a) The Distinction between Mental and Physical
Phenomena

Descartes makes a sharp distinction between bodies and minds: Bodies are extended, minds think. To think, Descartes tells us, is to doubt, to understand, to conceive, to affirm, to deny, to will, etc. In short, to think is to have mental acts. Brentano tries to improve on Descartes’s distinction. If what distinguishes the mind from the body are mental acts, then we must ask next how mental acts are different from bodies. Mental acts and only mental acts, according to Brentano’s famous answer, are intentional. Mental acts and only mental acts are directed toward objects. Ultimately, what distinguishes a mind from everything else is its direction toward objects, its intentionality. This insight, resurrected if not discovered by Brentano, has been the foundation of most recent philosophies of mind. It is a cornerstone of Meinong’s theory of objects as well as Husserl’s phenomenology. It has echoes in Heidegger’s philosophy, and it appears in Sartre. Moore uses it to refute idealism, and Russell bases his distinction between knowledge by acquaintance and knowledge by description on it.

Let me interrupt for a moment and clarify a distinction which I have presupposed. Descartes speaks of material substances; Brentano, of physical phenomena. I have talked about perceptual objects as well as of physical objects, and I have distinguished between the two. Chairs, mountains, clouds, apples, people, etc., are examples of perceptual objects; molecules, atoms, electrons, and quarks are examples of physical objects. Perceptual objects consist of physical objects. An apple, for example, consists of atoms and, ultimately, of the most elementary particles the physicist has discovered. With this distinction in mind, the realismidealism issue concerns the existence of perceptual objects: Are apples mental or not? It is clear, firstly, that if there are no perceptual objects, then there are no physical objects either; for the former are supposed to consist of the latter. It is also clear, secondly, that if perceptual objects are mental, then so are the physical objects of which they are supposed to consist. Finally, and thirdly, granted that perceptual objects are nonmental, it is an open empirical question of what sorts of ultimate physical object they consist.

Brentano, I said, speaks of physical phenomena, and he asks how they differ from mental phenomena. He presupposes that this distinction is exhaustive: some thing is either a mental phenomenon or else it is a physical phenomenon; there is no third possibility. This assumption, as we shall see in a moment, leads Brentano into trouble. Brentano starts out by giving examples of mental phenomena:

Every idea or presentation which we acquire either through sense perception or imagination is an example of a mental phenomenon. By presentation I do not mean that which is presented, but rather the act of presentation, thus hearing a sound, seeing a colored object, feeling warmth or cold, as well as similar states of imagination are examples of what I mean by this term. I also mean by it the thinking of a general concept, provided such a thing actually does occur. Furthermore, every judgment, every recollection, every expectation, every inference, every conviction, or opinion, every doubt is a mental phenomenon. Also to be included under this term is every emotion: joy, sorrow, fear, hope, courage, despair, anger, love, hate, desire, act of will, intention, astonishment, admiration, contempt, etc.

(F. Brentano, Psychology from an Empirical Standpoint, pp. 78-79)

What a difference there is between this list and Descartes’s! Descartes mentions such “intellectual” mental acts as doubting, affirming, understanding, and conceiving. There is no mention of emotions. Brentano, on the other hand, lists joy and sorrow, despair and anger, love and hate. Or compare Brentano’s inventory of the furniture of the mind with Hume’s assertion that “All the perceptions of the human mind resolve themselves into two distinct kinds, which I shall call impressions and ideas” (D. Hume, A Treatise of Human Nature, Book I, Part I, Section I).

But there is also something strange about Brentano’s list. He mentions hearing a sound, but also seeing a colored object. Hearing and seeing are mental acts of perception. What we perceive are perceptual objects, their properties and relations. We see tomatoes and their colors, we taste steaks and their flavors, we hear bells and their tones. At this point, Brentano does say that seeing a colored object is an example of a mental phenomenon, but it is clear from other places that this is a slip of the pen. According to Brentano, we cannot really perceive perceptual objects, but only their properties. For example, we cannot really perceive a tomato, but only its color. But this is not quite correct either. According to his view, perception somehow reduces to acts of sensing, the having of sensations. What we “perceive,” consequently, are not even properties of perceptual objects, but are properties of sensations. What we perceive, to be most faithful to Brentano’s view is something red, for instance, where the something is not a perceptual object but a sensation.

