Seeing Other Minds

By Steven M. Duncan

That there are persons existing as material things with a full complement of physical properties seems hardly controversial. However, there may be some disagreement about whether we should admit to persons predicates of a nonphysical kind. The usual candidates for these are the mental ones, and some philosophers and psychologists have been loath to admit them into their ontology because of their allegedly subjective, private and unobservable character. Even if we are willing to admit them in our own case on the basis of introspective evidence, it has been known since Descartes that, if the mental life of others is wholly private and unobservable in principle, there are serious problems about the justification of our belief that they have them, even if the belief is, for all practical purposes, unshakeable.[1]

There is a further point worth noticing. Most of us would claim to possess a good deal of knowledge, both general and specific, about other persons, especially the persons to whom we are closest and can be said to know best. Indeed, common human experience suggests that, just as we often understand others better than they do themselves, so too do others understand us better than we do ourselves. This, however, is the very opposite of what we would expect to be the case if knowledge of our own inner lives were transparent and that of others completely opaque to us. If our beliefs about the inner lives of others were as tenuous and crepuscular as the doctrine of inference suggests, we could hardly put as much trust in our judgments of others as we ordinarily do.

The case would be entirely different if at least some aspects of other minds were publicly observable. For then even the most empirically minded philosopher could admit mental events into his ontology without embarrassment. Further, given that we would have some evidence of a direct kind that persons other than ourselves have minds, the inferences we would make to those mental characteristics of others which we decidedly do not see (e.g., episodes of thinking or imagining) would be considerably strengthened. If we can know directly that other persons have some of the aspects of mentality without inference, it makes the inferences that we do have to make somewhat less egregious.

All of this suggests that our knowledge of the mental states of others is direct rather than inferential, and that our knowledge of other minds best understood on the model of our commonsense ideas about perceptual knowledge. This draws additional support from the fact that in ordinary speech we often use perceptual locutions to describe our awareness of the intentional states of others. For example, we often say things like "I could see how sad he was sad" or "I saw anger in his face" or "pride in his bearing." These locutions suggest that our knowledge of at least some aspects of other minds might be understood on a perceptual model. However, I take it that many philosophers would be inclined to reject the idea that I could see, in any literal sense of that word, the feelings of others. After all, they might say, perceptual verbs like "see" do not always indicate perceptual acts; why should we take them to do so in the case of feelings?

This essay is largely an attempt to sketch an answer to this question. I shall argue that there is a literal sense in which we see certain of the mental states of others, as these are expressed in externally observable behavior, just as they perceive our mental states in experiencing our behavior. Further, the mental states that are perceived in this manner are those expressible in and through our behavior. Nor does this involves inference, and that this is part of what is given as part of our experience of the behavior of others and their experience of our behavior. In so doing, I will be relying on the pioneering work of Virgil Aldrich (1904-1998), who first developed this account many years ago.[2]


Let us begin our investigation by drawing an analogy betweenpictures and persons. Beginning with pictures, we might note that there are at least two ways in which we might regard a picture. From the viewpoint in which pictures are regarded as material things, we would say that a picture is a two-dimensional array of dried paint-blobs covering a canvas stretched on a wooden frame. The picture, so viewed, has length, height, weight, location in physical space and a determinate physical microstructure. Everyone knows, however, that there is more to a picture than this, for a picture can also be viewed with respect to its content, i.e. what is to be seenin the picture. Viewed in this way, the picture has a sort of space, or constitutes a perceptual field, in which things appear and can be seen. Thus, I may look into a picture (a Norman Rockwell, say) and see a collection of individuals waiting at a bus stop - and what a collection they are! There we see a businessman, looking at his watch, tired from his day's work and obviously irritated that the bus is late. We also an old woman of the sort one often sees on the bus, to whom the years have not been kind and who sits on the bench provided by a local funeral home as though it were no concern of hers whether the bus ever arrives. There is also a plain young couple, staring straight ahead, holding each other's hand. We note that the young man looks pensive and even somewhat rueful, whereas the girl's expression radiates triumphant bliss. Then we turn to our last character, a young boy in a striped shirt. He wears a baseball cap and carries a glove; a ticket to the evening's game pokes conspicuously out of the back pocket of his jeans. He leans nonchalantly against a tree, and kills time by blowing a huge purple bubble out of an enormous wad of gum in his mouth.

