7 Mind as Feeling (1934)note

Traditionally mind is regarded as having characteristics of three fundamental kinds — cognitive, conative and affective — knowing, striving and feeling. This division is understood in two main ways. According to modern idealistic theory, cognition, conation and feeling are complementary aspects of all mental process, each implying the others and the three together constituting what we mean by mentality; and this is perhaps the most widespread view of the matter. At the same time we find in contemporary thought as well as in earlier theories the conception of the cognitive, the conative and the affective as different sets of mental processes. This is the view implied by Hume, for example, when he says that reason is no motive to action; reason is regarded as purely cognitive and passive, as having no conative or active character, and similarly as having no affective character.

It may be said, in fact, that all English thinkers of the eighteenth century were in difficulties as to the distinction and relation between reason and the passions — with the possible exception of Shaftesbury, who, recognising affections towards affections as well as affections towards outside things, made possible an account of mind as a system of affections. To Butler, on the other hand (and here his views are similar to those of Socrates and Aristotle), there could not be system unless there was a specially systematising affection, one which informed man as to his nature or showed him how he could live up to it. Now Butler, like his Greek predecessors, is unable to maintain a distinction between reflective principles and passions or to show what these principles are; this comes out in the ambiguous position of benevolence in his theory, and in his difficulty in distinguishing the operations of conscience and of self-love. He fails, in fact, to show that any of our motives (tendencies to action) are anything but passionate or how the object of a reflective principle could have any relation to the object of a passion — such a relation, in particular, that the two tendencies could conflict. Similarly, Hume cannot consistently uphold the existence of a reason which is not operative (i.e., which is not already among our actions and capable of affecting others), and which has not particular objects that it pursues in preference to other things and therefore passionately.

Berkeley, again, leaves the passions of the mind in a somewhat ambiguous position in his theory. They are, according to his first division, one main class of “ideas”; but he immediately goes on to argue that the

  ― 69 ―
mind is something over and above ideas (which it may “have”), that it is something of which we cannot have an idea but can only have a “notion”, since it is essentially active and ideas are passive. But this means that the “ideas” he calls passions are no more intimately related to the mind (no more “its”) than any other ideas; or, if he is going to adhere to the view that passions are in some special way mental, he must abandon the doctrine of the mental as an agent, but not an object, of ordinary cognition. The fact that he goes no further into the question of the passions, is a sign of the weakness of his position.


The above preliminary survey indicates that the upholders of the “aspect” theory have a certain advantage over those who believe in separate processes of cognition, processes of conation and processes of feeling. But in the end, in consideration of criticisms of a realistic character, we find that both these theories have to be rejected.

Modern realism is founded on the contention that knowledge is a relation, i.e., that it holds between two things and so cannot be a part of the “nature” of either. The main realist attack has been directed against the conception of what is characterised by being known, or the “idea”. Realism has denied that what we know need be in any way mental or in any way dependent on the mind which knows it (though realists have not perhaps seen clearly enough that the very term “idea” requires to be dispensed with). And it has thus also attacked the doctrine of the absolute idealists that we are what we know, that the whole field of which we are aware (“our world”) is equivalent to our consciousness and to our very selves. It has maintained, on the contrary, that what we know is part of an independently existing order of things, that the existence of a mind is one thing, and the existence of a field of things known by that mind quite another.

But the further implications of realism have not been so clearly grasped by realists in general, viz., that it has equally to reject what is characterised by knowing, or “consciousness”; that it has to say that what knows, as well as what is known, must have a character of its own and cannot be defined by its relation to something else. It has also to reject the whole “self-consciousness” theory of the idealists, who, in upholding the rationalist conception of the knowledge relation as belonging to the “nature” of the things related, brought the whole relation (and both terms of it) within the mind and tried to make a special character out of this internal distinction and relation — tried to make it generate the system which it characterised. This self-sustaining mind must be denied if we take relations seriously, and the rejection of the view that we are what we know must be accompanied by the rejection of the view that we know what we are.

