing”, it will only be on the assumption that relations somehow constitute their terms that we can use the term “sensa” to describe a class of things or a way of being. If, on the other hand, any class of things can properly be described as “sensa”, to speak of knowing them as “sensing” is to make the same sort of assumption, and is no more justifiable than to speak of knowing trees as “treeing”. I should maintain that there is no such thing as either sensing

  ― 33 ―
or sensa, since “the sensa which I sense” are taken to be those things, my knowing which depends on where and how I am, and since this (a) does not describe the things, (b) is true of all my knowing.

For Berkeley the things we know are “essentially related to our minds” and thus have a “relative existence”, as our ideas. The theory of sensa is likewise a theory of “relative existence”, in someone's or some “sense-field”, and of “that whose nature it is to have certain relations”. Dr Broad's theory, in this connection, does not, I think, differ greatly from other theories of sensa. Sensa are shown to be private and non-physical because of their dependence on certain conditions. Dr Broad does not commit himself to the view that sensa are mind-dependent. “The facts are on the whole much better explained by supposing that the sensa which a man senses are partly dependent on the position, internal states and structure of his body”. But certain examples, though they “do not suggest for a moment that sensa are existentially mind-dependent…do strongly suggest that they are to some extent qualitatively mind-dependent”.note Now dependence is presumably a relation, and if a certain existence or a certain quality depends on something, this does not justify us, rejecting as we do the theory of constitutive relations, in describing it as a “dependent existence” or a “dependent quality”. The existence or quality, though it might not have been but for that other thing, is independent in the sense of being distinct and having a character of its own. If Dr Broad's explanations were correct, we should have to say that a certain thing now exists because my body was in a certain position, etc., and has certain qualities because my mind was in a certain condition. Granted all that, the thing now exists and has these qualities, and no reason has been shown for calling it private or non-physical.

Whether the explanations should be accepted is made exceedingly doubtful by noting the ambiguity of the statement that “the sensa which a man senses” are dependent on his body. This may merely mean that what is dependent is “the fact that he senses these sensa”, i.e., his sensing them, i.e., his standing in a certain relation to them. That this should be dependent on where he is could occasion no surprise to commonsense, and would justify no statement about the dependence of the “sensa” themselves. The fundamental criticism is, however, that what exists because of me nonetheless exists, apart from or independently of me. The houses which would not have existed, had not men planned and built them (i.e., but for their minds and bodies), are physical and are not private to these men; they stand for other men to see them and may remain when no one perceives them at all. The argument from dependence commits us to the Berkeleian theory of “relative existence”; as does also the notion of a special “sense-field” in which a given sensum occurs. Dr Broad regards it as a merit of his sensum theory that it does not require the assumption of an absolute Space-Time. But “absolute” Space-Time is simply that in which things “absolutely” exist, and realism

  ― 34 ―
is committed to the rejection of “relative existence”, and so of “relativity”.

It may now be asked what reasons there are for supposing “sensible objects”, which differ from physical objects, and which are brought about and affected by persons to a greater degree than the latter. In arguing that there can be no adequate reason for such a supposition, I shall consider mainly the question of “sensible” shapes and sizes. Dr Broad explains “the notion of sensible appearance”, in regard, particularly, to shape, as follows: “We know that when we lay a penny down on a table and view it from different positions, it generally looks more or less elliptical in shape. The eccentricity of these various appearances varies as we move about, and so does the direction of their major axes.…It is a fact that we do believe [that there is a single physical object…which appears to us in all these different ways]. It is an equally certain fact that the penny does look different as we move about.”note There then arises a difficulty about the relation between the round penny and an “elliptical appearance”, or something “appearing elliptical”. As regards the latter alternative (which Dr Broad rejects and which he connects with the theories of Professor Dawes Hicks and Professor Moore), it seems to me necessary to point out that “appearing” is a relation, viz., that of being known or apprehended. So that what is apprehended in this case is that “something is elliptical”, and, since this interpretation does not allow us to speak of “an appearance”, the precise belief would seem to be that the penny is elliptical; a belief which is simply false. Now there are cases in which such a false belief is held, but in many cases it is not, so that it may be questioned whether anything “appears elliptical”. In any case, “appearing elliptical” does not state a relation between the penny and us, except when we are wrong.

“Appearing elliptical” at least involves apprehension of a state of affairs, but, according to Dr Broad, the “elliptical appearance” is apprehended without judgment, though it is apprehended as existing. This compromise is as unacceptable as that of Berkeley. We have something whose “whole nature” is apprehended (since an appearance is exactly what appears to a person), and then it is supposed to exist. As before, its nature and its existence must mean the same thing, and it must be perfectly indescribable. A similar point emerges in connection with the “different appearances” mentioned. Unless we think of a physical object as something which has to be known in its “whole nature”, there is no reason why it should not have different appearances, i.e., why different characteristics of it should not be observable from different standpoints.note And it cannot be denied that when we do know a physical object, we know a variety of distinct things about it. The recognition that, whenever we know, we know existences and that to know existences is to know

  ― 35 ―
states of affairs in which complex things occur, is sufficient to dispose of the theory of “appearances”.