This is what we see, I am supposing, when we look into some particular picture. Now it seems to me that the use of "see" or "seeing" here is a perfectly ordinary one; in this sense, to see is simply to perceive by means of the eyes. There is nothing private or subjective in my claim about what is to be seen in the picture: quite the contrary, this judgment can be confirmed by anyone with good eyesight simply by taking a look and thus adverting to evidence available to anyone.[3] I shall call the picture1, when considered as a material thing, the picture, and the picture considered with respect to what is to be seen in itqua intentional field, the picture2.[4]

Now, what is the relation between the picture2, and the picture1? No doubt, the first thing that needs to be noted is that the picture1 has all the material features or properties I mentioned before, while the picture2 has none of them. The picture1 may weigh seven pounds, but it would make no sense to ask what the intentional field of the picture weighs. Another significant difference consists in the fact that the picture2 is three-dimensional, within which all sorts of spatial relations are present which have nothing to do with physical space. I may truthfully say that the man in the picture2 is "within easy bowshot," even if the picture1 is so close that I wouldn't even be able to get the arrow off.[5]

Such are the primary differences, which make any reductionist account of the picture prima facie unlikely. At the same time, it seems to me that we must not overlook the intimate connection between the picture2, and the picture1. Indeed, the connection here is so intimate that it seems plausible to think that it is only in virtue of the picture having or being a picture2 that it rightly is called a picture at all. A picture could not represent anything unless it possessed an intentional field in which things could be seen in the way we have described. At the same time, the picture2 exists only as a function of the regional qualitiesof the material thing that occasions it; in other words, there cannot be pictures2 unless there are pictures1, as well. Neither, then, seems capable of existing, i.e., being what it is, apart from the other; as such, it hardly seems that a dualistic account of the picture will do. The alternative then would seem to be to allow that the picture1 and the picture2 are both instantiated in one and the same individual, such that "picture," when used without any qualification, refers to the unity made up of these two irreducible aspects.

Now we need to work the analogy with respect to persons. Here the essential idea is that the body of the person (i.e., the person qua material thing) functions like the picture1; it literally bodies forth or puts on display the passionate and affective states of the person, like anger, joy, and depression. On this account, the body of the person has/constitutes an intentional field where certain of the affective states of that person appear or present themselves to be seen. In this respect, then, the body of a person functions or is constituted as the analogue of the picture2. Thus, when we look at a living person we see not just the exterior surface of a material thing, but also the affective states that animate that body that are given along with it as part of the same perceptual experience. Again, "see" here means simply "perceive by means of the eyes," and the intent is that facts about the affective states of individuals will be just as much matters of perception as any other.

This suggestion can be clarified by reference to my earlier example, the (imaginary) picture of the bus stop and its denizens. What we primarily see, in being aware of the content of the picture2 are representations of material things, the surfaces of the bodies of persons. In this respect, the content of the pictorial field looks like (resembles) things that we might see directly in the ordinary perceptual space of the everyday world. (For example, we could use a photograph of someone we have never seen in order to identify him in a crowded room.) However, the representations do not consist simply in this. For recall that we described the businessman as irritated, the old woman as woebegone and unconcerned, the young man as rueful, the girl as blissful, and the boy as nonchalant. At the time, these judgments seemed just as natural and objective as the others we have been making concerning the per­sons in the picture. Suppose, then, we ask ourselves on what these judgments are based. Again, it seems to me that it must be what we see when we look into the picture; and so it is, when we consider how we would justify these judgments. In the case of the businessman, for example, we could point to the rigidity of his stance, the scowl on his face as he looks at his watch, the way he clutches his briefcase tightly in his other hand. In the case of the old woman, we could note her huddled, beaten posture, the slackness of the muscles in her face and the unfocused eyes that combine to create a blank expression. The young man's eyes are squinted and his lips are tightly pursed; the girl's eyes are wide open, glassy and her mouth stretched in a smug, self-satisfied smile. It is the boy's posture that gives him away, as he leans supinely against the tree, hands in pockets, lazily blowing his bubble, a look of indifference on his face.

In short, attending upon the faces and bodily attitudes represented in the picture are certain visualgestalts or regional qualities of an expressive kind. The persons in the pictures are expressive representations, which reveal not simply visage and visual form, but also attitude, emotion, and character. They do so, I suggest, by reproducing in the intentional space of the picture the very same visual gestalts or regional qualities which we see attending on the bodies of persons when we observe them giving expression to their attitudes, affective states, and character traits. If this is so, then our perception of other persons is not limited simply to an apprehension of the surfaces of their material bodies considered as such. Rather, they are the occasion for certain regional qualities of an expressive kind to supervene upon configurations of those bodily parts. (Indeed, we might note in passing that those expressive qualities do not need a human body to occasion them; a skillful cartoonist can do it with a stick figure or a line drawing.)