This does not mean that we cannot have knowledge of our minds (apart from the knowledge we can have of other minds); it means that

  ― 70 ―
we are not bound to know all that goes on in our minds. This side of the realist position has been most developed by psycho-analysts and has been neglected by the leading realists themselves. It clearly opens up the field of psychological inquiry, which is narrowed down by the assumption that we always know what we are doing. Unfortunately much of the old cognitionalist psychology still appears in the work of the psycho-analysts, and this is exemplified in the treatment of the unconscious (which we should naturally take as the unknowing) as consisting of processes of which we are not aware (the unknown) — the confusion being concealed in the expression “unbewusst”. When the necessary distinction is kept clear, we can see that it is possible for a mental process (having a character of its own; having, at least, “mentality”, whatever that may turn out to be) not to have the relation of knowing or not to have that of being known or not to have either relation. It is evidently the other view that wants proof. And if we reject the notion of “self-consciousness” (as an attempt to turn a relation into a quality), we clearly cannot argue that a mental process, by somehow “knowing itself”, is bound to be at once conscious (knowing) and known.

Rejection of cognitionalism, then, i.e., of the definition of a process by its relation of knowing, carries with it rejection of the theory of the three types of psychic process. There can be no merely cognitive process, no “reason” such as Hume assumes, no “intellect”, etc. — just as there can be no sensations, percepts, concepts, or other entities defined by the fact that, or the manner in which, they are known. But it also involves rejection of the three aspect theory, because we cannot call a relation an “aspect” in the same sense as a quality, and we are also entitled now to give credence to the evidence which would indicate that some mental processes do not know.


So much for cognitionalism. What of conation or striving? Is it more fitted than cognition to be an actual description of mental processes? It seems clear that, on the contrary, striving is also a relation, implying a striver and a striven for. Of course, if we regard conation as simply meaning activity, then it is just another word for process, and is not a means of distinguishing minds from any other existing things. Taking conation as striving, however, we find in the first instance that the conational theory of mind — exemplified in Alexander's “Foundations and Sketch-Plan of a Conational Psychology”note and again in Freud's recognition of the characteristic mental process as a wish; suggested, also, in the general position of the pragmatists — certainly marks an advance on the cognitional theory. This is particularly so in respect of the theory of knowledge itself (including the theory of error).

Cognitionalism upholds the doctrine of “ideas”, i.e., things characterised by being known, and thus involves us either in the coherence

  ― 71 ―
theory of truth or (when certain realist assumptions are implicitly made; when it is assumed that we somehow know, beyond “ideas”, what they represent) in the correspondence theory. The former theory fails because it, like any other, cannot be consistently unrealistic. If we rightly or wrongly regard a certain idea as cohering with other ideas, we are recognising the existence of actual relations among ideas and are therefore treating them as independently existing things. On the other hand, if to have an idea is to know all about it, if there is nothing more in it than we know, since, as Berkeley and Hume make out, “our idea” is just what we know, then we cannot be wrong about it; and the supposed relations, of coherence and incoherence, are either just ideas among others, or else are outside the region of ideas altogether, and in neither case do we have the adjustment or maladjustment which the theory requires.

The main error of this view — an error which appears also in the correspondence theory and in many other philosophical doctrines — lies in the assumption that there is a kind of knowledge which cannot be mistaken as contrasted with that which can; that, e.g., minds can receive “data” about which there is no dubiety, and can then “interpret” them in various ways which may possibly be mistaken. A sufficient objection to such theories is that they imply a kind of knowledge which both can and cannot be mistaken. Thus the recognition of the fact that “A is an interpretation of the datum B” must be at once the work of the fallible and of the infallible faculty. To avoid this difficulty we have to say that the fallible faculty can also be acquainted with “data”, which must therefore be matters of as much dubiety as “interpretations”. In short, the indubitable cannot be brought into relation with the “doubtful”, i.e., with any real issue, and the solution is that there is no infallible kind of knowledge.

This appears again in connection with the correspondence theory. Merely to have an idea which is like an outside thing is not on the face of it any better than having an idea which is unlike an outside thing; and the latter is not on the face of it error. Error arises only if we think the unlike is like, i.e., if we make a direct comparison between “ideas” and outside things; and this comparison is on exactly the same footing as a comparison between one outside thing and another. Thus the question of “ideas” does not arise at all in the general treatment of error, and it can arise in particular cases only if an “idea” is a certain sort of thing, existing perhaps in a special location (e.g., in a mind), but having the same type of independent existence, being on the same logical footing, as a thing existing in some other location (e.g., in a tree). The correspondence and coherence theories alike fail, then, because they cannot avoid admitting, at some point, the realist contention that we are dealing all the time with independent things — with what exists, whatever its location and character may be.