When we turn to Brentano’s examples of physical phenomena, we find another slip of the pen. He mentions “a color, a figure, a landscape which I see, a chord which I hear, warmth, cold, odor which I sense; as well as similar images which appear in the imagination” (Brentano, pp. 79-80). The example that does not fit is the landscape. A landscape is not a property of a sensation like color, shape, warmth, cold, etc. are. Of course, from my point of view, only the landscape belongs to the list. Only the landscape is a perceptual object, properly to be contrasted with mental phenomena. But Brentano sees things quite differently from us. A color, warmth, an odor, these are to his mind examples of physical phenomena. Even if we substitute the term ‘nonmental’ for Brentano’s ‘physical phenomenon’, his list sounds weird. Obvious examples of nonmental things are, rather, tomatoes, the Empire State Building, and a hair on Napoleon’s head.

But we are even more surprised when we read that images are examples of physical phenomena. What could be more obvious than that images are mental things? Would there be images, if there were no minds? Surely, something is fundamentally wrong with Brentano’s division. And it is not difficult to discover what this something is. Brentano has divided the world up, not into mental and nonmental things, but into mental acts and the rest of things. Brentano identifies mental phenomena with mental acts. As a result, whatever is not a mental act turns out to be a physical phenomenon. Since an image, as distinguished from an act of imagination, is not a mental act, it must be a physical thing. Since a sensation, something red, as distinguished from an act of sensing, is not a mental act, it must be a physical phenomenon. Even a pain, as distinguished from the act of feeling a pain, must be classified as a physical phenomenon. Brentano’s identification of the mental with mental acts leads to philosophical catastrophe.

Another thought may also have played a role. Perhaps Brentano was convinced that nothing spatial could possibly be mental. Since sensations can have spatial properties and stand in spatial relations to each other, as we explained in the last section, he may have concluded that they cannot be mental. Since they cannot be mental, they must be physical. And a similar argument may have convinced him that images are physical things.

We can easily separate Brentano’s important insight from his catastrophic mistake. Intentionality, the direction toward an object, we must realize, is not the essential characteristic of the mental, but of mental acts. Mental acts and only mental acts have objects. But there are also mental things which are not intentional and, hence, which are not mental acts. An image, for example, is obviously a mental thing; yet it is not intentional in the way in which an act of imagination is. A pain, as distinguished from the act of feeling it, has no object. Nor do sensations have objects. A mind consists of many things. Some of these are mental acts, others are not. And only the former are intentional.

Sensations, according to Brentano’s division, are physical phenomena. This is the bizarre corollary of his assertion that all mental phenomena are intentional. But something even stranger is yet to come. Brentano also subscribes, like almost everyone else, to the view that a color, warmth, cold, etc. are not to be found in the world of the physicist. He accepts, in other words, some version of the distinction between primary and secondary qualities. According to this distinction, color, warmth, cold, etc. are said to be mental things. Thus we have now the following situation: According to Brentano’s definition of the mental, these things (properties) cannot be mental but must be physical. According to his belief in the primary-secondary quality distinction, however, they cannot be physical. Thus they can be neither mental nor physical. They belong neither in the mind, nor in the nonmental world. They are in limbo. They are nowhere. This means that there really are no such things! The dialectic of the situation is obvious. Color, for example, is first shunted into the mind because it is not a primary quality of physical objects. When it is so relocated, it is called a sensation. But Brentano also thinks that mental things must be intentional, and color is obviously not an intentional thing; it is obviously not a mental act. Therefore, it cannot be mental either; it cannot really be a sensation in the mind. In desperation, Brentano concludes that there is no such thing at all as color.