Of course, when thesegestalts appear in pictures supervening on representations of those configurations they function as representations, or parts of representations, as one element in the overall effect by means of which the artist produces, e.g., a picture of an angry man. When thesegestalts appear in the bodies of persons, however, they are not representational, but merely expressive. The visual gestalts expressing the inner states of persons are not representations of those states, but rather (as I have already suggested) those states themselves in their publicly observable aspect. Thus, I cannot find it in myself to agree with Wittgenstein's oft-quoted dictum that "the body is the best picture of the soul," since this suggests that the intentional space of the body is like a mirror on man's inner life rather than a simple presentation of it. Opposed to this, it is this latter position that I should like to defend.

Now I suppose that this hypothesis may strike some of us as utterly fantastic; however, let me see if I can induce the proper intuition here by reference to some reflections about what we normally see when we view pictures and persons. To begin with, it must be admitted that what we see for the most part are surfaces of material things, such as tables and chairs. It is therefore perhaps natural to think that what we always see are the surfaces of material things just as such. However, if the argument of this paper to this point has been correct, this belief is quite erroneous. When we look at pictures, mirrors, reflecting pools, photographs, and so on we almost never see the surfaces of the material things that we perceive. Instead, what we habitually see are the contents of the intentional fields these objects possess and put on display. Of course, we can see the surface of a picture1 if we work at it, i.e., we can collapse the three-dimensional field of representation and its contents, the picture2, into a two-dimensional array of variously colored paint blobs by squinting, or getting so close to the surface of the picture that we can no longer see the content but can see the weave of the canvas, the cracks in the pigment, and the brush-strokes which create the effect. (A better trick is to see the surface of a mirror without attending to its content -- something I simply cannot do.) This, I think, is what it would mean to see the surface of a picture, or to see it as a picture1. However, this is not at all what it means, in the ordinary scheme of things, to see a picture. Instead, this would seem to be a trick we learn to perform by prescinding from the intentional contents we normally see in order to constitute the object in a different manner in our perceptual experience.

The same, I think, holds for the person. Normally, seeing a person is more like seeing a picture than it is like seeing a stone or a chair. Persons as we typically experience them present themselves to us as naturally expressive of a living principle that animates them. To see a person qua material thing would be to see the physical surface of the body of the person without attending to the expressive element present there, i.e., to see persons as bodies in motion and actions as bodily movements, in such a way as to ground only physical descriptions of what is occurring. It would seem that this is something that can be accomplished. A smile can be seen as a two-dimensional array of color patches, especially when we view it up close, concentrating on the pores and the wrinkles in the skin, and described in terms of relative distances between the lips, their length, the surface tension of the skin, and so on. Yet this hardly counts as being what, in the normal case, it means to seea smile. To see a smile as a smile is already to see it as expressive behavior, embodying happiness, or embarrassment, or wicked delight in seeing others suffer, or what have you. This, I submit, is the character of our experience of persons in the ordinary case, as opposed to the act of precisive abstraction that constitutes the body of a person as simply as series of postures or gross bodily movements.

A similar point could be made with respect to the role of inference in our seeing of pictures. It would certainly be a fantastic hypothesis, not to be adopted without serious cause, to suppose that we infer to the intentional content of pictures from seeing the physical surface of the picture qua material object, since it seems to conflict with the evident facts of experience that first, it is the content, not the surface of the picture which we immediately see in the normal case, and second, the drawing of such inferences does not seem to play any part in our experience of pictures. It seems to me no less fantastic to think that we infer to the feelings of others from observing bodies in motion or muscular movements, and for the same reasons. For in the normal case, our behavior is perceived as already expressive, as having intentional content. Further, if we attend to our experience, I think that we will not find that we in any conscious way draw inferences of the kind described by those in the Cartesian tradition. Of course, one may wish to claim that we cannot justify our knowledge claims in this area without appeal to inference. This, however, is another issue.