Thus an “elliptical appearance”, in respect of the penny, can only mean a false belief. I have said that in many cases this false belief is not held; what visibly appears to us is the round penny (or the penny's being round), even though the round surface is not at right angles to the line of vision. The assumption underlying the whole theory of differing shapes and sizes, seen from different directions and distances, is that we look out at, or there visually appears to us, a plane projection of the visual field. It is quite certain that a penny may be so placed that its projection on a plane perpendicular to the line of vision is elliptical; it is equally certain that the further a penny is beyond such a plane, the smaller its projection on the plane will be, and that it may be so near the plane that its projection is larger than that of the moon. But if, as is the case, it is not true that, when we look out, we either look at or see things arranged in a plane, if we do see things at various distances and at various angles to one another — in short, in three dimensions — then the contention that a thing looks smaller as it retreats, or that a round disc looks elliptical when it is oblique, is robbed of its force. Since we see things in three dimensions, there is no reason why we should attribute to a thing itself the shape of its projection on a plane perpendicular to the line of vision, or see that shape at all.

In cases where there is said to be an “elliptical appearance”, there really is something elliptical, viz. (assuming the surface affected to be plane and perpendicular, or sufficiently near the perpendicular, to the line of vision), that part of the surface of the retina on which the rays of light from the object fall. As we are not looking at the retina, this does not affect the question directly. But it is sometimes assumed (on a theory similar to Berkeley's) that the retina is affected in precisely the same way, no matter how far the light has travelled; that consequently we cannot distinguish distances by sight, so that any part of a visual object, or field, must “visually appear” to be at the same distance as any other part. (What distance this could be is quite obscure.) It is, however, perfectly conceivable that rays from different distances should affect the retina differently; even though “the picture imprinted on the retina” remained the same, the effects might differ in other respects. The fact is that we do see things at different distances, and if it is alleged that this must be due to something not given by vision, the answer is (a) that what it is due to is quite irrelevant, (b) that the objection involves the attempt to maintain that we cannot see what we actually do see.

But, though reference to the retina is irrelevant, there may still be something elliptical to be considered. We commonly see things against a background, and if that background were perpendicular to the line of vision, the shape of the part of the background concealed by the thing would be that of the thing's projection on a perpendicular plane. In this way an oblique penny may conceal an elliptical part of the wall of a room — and this also happens in cases where we are looking obliquely at

  ― 36 ―
the wall. Now we are just as capable of observing that an elliptical part of the wall is concealed as of observing the round penny; and the concealed elliptical part is just as much a physical object as the penny. In such a case, on a casual glance, we may fail to distinguish the distances of wall and penny, and suppose that the penny is elliptical. There is something elliptical in the same direction as the penny, something moreover of which we only see the shape, and there is a consequent possibility of our attributing that shape to the seen penny. If, however, something in the appearance of the object suggests that it is a penny, then we doubt the supposition we have made, and by stricter attention observe that the penny is not in the plane of the wall but is oblique and round. The previous mistake may be described by saying that we had “displaced” the elliptical shape from the wall to the penny, just as we might displace the red colour of red spectacles to the things we saw through them. Also, the fact that we know that the penny is round need not prevent us from making the mistake; it would only require to be two different processes which had the two beliefs, and we should attribute “knowledge” to that process which was able to overcome the other when they came into conflict.

It is possible for us, then, correctly to distinguish something elliptical from something round, the two being physical objects occupying different places; whereas, if we could only distinguish things in accordance with their projections, almost all our observations would be mistaken. It may be said that we can judge or discriminate best the shapes of surfaces which are perpendicular to the line of vision.note But though we could less easily distinguish a circle and a nearly circular ellipse if they were lying obliquely to our vision, that would not prove that we see their projections. We may tend to err by assuming that the easier conditions are fulfilled, but it is possible, when we are presented with an oblique circle, to “see it circular”. Again, we can judge sizes best when the things compared are close together; but we can see a distant tree larger than a man near at hand, who, if he stepped aside, would conceal the tree, and the relative sizes of the projections only appear to us in terms of concealed portions of a common background. Improvement in discrimination is possible, and may come about with the aid of other senses, as well as through the movements of the observed things and of ourselves in observing them. But it could never begin if we saw a flat picture.