The doctrine of striving, on the other hand, permits us to distinguish cases where we are mistaken from cases where we are not. We are right,

  ― 72 ―
it may be said, when we get what we strive for and wrong when we do not. Error, at least, is comparable to missing one's mark (mis-taking); and here the Freudian theory of errors as satisfactions of (unacknowledged) wishes is important. We are in error when we treat A, which is not B, as if it were B; when we mistakenly use it as a B — e.g., when we treat a red danger signal as if it were a green safety signal, or treat as a friend a man who will actually deceive us. This clearly implies a certain knowledge of both A and B, i.e., error is impossible without knowledge; knowledge is primary, error secondary — but, of course, that is actually the case. It is to be noted that we are sometimes undeceived and sometimes not. The fact that we do not always find out afterwards what mistakes we have made, and the fact that we do not always know what we are doing or what our actions will lead to, are among the obstacles to a thorough discussion of error.

Error, then, is exemplified in misuse of things (using a pruning-knife to cut down a gum-tree or a sledge-hammer to crush a wasp), and arises in our striving, with the means at our disposal, to satisfy our wishes; or, as I should put it, in our motives themselves, our tendencies or mental tensions, striving to find an object, to find outlet or release. We believe what eases our minds,note whether it is true or false. Freud's theory of the wish and Alexander's identification of the theoretical and the practical (treating judging, e.g., as Descartes also does, as a form of willing) have prepared the way for this development of psychological theory, in spite of incidental cognitionalist confusions. The theory of the mental tension appears also in James's account of the “active gap” in our “consciousness” when we are striving to recall a forgotten name. And if we take all knowledge as discovery,note then we have a general recognition of tension; and that is known or found, we may say, which releases the tension and sets our minds at rest — or, at least, a part of our minds; we have, of course, many quests, not only successively but at any given time.

This conational view is also in accordance with the whole modern doctrine of interests, as guiding our lives and constituting our mental character (though acceptance of that doctrine has to be distinguished from the adoption of any special theory of “instinct”). Even those who speak of “intellect” have still to recognise “intellectual interests”, a “passion for truth” and the like. But when all has been said that can be said on the conational side, we find that we have still not learned what are the qualities of mental processes themselves, what it is that may be in tension or relaxed. Striving, like knowing, is a relation, and the mental quality (mentality) is still to seek.

  ― 73 ―


This brings us to the third element in the original classification — feeling (affect, emotion). And this, I think, enables us to solve our difficulties; it gives us a basis for a general descriptive account of mind, i.e., we can recognise “affects” as real qualities of mental processes (or, what comes to the same thing, as real mental processes), whatever relations, knowing, striving or other, they may have to other things.

McDougall (in Social Psychology) does something to suggest this view, but he makes a very unsound division of mental process into a cognitive (leading up, afferent) part, an emotional (central) part, and a conative (efferent) part — as if the process as a whole could have the relations of knowing and striving only by having a knowing part, etc., and as if the central part were not as definitely related to outside things as afferent and efferent nervous processes are. In this division, in fact, we find traces of cognitionalism and also of physiologism, i.e., of the kind of view which expresses mental facts in terms of the physiological processes to which they are related (as knowledge is called “sensory” because it is related to sense-organs) — though the rejection of physiologism does not imply that mental processes are not themselves physiological. Nevertheless, for McDougall, the central and most distinctively mental part of the process is emotional; and so, if it is mind that we regard as having the relations of knowing and striving, we may go on to express the position by saying that emotions (or feelings) know, emotions strive and, in general, interact with other things.

Interacting is, of course, something that all things can do; and it may be that the relations which we can in the end recognise as knowing and striving are not peculiar to mind at all (cf. Alexander's account of knowing as “compresence”), and that we have thought otherwise solely because we have thought into the mind's relations something of its own emotional quality. It is enough, however, for our present purpose that an emotional thing can have these relations; and from this point of view we should reject the cognitionalist formula of Freud about “an idea becoming charged with affect” and speak instead of a feeling finding an object.