It will be clear by now that I do not regard myself as a reductionist, attempting to collapse the mental into the bodily and ultimately into the physical. Rather, I want to use "behavior" as the cover term for what we perceive when we see persons in the normal fashion. "Behavior" used in this way retains the idea that what we are doing is seeing or observing, notinferring in any sense which does not turn all perceiving into inferring. However, it rejects the view that what I see or observe are bodily movements or changes rather than expressions of affective states. In fact, where behavior of this type is concerned, "bodily movements" and "inner feelings" are related like the picture1 and the picture2 and are, when treated by themselves, mere abstractions. Each depends on the other for its being what it is; the body of a person is such only by fact that it is naturally seen as expressive of an animating principle which consists in such cases in an affective state. On the other hand, that animating principle and those affective states could hardly exist without a body to occasion them.


So much for the model, such as it is. No doubt, there is much here which still requires clarification. However, perhaps I have said enough already on my own behalf, and it is now time to consider some of the objections which by now have suggested themselves to the reader. In considering the objections to my view, I shall proceed as though an opponent and I were engaged in a sort of dialogue. I shall now give the hypothetical opponent his first chance to speak, and suppose that he begins as follows: "Nothing in your position so far really distinguishes you from the so-called crude behaviorist. Indeed, it seems that you have left out of account something absolutely essential to any non-reductionist account of the person. For, after all, is there not, in addition to the angry behavior, however "expressive," the feeling of anger, a subjective mental event happening only inside the person somewhere? The introspective evidence of ordinary experience seems to demand that we admit such a category of events into our theory; yet to this point, no provision has been made for it."

This is not all there is to the critic's complaint. On the contrary, he really hopes to catch us in a rather nasty dilemma. If we admit subjective mental events into our theory, then it seems that we admit that there is an element of the affective state of the person that we cannot observe, even in principle. On the other hand, if we deny or discount the presence of such subjective mental events, we shall be in a position that not only contradicts the whole weight of our experience, but also gives the lie to our attempt to avoid the reduction of behavior to mere uninterpreted bodily movements.

Now the short way with this objection is, I think, to grasp the first horn of the dilemma and try to "throw the bull." On this strategy, one admits what experience seems to require, i.e., that there are such things as subjective feelings, but argues that no disastrous consequences follow from this. For on the view developed here, I think it will be enough to turn the objection to assert that it is precisely the subjective feelings of individuals that are expressed, or put on display, by behavior. That is, the subjective feeling itself is the animating principle making what would otherwise be characterless, meaningless bodily activity into expressive behavior. Thus, the subjective feelings of others are publicly observable for all of that. This, however, does not cancel their subjectivity. It is still true that we do not perceive the feelings of others in the way that they do; and we do not feel their feelings, for if we did, they would not be uniquely their feelings, occurring in the stream of consciousness to which they have privileged access. Nevertheless, it does not follow from this that we do not perceive them at all. On the contrary, we perceive them by seeing them in and through the behavior by means of which they are expressed.[6]

That is the short answer. However, I suspect that it will not find favor without some further clarification; it is perhaps a defect of my view that I have not said more about what it means, e.g., to feel joy or anger. Let me repair this defect by appeal to a further speculation, adapted from Austin Farrer's discussion of action in his book The Freedom of the Will.[7] Farrer follows Ryle in giving a dispositional account of actions. The disposition identical with any action, however, is not simply one of mind or brain but rather of the body as a whole. The root of the action is in the brain as its cause, but the action itself as what Aristotle would have called its first actuality is spread throughout the body and resides as much in the nerves and muscles which will be called upon in the action as it does in the brain. This diffuse disposition Farrer calls the action pattern.[8]This notion can be clarified with respect to a simple example. Suppose you are roaring up to an intersection when the light goes yellow, you still being a hundred feet away. One's recognition of the meaning of the yellow light galvanizes your body into action, and you become primed to do one of two things: slam on the brakes or push harder on the gas pedal in order to get through the light before it goes red. Here are two action patterns coming to life or passing from first to second actuality; we feel them both as first anticipatory "tingles" in our nervous system. Our exercise of free choice is in our decision to activate one of them and to extinguish the other, to press the brake if we are prudent, or the gas pedal if we are daring.

In Farrer's analysis of the action-pattern, we have seen that it contains three elements. First, there is the context in which the bodily response represented by the action pattern becomes relevant, or in which it occurs. This context need not actually obtain, but the person who is "galvanized into action" must so believe. Second, there is the action pattern, pre-existing in the body of the person as a disposition to be galvanized in a certain way. Thirdly, there is the action itself, the second actuality of the action-pattern that is publicly observable.