I have considered at length the case of “elliptical appearance” in order to show the kind of mistake that is possible (though not necessary), and the possibility of correcting it by means of other judgments. I regard the general theory that I have advanced, as showing that an account along similar lines could be given of more difficult cases. In general, it cannot be maintained that in this (or any other) sort of apprehension, judgment, i.e., apprehension of states of affairs or situations, is not

  ― 37 ―
involved, since (a) it is always something that appears elliptical or smaller; we do not apprehend “ellipticity” or “smallness” by itself; (b) that something is always taken to be in some particular place. Any such judgment will be either correct or mistaken; but correction will only occur by means of judgments of the same order. At no time in the process of making our observations more precise, i.e., of discovering new distinctions and connections, as well as previous errors (and it is just in these ways that, on any view, we extend our knowledge of physical objects), do we suppose that we are not observing the things themselves and their actual shapes and sizes: at no time do we distinguish a “datum” or “sensum” from a thing. There is no thing or quality, then, which we can suppose ourselves to know “all about”; discrimination and association are always possible — whereas a “datum” could enter into no proposition. The same considerations are applicable to all the so-called sensa. The artist comes to discriminate and know colours better; and we can apprehend by sight many other qualities besides colour. What we see, like what we apprehend in any other way, is always complex, always a state of affairs; and the physical object is no more to be supposed to lack the “secondary qualities” than to lack the shapes which we see.


The fundamental reason for rejecting the term “consciousness”, or “awareness”, is that, like “sensum”, it involves the notion of “relative existence”. This is brought out very clearly in the account given by Professor Dawes Hicks (l.c., p. 319) of the theory of Meinong. “With the doubtful exception of certain feelings and desires, he lays it down as a characteristic feature of the psychical, in contradistinction to the non-psychical, that it is directed upon something…A physical event can be described in and for itself. Not so a mental event. To speak of an act of awareness simply would be to speak of that which is never met with. Awareness in and for itself has no existence, and, indeed, no meaning; a ‘something’ of which there is awareness is its indispensable correlative.” The natural conclusion would seem to be not that a mental event cannot be described in and for itself, but that it might possibly be described as feeling or desire, and that, however described, it may have the relation “awareness” to something else. Yet we find so realistic a thinker as Professor Alexander declaring that consciousness is the sole quality of mental acts, and denying that the unconscious, i.e., any process which does not know, is ever mental.

That he is really setting up mind as “that whose nature it is to know” is made quite clear in the account which he gives of experience.note He begins, realistically enough, by asserting that “any experience whatever may be analysed into two distinct elements and their relation to one another”. But, he continues: “The two elements which are the terms of the relation are, on the one hand the act of mind or the awareness, and

  ― 38 ―
on the other the object of which it is aware; the relation between them is that they are together or compresent in the world which is thus so far experienced”. (My italics.) Awareness, then, means both the relation itself and one of the terms.note It is as if we should say that the terms of the relation, paternity, are on the one hand the father or the paternity, and on the other the child of whom he is the father. But the last phrase italicised shows the identity of Professor Alexander's argument with that of Descartes. In each case an account is to be given of knowing or experience, and in each case it is assumed that what is found, by the observer of the experience, to be involved in it is experienced by the person having the experience; e.g., all that I know about your knowing must be known by you in knowing. This is what James calls “the psychologist's fallacy”. Certainly, the two terms are required for the experience, but this does not mean that both are experienced.

If they were, the distinction of -ing and -ed would disappear. Yet Professor Alexander actually uses this distinction to support the view that “the two terms are differently experienced. The one is experienced, that is, is present in the experience, as the act of experiencing, the other as that which is experienced.” In other words, the one knowingly knows, or is known as knowing; the other is known as known. And to complete the parallel with Descartes, we have the statement that “my awareness and my being aware of it are identical”. Now, no doubt, if an experience is experienced or known (though there is no more reason for saying that this must be so, than for saying that a marriage must be married), the knower must be known as knowing and the known as known. But this gives no ground for saying that in any experience the knower knows his own knowing, or that there are two ways of knowing, enjoyment and contemplation, such that the mind “experiences itself differently from [the physical things which are objects to it]. It is itself and refers to them”.note No such identification of the character of a thing and the relation of knowing, or experiencing, is possible. It merely makes “knowing” ambiguous, and resurrects the notion of “that which knows itself”. For mind to be itself is not to know at all; and thus no definite meaning can be found in the terms “enjoyment” and the “quality of consciousness”.