Brentano is a Berkeleyan without ideas (sensations)! His philosophy transcends idealism. In his ontology, there are mental acts and nothing but mental acts. There are no colors, no smells, no pains, no images, etc. Beyond mental acts, there may or may not exist a world of physical objects, but our belief in such a world is mere conjecture. Is it not surprising that the student of Aristotle and opponent of Kant should end up with an idealism more extreme than Berkeley’s? Two important lessons can be learned from Brentano’s plight. Firstly, we must not accept Brentano’s characterization of the mental: There are mental things which are not mental acts. Secondly, once again we are made aware of the pervasiveness of the primary secondary quality distinction. Brentano may have been an Aristotelian at heart, but his mind belonged to the “empiricistic” tradition of modern science. According to this tradition, color does not exist in the “external” world.

But color is not in limbo. It belongs to the world. However, it is neither exclusively a part of the external world, nor is it exclusively mental. As a property of sensations, it may be said to be “in the mind.” As a property of perceptual objects, it may be said to belong to the “external” world.

b) The Thesis of Intentionality

Every mental act (though not every mental thing) is directed toward an object. I shall call this “the thesis of intentionality.” Once pointed out, the thesis is a truism. There is no idea which is not an idea of something; there is no desire which is not a desire for something; there is no belief which is not a belief in something; and so on. No philosophical argument could persuade us otherwise. But the consequences of this truism have great philosophical significance.

Mental acts have objects. But in what, precisely, does this “having” of objects consist? There is an obvious and to my mind correct answer: Every mental act is related to an object. There is a unique relationship—let us call it “the intentional nexus”—which holds between every mental act and something else, the so-called object. Without the intentional nexus, there would be no world for a mind. But this obvious answer runs immediately into a most intractable difficulty. Sometimes, what we see does not exist, what we desire does not come to pass, what we believe is not the case, and so on. In short, sometimes the objects of our acts do not exist, are not facts, are not the case. Some acts, as I shall put it, have nonexistent objects. But how can a relation, the intentional nexus, possibly hold between a mental act and something that does not exist? Must not a relation always establish a connection between existents? How can there exist a connection to something that is not there? Is not a relation with a nonexisting term, as some philosophers have claimed, an absurdity? On the other hand, if the intentionality of mental acts does not consist of a relationship between the act and its object, what does it consist of? How else can we possibly understand the “directedness toward an object”? What else can we make of the fact that every mental act “has” an object? I really see no alternative to the relational interpretation of the thesis of intentionality. But irrespective of whether I am right or wrong, the dialectic of the situation is clear: While a relational account of intentionality must face the problem of nonexistent objects, a nonrelational account must explain the directedness of mental acts.

But it is not only the problem of nonexistent objects that has moved philosophers to throw themselves upon the second horn of the dilemma. A relational account presupposes the ontological recognition of relations. One who rejects the existence of relations, for whatever metaphysical reason, cannot explicate intentionality in relational terms. Brentano, I think, could never shake off an Aristotelian prejudice against relations. His students Meinong and Husserl freed the theory of intentionality from the burden of this prejudice.

Recall the Cartesian mystery of how representation is possible without identity or similarity. The intentional nexus solves this mystery. An idea is neither identical with nor similar to its object. This is the Cartesian insight. Yet, it represents its object in that it intends it. This is the insight of Brentano and his students. For an idea to represent an object is to have that object; and to have an object, is to be related to it by means of the intentional nexus. You may think that this is no solution at all, but merely an invitation to accept the Cartesian mystery. And you are right, up to a point. We have here a situation quite common in philosophy. We have the question: How is such and such possible? And the answer seems to be nothing more enlightening than: Well, it just is! How is it possible for the body to interact with the mind? Well, it just does! We must take a closer look at the question in order to understand the true nature of the situation. The question really is: How is it possible for the body to interact causally with the mind, if such interaction must be thought of as taking place in space, and minds are not spatial? The answer is, not trivially that such interaction takes place as a matter of fact, but rather that it can take place because causal interaction must not be thought of as taking place necessarily in space. The mystery arises because of an assumption, and the answer claims that this assumption is false. Causal interaction must be understood not as spatial interaction, but as lawfulness. This is the enduring lesson of Hume’s analysis.