Now it seems to me that much the same model can be applied to the case of the passions. Once again, we want to distinguish three elements or "moments" in our analysis. First, there is a context in which, for some individual, one of the passions, e.g., anger, is aroused. This context will consist of actions, beliefs about those actions, and perhaps other kinds of facts as well. Second, there will be the anger, pre-existing in the body of the persons as the disposition to become angered in certain situations. Finally, there is the "being angry," i.e., the second actuality of anger consisting in the full-blown explosion of the body into angry behavior.

Now, on this model, what the feeling of anger or the feeling angry consists in is a kinesthetic awareness of the passion's "coming to life" in the body. We can feel what others can see, i.e., our nostrils flaring, our chest expanding, our face reddening, shouts as they explode from the larynx, and so on. The "coming to life" of the passion in us and our feeling of that movement begin at the same time and grow up together. They are related not as cause and effect, but (once again) as picture and content. On the other hand, it is by virtue of the feeling that it expresses that the bodily source is able to be a bit of behavior, rather than mere bodily movements without sense or character -- and how different our experience of persons would be without them!

The position I have been sketching here is far from obviously true, and for all I know it may be quite controversial. Nothing hangs here on whether this speculation is true. I simply want to propose that there are conditions under which it would make perfectly good sense to say that I can perceive the feelings of others, by seeing them in the behavior expressing them. So, then, the subjectivity of those feelings does not bar our perceiving those feelings, and so does not represent a problem for the view developed here.


Without doubt, there is more to be said on both sides of this dispute, but I think that this is not the place to say them. Instead, let me allow my hypothetical critic to drop his present line, and onto another one, which is equally important and which in the end may be more fruitful. In this section, I wish to consider another important objection which might be urged against the position I have been developing; to do this, I will put some more words into the mouth of the hypothetical critic: "According to you," he may say, "my seeing that someone is angry or depressed is a phenomenological "given" or a perceptual fact. Even so, error is possible where my ascriptions of mental states (e.g., feelings) to others is concerned. I do not always successfully tell the moods of others, or report them accurately. However, if this is a matter of perception, it is very difficult to see how error could be possible. For, after all, is not the fact that error is possible which serves as the primary evidence for the claim that perception, of whatever kind, involves inference?"

To begin with, the notion of error itself needs to be looked at. It is true that epistemologists in the past have attempted to exclude error in the perceptual context by excluding inference, and that in large part this sort of approach has been abandoned. However, it does not follow from this that wherever error is possible, or actually occurs that we have a case of faulty inference. Take arithmetic calculation as an example. Suppose a clerk adds a long column of figures and gets the wrong answer. Is the error here best explained on the model of a faulty inference? Well, perhaps something could be worked up on those lines, but again it seems to me that this is certainly not our ordinary way of viewing such errors. Rather, mathematical operations areprocedures that if properly employed within accepted limits, can be counted upon to yield the correct answers to problems. Mathematical errors, then, are mistakes, that is, sloppy use or outright abuse of the methods and procedures that constitute mathematical calculation. It is for this reason that the mere fact of error in mathematics is never regarded as a good reason for thinking that the procedures or methods themselves are in any way doubtful or untrustworthy. Instead, we identify the source of the error with the person who makes (careless or incompetent) use of the methods and procedures. Our unwavering confidence in them is demonstrated by the way in which we attempt to uncover those errors, which is basically to reapply those procedures to the problem, this time very carefully and self-consciously, with an eye to those places at which the crucial slip may have occurred (e.g., "Did I carry the two?").[9]

Something like this seems to me to apply to perception itself considered in a general way. Perception itself consists of a set of procedures or techniques for the proper use of the senses in acquiring information about the external world. The careless use or misuse of these will lead one into error. These errors are also mistakes, and can (at least in principle ) be detected by using the senses in order to check themselves. Perceptual errors of this kind are possible because sense perception occurs in a context provided for it by physical mediating factors in the perceptual situation, such as the amount of light and physiological facts about the perceiver himself. There are mistaken perceptual judgments, e.g., about what we see, but the great majority of these can be eliminated in an intra-experimental way.