Such expressions sound harsh, but only to ears attuned to cognitionalism. The real confusion comes from the opposite quarter, in the recognition of “reason”, “intellect”, “sensation” and so forth — as when McDougall, after making the instincts the native mental forces, introduces a complicating factor of reason,note instead of recognising that what reasons is simply a complex of the emotional activities he has already dealt with, and not a new faculty springing from nowhere. Once we have rejected “constitutive relations”, once we have seen that what knows and reasons must have qualities of its own, we can say that emotion is as likely to know as anything else; we cannot reject it as a possible knower merely because it is a quality, since we should be thrust back on the supposition of something which consists in its relations.

  ― 74 ―
The objection that feeling also is a relation comes from the same cognitionalist source. On the theory of mind as feeling it is not denied but asserted that feelings have relations and hence that we can be angry at, afraid of, pleased with, something or other (this being to say that we are angry, knowing something, or that our anger has an object, i.e., knows something, etc.). But to have a relation is not to be a relation. The term “feeling” has certainly been employed to signify some sort of “immediate experience” (erlebnis) in which knower and known are one. This appears in Alexander's use of the term “enjoyment”, and also in the use of the term “feeling” (“sensation” being sometimes substituted) by such thinkers as Bradley, but especially by James, who, in his Principles of Psychology, speaks of feelings of red, of green, etc., and again of feelings of and, but, etc. Such views fall with the rejection of relativism (the belief in constitutive relations), whilst at the same time they testify to a certain recognition of the “inwardness” of feeling. What should be admitted is that feelings (e.g., anger and fear) are qualitatively different from one another, though they still have the general feeling-quality in common. It will not surprise us then to find that, besides having similar relations to outside things, they also have different relations to these things, e.g., that they have different objects, that one seeks what another avoids, etc.

The thorough-going rejection of cognitionalist doctrine involves the recognition of the following facts: (a) that a mental process may exist in us without our knowing it (as when we find out afterwards that we were angry or afraid); (b) that a mental process may exist without knowing (as in what are called “nameless fears”); (c) that nothing mental is simple or passive, but that we have a vast complication of tendencies (tensions) which pass through one another, and become variously organised, in pursuits and aversions, strivings and capitulations, sentiments and interests of all descriptions; that “intellectual pursuits” are thus operations of the love of truth (the inquiring spirit), developing from original scattered curiosities — for we have no reason to suppose that all curiosities are parts of one curiosity, all angers the work of a single faculty of Anger, etc. We thus have a conception of mind as a society or economy of impulses or activities of an emotional character. This conception of our “motives”, conscious or unconscious, as emotions will, I am convinced, give coherence to psycho-analytic doctrine, and, though this is a point of less immediate importance, will be found to work in with the physiological examination of those brain processes (in their relation to bodily processes in general) which are “the seat of” the emotions, i.e., which are the mind.

We may here refer to the James-Lange theory (which is, of course, of a cognitionalist character) of emotions as “sensations of certain bodily processes”. In the first place, “organic sensations”, regarded as objects, are the organic processes themselves and thus are distinct from the special class of emotional or mental activities, even though the latter are also organic; e.g., if we mean by hunger what is going on in the stomach, it is not mental. On the other hand, if by “organic sensation” we mean

  ― 75 ―
what knows the organic process in question, then we have still to be informed where and what that knower is — what qualities it has. It is not, of course, the case that we are sorry because we weep, etc., but it may quite well be the case that, on a given occasion, we find out that we are sorry by noticing that we are weeping, find out that we are afraid by noting the condition of our breathing, and so on. We are not, as has been seen, bound to know directly what is going on in our mind, and in such cases we may discover it inferentially. But this inference depends on the previous observation of a connection between the two, between the (central) emotional process of sorrow and the (peripheral) process of weeping; which implies that we have a direct acquaintance with both terms of this relation.

The same circumstance, that we are not bound to know what is in our mind, accounts for the fact that many emotions have no names or are named only from their objects (e.g., as “love of” something), these objects, which the emotions themselves are interested in, being naturally what is known when they are operating, and attention, by other emotions, being directed to the former emotions only rarely — e.g., when they are obstructed. Progress in psychology may therefore be made by the actual discovery of the emotional character of sentiments or motives, i.e., of what is in our minds, as contrasted with what is before our minds, when we engage in certain pursuits.