In our case of intentionality, the mystery arises because it is similarly assumed that there are no other relations than similarity and (its limit) identity. The Cartesians were faced with a mystery because they could not conceive of any other relations between ideas and their objects than those of similarity and identity. What we have, therefore, is not just a stubborn insistence that ideas do represent their objects. There is philosophical progress. A fundamental philosophical prejudice has been destroyed. In the case of causality, what has been abandoned is the “push-pull model” of causality. In our present case, what has been removed is a pervasive prejudice against relations other than those of similarity and identity. It is not by chance that Brentano’s students arrived simultaneously at a better understanding of the nature of intentionality and of the importance of relations. In order to accept a relation as peculiar as the intentional nexus, one must first acknowledge, cheerfully and without reservation, the existence of the category of relation.

It was another student of Brentano’s, K. Twardowski, who introduced the by now familiar distinction between a mental act, its content, and its object (K. Twardowski, On the Content and Object of Presentations). An act of presentation, for example, is a mental individual thing. It is not spatial, but it is temporal. This act has two important properties which I shall call “kind” and “content.” It has, firstly, the property of being an act of representation; this is its kind. This property distinguishes it from acts of judgment, for example. An act of judgment would have the property of being a judgment. It would be of a different kind. It has, secondly, a content, that is, a property which determines its object. For example, the presentation of a certain tree is different from the presentation of a certain elephant. And this difference, according to the distinction under discussion, is not just an “external” difference between the objects, but is an “internal” difference between the two acts. The acts themselves are qualitatively different. Since it is possible for different kinds of acts to have the same object, there are all kinds of possible combinations. Someone may believe while someone else doubts that Mt. Everest is the highest mountain on earth. In this case, we have acts of different kinds, but with the same object and, hence, with the same content. Or it may be the case that someone believes that Mt. Everest is the highest mountain on earth, while someone else believes that Salzburg is the birthplace of Mozart. In this case, the two acts are of the same kind, but the contents differ and hence, the objects. And so on.

As one can see from these last examples, the term ‘object’ has to be taken in a wide sense. The object of a belief is not an individual thing, but a circumstance, a state of affairs. The kind of mental act which intends such states of affairs I shall call “propositional.” One of the most important theses of this investigation is that perceptual acts are propositional. Acts of seeing, for example, intend states of affairs. The importance of this thesis will become clear when we discuss how we are acquainted with abstract entities, that is, with things which are neither spatial nor temporal.

A second point to note is this. The content of an act varies with its object, and conversely. Content and object determine each other. It is therefore the content, not the act as an individual thing, which stands in the intentional nexus to the object. The intentional nexus thus holds between a property of the act and the act’s object. In the case of propositional mental acts, the intentional nexus obtains between a property and a state of affairs.

The so-called content of an act of presentation is, I believe, what some philosophers have called an “idea,” a “concept,” or a “notion.” The idea of the moon, for example, as it is supposed to exist in the mind, is a certain property of all of those acts which are acts of presentation of the moon. Thus my idea of the moon is identically the same as yours, even though my act of presentation differs numerically from yours. It is clear that ideas must be distinguished from sensations. An idea is nothing “sensible.” It does not have a color, a shape, or a size. The idea of an equilateral triangle, for example, is neither equilateral nor is it triangular. It has no shape at all. But the sensation which is under normal circumstances caused by the perception of an equilateral triangle has a shape. Our sensations may or may not resemble the things which cause them, our ideas never do. This, of course, is the heart of Descartes’s contention that ideas represent without similarity. It is also the view which Berkeley attacks in the Introduction to the Treatise.