Now it is not denied that there are cases of inference involving perception, one paradigm case of which is the inference to what we do not see, or even cannot see, as when I argue from the cloud-chamber track to a sub-atomic particle with highly unusual properties. However, my only point here is that this is only one way in which the senses can be used, not the model to be applied to every perceptual context. Consequently, the mere fact of error does not prove the presence of inference in every case of perception, and this same conclusion will hold in the case of our perception of the feelings of others as well. For on the model we have been developing here, seeing the feelings of others does not involve any inference to the unseen. Therefore, whatever the problems which plague that kind of inference, they are not the problems which afflict our knowledge of the feelings of others. This does not imply that error is impossible where my judgments of the feelings of others are concerned; the point is simply that the model excludes a certain sort of (particularly annoying) error.

How then is error possible where judgments of the feelings of others are concerned? I think it is possible for anger, depression, etc. to exist simpliciter, insofar as such affective states can be induced by drugs and so on. For the most part, however, there is no such thing as anger simpliciter, but only "anger-at-x-on-account-of-y," which is to say that affective states are intentional states as well. Intentional states of this kind occur, as was pointed out previously, in the context provided for by facts about the immediate situation, the beliefs of the person, knowledge of his character, and so on. Familiarity with the individual and with the context surrounding the affective state often plays a large part in successfully identifying the state that is being presented. Lack of such familiarity may well cause us to hazard a guess about the relevant facts, and thus lead us into mistakes in judgment. There is nothing in such errors, I think, which could not be in principle removed by more acquaintance with the facts of the case.

There are other, more mundane sources of error here as well. Simple lack of attention may lead us to ignore or fail to advert to everything being presented in the intentional field of the body of a person. Sometimes the behavior of others is enigmatic, i.e., we see it as expressive even though we cannot tell what it is that is being expressed. In cases like these, the feelings involved may be so confused or subtle that the person himself may be unable to tell us what they are. Certainly, there are cases in which our ability to perceive the feelings of others is more like that which belongs to the cultured palate of food and wine rather than a simple ability to make coarse distinctions between strong emotions, and it may well be that some individuals lack the required sensitivity, or acquaintance with the subject. There may be others who possess such abilities in superabundance, as the following anecdote from the annals of painting attests:

"Lord Mulgrave, whose name was Phipps, employed Stuart to paint the portrait of his brother, General Phipps, previous to his going abroad. On seeing the picture, which he did not until it was finished, Mulgrave exclaimed, "What is this? This is very strange'." and stood gazing at the portrait. "I have painted your brother as I saw him," said the artist. "I see insanity in that face," was the brother's remark. The General went to India, and the first account his brother had of him was that of suicide from insanity. He went mad and cut his throat. It is thus that the real portrait painter dives into the recesses of his sitters' minds, and displays strengths or weaknesses upon the surface of his canvas. The mechanic makes a map of the man."[10]

"Well," the critic may interject, "perhaps we can build such a case. There is another way in which error may be introduced into our judgments of the feelings of others, i.e., the kind of error which results from our having been deceived by someone as to what their feelings are. This happens not only in those cases in which people deliberately repress their emotions, but also and most typically in those cases in which they pretend or feign feelings that they do not have in order to get us to believe that they really have them. If they are skillful enough at it, they may well convince us that they do have that feeling and thus succeed in deceiving us. If so, then our error looks very much like a faulty inference from bodily movements to an inner state which is not there. For how are we to square the contention that we see anger in the deceiver's face, or in what he does with the admission (which is absolutely required here) that the individual in question is not angry?"

This is a legitimate question, and one to which any account of the kind being offered here needs to make a response. We need to sketch something of the phenomenology of acting, and treat feigning and pretending as instances under this general account. In acting, as it occurs on stage within the context provided for it by the conventions of the theater, there is no problem about saying that the actors express emotions of various kinds, or about the public character of judgments such as "This actor expresses the sadness of Lear."

However, when we use "express" in this way, we clearly use it in a sense different from the one operative in the normal case. We do not mean, I think, to say that the person expressing sadness in such a case is sad, for then he would not be acting.[11] On the other hand, neither do we mean that he literally expresses the sadness of someone else, since no onecould literally express the emotions of another, whether real or fictional, living or dead. Rather, let me suggest that the actor expresses the emotions of the characters he portrays in somewhat the same way that pictures and sculptures do. The sadness of the picture and the sadness of Lear as the actor puts it on display are present there in the intentional field of the picture or the person without being there as the expression of the emotions of that which does the expressing.