These, however, are matters of detail. The main points are: (1) that knowing and striving, as relations, cannot be the character (“mentality”) of the mental; (2) that feelings, as qualitative (a point which is illustrated by the qualitative distinctions among feelings), are capable of characterising the mental — as well as of having relations to other things; (3) that we must assume that they do, that feeling is mentality, unless we are going to suppose that some entirely unsuspected character of mind has yet to come to light; but (a) as we have seen, there are multifarious suggestions that feeling is already, if only confusedly, recognised as the mental quality;note and (b) we do recognise and speak of minds and therefore we must already have recognised some mental quality. To say that we know mind only as “that whereby” certain effects are produced or arrangements made is to say that we do not know mind at all — for how, except by observation, do we know what sort of thing would have these effects or that there is a thing, of some peculiar kind, to have them? In fact the rejection of the belief in constitutive relations implies that to know a thing is to know some of its qualities. We are thus in a position to say that mind is feeling, and that it is such feelings that have objects.

Having this basis, we can go on to discuss how feelings develop and interact, how they are affected by the bodily organisation in general, and how by things outside the body, including feelings in other minds — in which connection McDougall's theory of “sympathetic induction” is of

  ― 76 ―
interest — and for this discussion we shall also have to take into account how they affect these other things. This discussion (including, of course, an account of what various feelings there are) will be psychology.


Some further remarks on the connection between belief and behaviour may help to remove difficulties in regard to what has been said about conation. It was contended that on the view that it is certain interests that know, certain feelings operating so as to secure “outlet” or relaxation of tension, an account can be given of error and other difficulties which the ordinary theories of knowledge cannot meet. In terms of the formula that we believe what eases our minds, it appears that we shall refuse to accept situations which we find intolerable. Now this attitude of opposition to situations which occasion us uneasiness or distress can operate in various ways according to personal as well as external factors (the breadth of our interests, the difficulty of the situation, etc.). The simplest case is that in which our uneasiness expresses itself simply by removing the obstacle, so that uneasiness ceases, i.e., we are satisfied with what is the case, and there is no question of error. But, since error does occur, we require a more general formulation of the position, covering all cases.

It may be said, then, that there is in general a tendency to remove obstacles; or it may be enough to say that there is striving. Now, in terms of the conational theory outlined, what is striven for will be identical with what is believed. As Alexander points out, if we will that a certain person should leave the room, then the situation we have before our minds as occurring is just that person being out of the room; the object of our judgment is identical with the objective of our willing. The success of the judging, the “click” of conviction, the “satisfaction that” the event is so, is identical with the success of the willing and the relaxation of tension.

Obviously there may be complications in the case; while having as our object the person's absence, we may recognise that he is not yet gone. But, in the first place, we can have opposing interests and thus opposing judgments. And, secondly, we can recognise phases in a situation; we can find the person's presence temporarily tolerable (and therefore admissible as a fact) because it is a stage in the process leading up to his absence, when we shall be quite satisfied. In other words, we can distinguish a main objective from subordinate objectives which are means to it or preliminary conditions of it, and on the theoretical side this is expressed by saying that we recognise not merely single situations but implications (or, more generally, the passage of one situation into another), this being the state of affairs to which the term thinking is ordinarily attached, just as conduct is commonly taken as the carrying of a line of action through a series of stages.

The general position is, then, that we regard what we want as brought about, and if our action (or simply our situation) is such as to bring it about — if we get what we want — then we have a true belief. In certain

  ― 77 ―
cases, however, our action is unsuccessful; and then, though there are various possibilities according to the variety and development of our interests or demands, our primary tendency, Alexander contends (in line with the views of the Freudian school, particularly as expressed by Freud himself and by Ferenczi), is to regard the wished-for result as brought about. That is to say, we obtain a certain satisfaction or release of tension under a condition of hallucination or illusion. Such a condition, it is to be observed, is not logically different from error in general. To suppose that it is is to suppose some objective situation of a simpler type than the proposition (hallucination being conceived, e.g., as a “datum” which does not “correspond to” reality).