Both ideas and sensations must be further distinguished from images. Images, like sensations, can resemble what they are images of. They, like sensations, are individual things rather than properties. The mental image of an equilateral triangle, for example, has a shape. Ideas represent their objects because they intend them, even though they do not resemble them. Sensations do not represent anything, but they may resemble their causes. Images are like sensations in that they do not intend anything but resemble the things of which they are images. However, what makes an image of the Eiffel Tower into an image of the Eiffel Tower is not solely its resemblance to the Eiffel Tower, but a mental act of imagining or of remembering. When you imagine a mermaid, there occurs a certain kind of mental act as well as an image. It is this mental act of imagining which “connects” the image with some object; in this example, with a mermaid. In imagination, we imagine something by means of something else, the image. Imagining is a more complicated kind of mental act than, for example, believing or desiring. It resembles the two fundamental kinds of mental act that make language possible, namely, meaning something by something and understanding something by something. For example, we mean a certain shape by the word ‘square’ and a certain shade of color by the expression “olive green.” In these two cases, the words (certain shapes or noises) correspond to the image, what is meant to what is imagined. In memory, too, the memory image gets its direction toward an object through a mental act, namely, an act of remembering.

So much for the important distinction between concepts (ideas), sensations, and images; a distinction that was often hidden under the blanket term “idea.”

c) In the Footsteps of Brentano: Moore’s
Refutation of Idealism

According to Twardowski, the content of an act is part of the act, but its object is not. According to our analysis, the content is a property of the act, its object is not. While the object of a given mental act may or may not exist, the content always exists. If you conceive of Mt. Everest, then the object as well as the content of your conception exist. But if you conceive of the golden mountain, then only the content of your conception exists. Philosophers have said such things—and Brentano was one of them—as that Mt. Everest exists, not only in your mind, but also in reality, while the golden mountain only exists in your mind. In terms of the distinction between content and object, we can make sense of this kind of talk. There are really not two kinds of existence, existence in reality and existence in the mind. There is only one kind. The golden mountain does not exist, period. Only its idea exists.

Brentano speaks of the “immanent objectivity” of mental acts and of the “mental inexistence of objects” (Brentano, p. 88). When you conceive of the golden mountain, the golden mountain is the immanent object of your conception; it has mental inexistence. As we use these terms, an immanent object is a contradiction in terms. The object of an act is always extraneous to the act. What is immanent is the content.

The content of a mental act could not exist without the act, while its object does not depend on the act for its existence. Mt. Everest, for example, does not depend for its existence on your conceiving it. It would exist even if there were no mental acts at all; even if nobody had ever seen it or thought of it. But the idea of Mt. Everest would not exist if there had never been a mental act of which it is the content. This simple truth forms the basis of Moore’s famous attempt to refute idealism (G. E. Moore, “The Refutation of Idealism,” pp. 1-30). Moore distinguishes between the sensation blue (notice, not blue sensation!) and the awareness of this sensation. He argues that since the sensation blue is an object of the act of awareness and not a part of it, the sensation could exist without the awareness. In our words, since the sensation blue is the object of an awareness, not one of its properties, the existence of the sensation does not necessarily depend on the existence of the act. Berkeley’s esse est percipi is therefore not necessarily true, and idealism has been refuted.

The first thing to notice is Moore’s insistence that the idealist must treat the esse est percipi principle as a necessary truth. What kind of necessity is here involved? It is clear from Moore’s argument that it must be ontological necessity. We may agree that it is a law of ontology that a property of mental acts can only exist if mental acts exist. It follows then that the sensation blue, if it is a property of mental acts, cannot exist unless the acts exist. It is then necessarily true that a sensation cannot exist unless it is experienced, that is, unless it is the property of an experience. By pointing out that the sensation blue is not a property of acts of experiencing, not a “content” of mental acts, but the object of such acts, Moore shows that the ontological law does not apply to sensations. He thus shows that one argument for the necessity of Berkeley’s principle, as applied to sensations, is not sound.