So, then, we see anger in the actor's face, but it is not the actor's anger which we see; instead, it is the anger of the character he portrays. Yet what is the nature of this presence? On my view, the actor turns his body into a sort of picture, or a "pictorial space," in which the emotions of others are represented by means of bodily movements. Acting, then, is an affair of representation rather than expression. The bodily movements displaying the intentional state of the character being represented are simulated by the actor in the course of a representational act on stage. So, then, just as a picture can portray the agony of Christ in the crucifixion by representing Christ hanging on the cross, so too can the actor portray the sadness of Lear by representing him on the stage. It is possible for the actor to dissemble in this way because it is possible for us to see persons in other than the normal fashion. We can see the body of the person as a material thing, i.e. look at the person as nothing but a set of bodily movements, observing those movements in such a way as to prescind from their intentional content. The bodily movements that this kind of seeing abstracts from the normal perceptual situation we can learn to copy and, if we are skillful enough, to reproduce at will or on command. This, in fact, is what the actor does in his work.

Now the deceiver is an actor, too. However, he appears only outside of the context of the theater. In the theater, institutional constraints prevent any confusion of expressive with representative acts. As long as we understand the conventions of the theater, we will not be fooled by the actors or make the mistake of the fabled bumpkin who leaped on stage to save the heroine from what he imagined to be real danger at the hand of a villain. Outside of the theater, these institutional constraints are absent and the deceiver can take advantage of the similarity between the relevant expressive and representative acts in order to trick us. He represents himself as a person of a particular sort and uses the intentional field of his body as a sort of picture space in which he simulates the character and emotions of the person whom he would like us to believe that he is. We see anger, sure enough, but no longer protected by institutional constraints, we do not realize that it is anger represented, not angerexpressed, which we see. This is what creates the room for error here.

Again, I have no doubt that the phenomenology of acting which I have been sketching here is far from obviously true, even though I regard it as having some degree ofprima facie plausibility. However, it is once again my intent here, not to argue for the truth of this particular phenomenology, but instead to argue that we can give an account of how people can deceive us with respect to emotions and feelings that does not conflict with the general thesis I have been arguing for, i.e, that some of the affective states of others are directly given in perception. This, I think, is all that is required to turn the objection at hand.

"Very well," the critic may reply, "but doesn't this admission of yours entail that your model is almost devoid of epistemological worth? Presumably, the reason for proposing that we perceive some of the affective states of others is in order, in some way, for our knowledge of other minds. However, it would seem that the difference between real expressions and feigned ones is not one that is given perceptually, but can be specified only by relation to that part of the intentional state that the deceiver simulates, i.e., the subjective feeling itself. You set out to avoid the problems attendant on having to ground our knowledge of other minds on inferences from the seen to the in principle unseen. You may have avoided that problem, but your view must in the end founder on a cognate one, i.e., that of distinguishing cases of deception from genuine cases of expression. Insofar as this can be done only by reference to the presence or absence of the relevant subjective feeling, you have not given us any solution to the problem, only shifted its emphasis. So, when all is said and done, we shall have to conclude that you have made no gain over the traditional inferential account of our knowledge of other minds."

Now I think that there is a good deal of merit to this objection, in that I think that it compels me to admit that the gain over the traditional account is somewhat slight. However, I am not convinced that it represents no gain at all. For, if the account which I have been sketching is correct, at least there is no problem,in principle, about how we can know that other people are angry, happy, depressed and so on, since in some cases it will turn out that we simply see that they are. Of course, people may try to deceive us, but so long as it is at least in principle possible for us to uncover the fraud, this should leave the great majority of our judgments about the intentional states of others unaffected, for in most such cases, as in most cases of mathematical calculation, there is simply no reason to doubt that our judgment is correct.

Yet is it in principle possible to uncover the fraud? The objection seems to suppose that, if the only difference between expressive and representative acts is that in the one case an intentional state is displayed and in the other it is not, we could only have conclusive reason to reject certain expressive acts as spurious if we could somehow introspect the minds of others. Yet this hardly seems to be true. Suppose that we were to witness a man who gave evident signs of grief at his wife's graveside in the bodily attitudes, facial expression and tone of voice he displays during the funeral. Yet suppose we were to discover later that he put poisoned ice cubes in his wife's Scotch-and-Soda, or find him rubbing his hands with glee at the thought of the fortune that will be his due to her insurance policy. Here, I think, we would have perfectly good evidence that the man's grief is mere dissembling, undertaken to prevent (or dispel) the suspicions of others. In a like manner, we will doubt the reality of a particular woman's fear of mice, no matter how convincingly staged, if we later discover that she keeps mice of her own as pets, and gives them free run.