One possible way of obtaining ease of mind or release of tension, then, is the false belief that the object has been achieved; but the fact that it has not may result in a continuance or re-establishment of tension. We know that this is not bound to be the case; there are people who retain false beliefs throughout their lives or at least are never disabused of certain errors — indeed, this could be said of any of us. But still there are cases where our satisfaction proves evanescent, where we are undeceived, and the demand for the thing itself, for “real” as opposed to “hallucinatory” satisfaction, is reinstated. In such cases the object may eventually be obtained or we may secure some substitute for it — in any case, the tension has to be diminished in some fashion, or the position really would be intolerable.

This process of finding substitutive satisfaction has been dealt with in detail by Freud and his followers. The point to be emphasised here is that it is just the process of developing a theory of things, so far as we do develop a theory; and it may be incidentally observed that the distinction of “illusion” from “reality”, i.e., the recognition that we are capable of falling into error, is a considerable theoretical step. But, as Freud has also shown, unsatisfied tendencies may remain in a state of subdued tension, of repression, in which they do not secure outlet but, on the one hand, draw away energy from the operation of the interests in general, leaving the person comparatively inactive, and, on the other hand, interfere with the other interests, altering the direction of their activity, and thus precipitating (a) those confusions and mistakes which Freud deals with under the heading “Psycho-Pathology of Everyday Life”, and (b) the dream form of hallucination. We can distinguish, then, a number of different types of reaction to a unsatisfactory situation — there is simple hallucination; there are various forms of self-deception and confusion, whereby we contrive to hold contradictory beliefs at the same time; there are various methods of substituting one object for another; and finally there is the possibility of such a rearrangement of tensions (i.e., such a development of the mind) that repression and dissatisfaction are overcome — though we may admit that this development will never go so far as to enable all our tendencies to find outlet.

In line with these distinctions, we may distinguish from simple error various forms of “interpretation” of the things we deal with — as when

  ― 78 ―
it is said that we interpret the unfamiliar in terms of the familiar. Here, on the one hand, we can have that dictation to Nature, or looking for simple solutions, of which Heraclitus accused the Pythagoreans, and which is bound to land us in error; but we can also have simple insistence on special uses of certain things, emphasis on some aspects of them at the expense of others — a preference which need not in itself amount to misuse or confusion. Granted that we have interests and that it is these interests that operate on things and give us our knowledge, then it is naturally the case that we select certain characters of things for special attention and neglect others, and thus that we select those most familiar or easy to deal with or most satisfactory to us. This selectiveness is in fact the condition alike of error and of discovery.

Now in this connection Heraclitus upholds the replacing of desire by understanding, and similarly Freud speaks of passing from the pleasure-principle to the reality-principle, getting an interest in things as they are as contrasted with what we should want them to be. Clearly we must be wary of over-stressing such a transition, since understanding, or adherence to the reality-principle, is still the operation of an interest, and desire (or what follows the pleasure-principle) even in the first instance finds things to be the case, whether correctly or incorrectly. It is not a question, then, of a theoretical interest coming out of a non-theoretical interest. We may admit that peculiarly scientific interests can be developed, but we shall still have to say that, as interests, they have special objects, and that there are special conditions of their finding outlet. It is therefore not surprising that we find throughout the history of scientific inquiry a recurrence of that dictation to Nature (as contrasted with “expecting the unexpected”) which Heraclitus condemned in the Pythagoreans.

Of course, inquiry, confused or otherwise, expresses itself in saying that something is the case, and such an assertion must be met in the first instance by showing that it is not the case; only after that has been done can we go on to the question of how the particular inquirer came to prefer his erroneous view. But, while that is the form of argument, of refutation or proof, the carrying out of arguments or inquiries is still the work of our wants or interests, and the notion of a “dispassionate reason” rests on a confusion between the objectivity of the issue (i.e., the simple contrast between truth and falsity, occurrence and non-occurrence) and the activity of the inquirer into that issue. And along with the notion of dispassionate reason we must reject the notion of “reasonable conduct”, or conduct enjoined by reason. The most that can be meant by reasonable conduct is the thing that is to be done, i.e., what is demanded by certain interests; and, since this will obviously differ from what is demanded by other interests, we see that “what is to be done” is a relative expression and that there can be no absolute “duty”. It appears, then, that the clearing away of cognitionalist confusions is also of advantage to the ethical inquirer — as it must be to any inquirer whose field includes any part of human behaviour.