But he does not show that this principle is false. And, indeed, I think that the principle is true for sensations, not as a matter of some kind of necessity, but simply as a matter of fact. Sensations occur only “in minds”, they do not float around all by themselves. But to occur “in a mind” is the same as to be experienced. This, as we shall see, follows from our conception of what a mind is. Therefore, there are no sensations which are not experienced. And if Berkeley were correct in claiming that perceptual objects are nothing but bundles of sensations, then the idealist’s conclusion that such objects cannot exist unperceived would indeed follow. This shows, I think, that Moore’s main argument hardly touches on the issues. The important question is not whether Berkeley’s principle is necessarily true for sensations, but whether it is true. Moore merely shows, if he shows that much, that unsensed sensations are ontologically possible, not that there are any.

But it should also be obvious that the truly important question is not even the question of whether or not there are unsensed sensations, but whether or not there are perceptual objects. The question is not whether or not a blue sensation can exist when it is not experienced, but whether or not there are apples. Only if we identify apples with bundles of sensations, as Berkeley does, do the two separate issues get conjoined. And it is precisely this identification which the realist must reject. For the realist, sensations are mental and do not exist unsensed, while perceptual objects are not mental and can exist unperceived. When Berkeley identifies perceptual objects with bundles of sensations, he invites the obvious objection that this identification must be mistaken because an apple can exist unperceived, while a bundle of sensations cannot exist unsensed. I think that this objection is the most decisive refutation of Berkeley’s idealism. I also think that Berkeley was quite aware of the force of this reply, but could not think of a way to avoid it. Be that as it may, it is somewhat ironic that Moore’s refutation of idealism threatens to disarm our most telling objection to Berkeley. If Moore holds that sensations actually exist unsensed, then we can no longer argue that they are in this respect quite different from perceptual objects and hence cannot be identified with them.

At the end of his paper, Moore claims to have shown that there is no question of how we are to get outside the circle of our own ideas and sensations. “Merely to have a sensation,” he says, “is already to be outside that circle. It is to know something which is as truly and really not a part of my experience, as anything I can ever know” (Moore, 1970a, p. 29). His claim is of course only true if he means by “ideas and sensations,” not Berkeley’s ideas, but mental acts of having ideas and sensing sensations. To sense a sensation, we can agree, is to “transcend” the sensing toward the sensation. But with this meaning, a Berkeleyan can be perfectly comfortable. He claims that we do not get outside the circle of Berkeleyan ideas, that is, of sensations; and Moore has done nothing to refute this claim. The important question is: How do we get outside the circle of mental things to the perceptual objects? Moore turns to this question in the very last paragraph of his paper. There he maintains that we make contact with perceptual objects by means of the same peculiar relation of awareness through which we are acquainted with sensations. Berkeley is mistaken when he claims that we can be directly aware only of our own ideas. And then Moore states: “The question requiring to be asked about material things is thus not: What reason have we for supposing that anything exists corresponding to our sensations? but: What reason have we for supposing that material things do not exist, since their existence has precisely the same evidence as that of our sensations?” (Moore, 1970a, p. 30).

This may not be a refutation of idealism, but it is a step toward realism. Our acquaintance with perceptual objects, Moore tells us, is as direct as our acquaintance with sensations. It is not mediated by sensations. This takes care of the skeptical devil. Furthermore, material things in space are sharply distinguished from sensations; the former do not consist of the latter. This takes care of the idealistic deep blue sea. After such a splendid beginning, we may wonder, why does Moore so desperately try in later papers to stick sensations to the surfaces of perceptual objects? Was it because he could not see his way clear to defend this sort of realism? Was it because he never really believed in it in the first place? We only know that in the end he frankly admitted that although he believed in realism, he did not know how to defend it. My essay is an attempt to succeed where Moore and many others have failed. I follow in Moore’s footsteps, trying to show that they lead eventually to a solid realism that can be defended against all skeptical and idealistic objections.