Examples of this kind readily come to mind for the other passions as well, but perhaps it will be more to the point to try to say, in a general way, what such examples should be taken to show. It is, I think, simply this. The affective states put on display in expressive and representative acts are not somehow isolated from the other aspects of the person. On the contrary, the actions, professed beliefs and passions to which human beings are subject create in us expectations for observed behavior. Failure of those expectations may well lead us to re-examine our judgments concerning the character of that person, and there is nothing in principle preventing us from concluding that the emotions or feelings that an individual puts on display are false or insincere, even in the face of the individual's claims to the contrary. For the most part, people cannot deceive us simply by feigning certain emotions, because these emotions where genuine imply certain things for his belief and action. These can be tested for in many cases; consequently, there is nothing requiring that, in order to detect the fraud, we should be able to perform some impossible perceptual act. It does not follow that we will be able in all cases to determine whether or not our suspicions of others are well founded, and in these cases skepticism will, I think, be justified. However, the impossibility here will bede facto, not in principle.

Of course, if my hypothetical opponent is a certain sort of skeptic about other minds, the foregoing will hardly satisfy him. Instead, we shall count upon him to reply as follows. "I can't see how anything you've been saying really addresses the epistemological problem you set out to solve. For on your own principles, it is perfectly possible for the supervenient gestalts you have been mentioning to be present even where the relevant inner states are absent, as in the cases of pictures and actors. So, then, for all you know, the bodies of other persons could be exhibiting all the "inner states" that you claim they do and yet for it still to be the case that they have no minds. For you have not excluded the possibility that they might simply be "living representations" of persons who never have and never will actually exist. Your claim to perceive the "inner lives" of others, then, is as hollow as Descartes' to infer it analogically from his own."

To give an adequate reply to the skeptic here would require that I raise a much more general issue than the one I have been considering here, i.e., the nature of sense-experience as evidence for ordinary existential judgments, a subject that obviously lies beyond the compass of this essay. As such, I shall allow myself only a few comments indicative of my general line of response to thisobjection. In the first place, in no way do I regard the relationof sense-experience to ordinary existential judgment as one of entailment. As such, I neither desire nor need to contest the skeptic's claim that there is an ineliminable logical possibility that I might be wrong in affirming the existence of other minds. However, I do want to claim that our (putative) experience of other minds is such as to provide presumptive evidence for the existence of those minds, and that insofar as that presumption is not overridden by other considerations we can reasonably claim to know that other people have minds.

Making good on this boast is another matter, of course. But if I am right, there is no reason why, at least in principle, the evidence for the existence of other minds could not be of just the same type and just as compelling as the evidence for the existence of tables and chairs. Were this to result, the skeptic about other minds would find himself in the uncomfortable position of either having to abandon his doubts about minds or consistently extending them to cases that seem much less controversial. I would hope that these considerations would be enough to bring most skeptics on this issue back from the brink; with the rest of them I must reserve discussion for another time.


It is time to dismiss my hypothetical opponent, in order that he may be replaced by fleshly counterparts. Before I give them their turn to speak, however, let me recapitulate the argument of this paper. In the first part of the paper, I sketched Aldrich's phenomenology of aesthetic perception, and his application of that phenomenology to the body of the person. The existence of a subjective feeling as a component of affective states was not denied, but the claim that these feelings are in principle unperceivable was denied. It was also denied that the possibility of erroneous judgments about the feelings of others, either in general or in the particular case of deception, shows that these judgments areeo ipso inferred.

In retrospect, what have I proven? I suppose that the answer must be that I have shown palpably little. At most, I have presented some motivation for the view that there is no special epistemological problem about other minds, which is to say that the problem about knowledge of other minds can be reduced to problems about perception generally. It is not supposed that these problems, in turn, are easily solved, only that a shift in emphasis, i.e., from an inferential to a perceptual model, may help to get us clearer about the actual structure of our knowledge of other minds. The paper will have been a success, I think, if it has managed to sketch an alternative to the Cartesian dilemma offering us only a choice between skepticism on the one hand and a difficult to justify inference from my own world of private thoughts to that supposedly hidden away in the body of another person on the other. If so, then perhaps I have gone some way toward justifying the plain man's view that there is something ingenuous about this problem and its related problems in traditional epistemology.