Unless, then, mind can be contemplated by mind and found to have certain qualities, we cannot know minds at all or speak of their knowing. It is precisely the Cartesian type of theory that leads James to argue, since he finds that only one term is experienced (is -ed) in an experience, that consciousness does not exist. But, if so, no satisfactory account of the terms “knower” and “known” can be given; James's theory of intersection would make them interchangeable. It has to be admitted, in fact, that we do observe situations of the sort “A knows B” (whenever, e.g., we take part in a discussion). And this implies that we know A, as well as

  ― 39 ―
B, as having a distinctive character, and not simply as knowing. Such characters of mind are found whenever we say that anyone is angry or pleased or afraid. It is, of course, argued that these characters are “attitudes” to things, i.e., involve relations; that anger, e.g., is always anger at something. Now anger, or any other feeling, has always an occasion, and a man, in being angry, may know what the occasion of his anger is. But he need not do so; it is admitted that a man may not know “what he is angry at”. To say, in the face of this fact, that he cannot be angry without being angry at something, clearly depends on mere prejudice in favour of a theory of mind as essentially knowing. We have, then, empirical grounds for distinguishing between what a mind is and what it experiences; and we see that it is possible both that a mental process should know without being known and that it should be known without knowing (and, for that matter, that it should neither know nor be known).

This view is supported by the Freudian theory of the “unconscious”. The term seems often to mean processes which are not known, instead of processes which do not know. But what is really meant is that the “object” of the process in question is not known. How, then, can it be said to have an object? It has to be remembered that Freud speaks of unconscious processes as “wishes”. Now a “conscious” wish is for a certain state of affairs or occurrence; that is its “objective”. To complete the theory, then, we have to identify objects with objectives, things known with things sought. This, it seems to me, is what is done by Professor Alexander in his “Foundations and Sketch-plan of a Conational Psychology”.note Thus he treats judgment as simply the theoretical form (i.e., the form in which the reaction does not directly affect the thing known) of will, and will as having as its object the state of affairs it is striving to bring about. “In all practical volition the cognitum is a proposition.” “This proposition states the so-called end of the volition and states that end as attained.…The object in question is not necessarily conceived as future. It is the business of the act of will to secure its future existence. What is as a matter of fact future is thus made actual and present” (pp. 265,6). And Professor Alexander holds (p. 245) that “theoretical and practical conation cannot be divided sharply.”

In terms of a theory of this kind, we may say that an unconscious process has a tendency to bring about some state of affairs but has not done so; and we may be able to find out what it would bring about if it were not obstructed or “repressed” (just as we can find out that a person would do something if he were not prevented), and so to describe it as an “unconscious wish” for that state of affairs. But still this is not its character, and it is important to observe that we can know ourselves or other minds as of a certain emotional constitution, whatever this brings about and knows. It is possible that all mental processes are of the nature of wishes, but in order to specify any one such wish we require to know what it is, as well as what it is for. We may know that a man is in a rage,

  ― 40 ―
while his rage has not yet found anything on which to vent itself; and a repressed wish, while it does not attain its objective, can be known to exist and to have definite effects on other processes.

The theory of the wish itself indicates that knowing is not an inseparable feature of mental processes, but at the same time it enables us to give an account of knowing which is in accordance with the plurality of these processes and supplies an answer to certain difficulties. According to the realist theory “the known” consists of independent things in space. But, it may be asked, if what I know when I look at a chair is just something out there, which would be the same whether I looked at it or not (as I certainly take for granted in looking at it), how is it that I know that chair and not other things; how is it again that I know certain characters of the chair and not others; when all these things and characters are equally out there? Must not those objects which I “select” be attached to my mind in some special way, which does not affect those equally present things which I do not know? On this basis, Professor Alexander's description of knowledge as “compresence” might be criticised, for, though he means by “compresence” presence in the same motion, it is a fact that we often recognise things to have been present which we did not notice at the time. The answer to the question depends partly on what has been said regarding the “whole nature” of things, i.e., on taking things in propositions or states of affairs, there being distinguishable states of affairs in any situation whatever. In saying that specific features of our minds “select” specific features of our surroundings, we are only saying what can be said of any two things that come into relation. We can point out, for example, that the Earth and Moon move in relation to one another in terms of their masses, and that all other qualities of either can be neglected. If this mutual selection of masses is said to be our abstraction from the total situation, the answer is that no other type of relation could be stated or conceived; that we know things only as having specific characters and as occupying Space and Time. But the selection which we call “knowing” is made more precise if we can say that we pursue states of the things that surround us and they satisfy processes in our minds. It is still being stated in terms of the relations of two complex things, and leaves “subject” and “object” perfectly distinct and independent. And it is precisely in terms of the complexity of knower and known that an account can be given of error, which cannot be done on the “whole nature” theory. It is required that both knower and known should be changeable and should have internal distinctions.

In short, the foundation of the realist position is logical, and if this logic is not impugned, then, whatever the difficulties of any special problem, it must be capable of being worked out in accordance with that logical basis. A theory of “sensa” or of “consciousness” could not be accepted merely because it enabled us to give a simple account of some limited range of facts. It would sooner or later be found to conflict with a logic of propositions; while that logic itself assists us to give a definite theory of the nature of “subjects” and of any particular class of “objects”.