Mental Causation: The Mind-Body Problem

PLATO AND Aristotle were both deeply curious about the mind and its causal relations to the physical world. Neither saw the mental causation problem quite the way we do, since neither conceived of the physical world in quite the way physics does now. But as this chapter will show, their (distinctively different) views lead very naturally to puzzles about how the mind can cause things. Neither view quite makes it past the Scylla of dualism or the Charybdis of monism: dualism takes the soul too far away from physical things to cause them, while monism seems to make all the causation exclusively physical.

Plato and Aristotle both talk confidently about thoughts causing actions. While this idea has been challenged (see below, chapter 4), its roots are deep in our conception of ourselves and how we relate to the world around us.

We are all familiar with battles between reason and desire. Socrates asks whether there are thirsty people who don’t wish to drink (Republic 438c, Plato 1992). Indeed there are. (A sign on a faucet that reads “nonpotable water, do not drink” won’t take away a person’s thirst, but she won’t want to drink there.) Yet there is something paradoxical about this: the word “thirsty” means “wishes to drink.” So we are imagining people who wish to drink and do not wish to drink. How could that be? “It is obvious that the same thing will not be willing to do or undergo opposites in the same part of itself, in relation to the same thing, at the same time. So, if we ever find this happening in the soul, we’ll know that we aren’t dealing with one thing but many” (436b). In other words, since no one thing can both wish to drink and not wish to drink (in the same way at the same time), no one thing can have both of those two characteristics; we thus manage this by being more than one: one part of the soul wishes to drink, and another does not wish to drink.

Plato uses the word “psyche” for this entity that has more than one part. This word is usually translated as “soul.” His word is far more general—covers a far wider range of phenomena—than does our word “soul.” Psyche is that part of us that takes in perceptions from the world around us, desires things, reasons about things, has battles with itself about things. This is essentially what we now call the mind.

Thus Socrates uses a simple observation to discover something surprising about the soul (the mind), that it has parts: “We’ll call the part of the soul with which it calculates the rational part and the part with which it lusts, hungers, thirsts and gets excited by other appetites the irrational appetitive part, companion of certain indulgences and pleasures” (438d). There is one further part of the soul, the “spirited” part. The form of the argument is the same: two people who are exactly similar in both the rational part and the appetitive part may nevertheless do different things. Socrates recalls Leontius, who (a) had an appetite to gaze on corpses but (b) thought that it would be rational not to: “For a time he struggled with himself and covered his face, but, finally, overpowered by the appetite, he pushed his eyes wide open and rushed towards the corpses, saying, ‘Look for yourselves, you evil wretches, take your fill of the beautiful sight!’” (439e–440a). Socrates’ suggestion is that the appetitive part wins, aided by another part of the soul that stirs it on. This Socrates calls the “spirited” part (thymos). Sometimes the spirit acts in defense of reason, sometimes in defense of appetite; hence we should (according to Socrates) infer that it is a part different from either.

The three-part soul enables Plato to explain evil in a flexible and persuasive way. Sometimes people do bad things out of ignorance; but often they know full well what they are doing. This is weakness of will, or akrasia: intentionally doing one thing, even though convinced that one would prefer doing something else. Plato’s explanation is that one acts akratically when one’s action comes from a part of the soul with which one is not fully identified. (I express this in terms somewhat foreign to the way Plato or Aristotle thought about the issue. Aristotle thought that the akratic action must always be the worse action. But if akrasia is simply acting against one’s own best judgment, it could happen that what one ends up doing is, objectively, the better action [Davidson 1970a].)

The terms Socrates uses to describe Leontius—“struggle,” “overpowered,” and even “pushed”—are all causal terms. Psychological struggle, then, is what happens when various parts of the soul interact causally. Equally, when one part of the soul wins out, that part, “victorious,” causes its favored action. Leontius’s spirited part is causally responsible for his eyes widening (“pushed his eyes wide open”) and for his rushing headlong toward the corpses.

If that is what happens in conflicted action, it would seem to be what happens in unconflicted action as well. Even if Leontius had remained serene, averted his gaze, and walked away, his reasoning part would have done that causal work, that is, caused his actions. Hence the perfectly familiar experience of doing (or failing to do) as you choose shows that your actions are always caused by your reasons.

If “like causes like” (Makin 1991), then, since reasons cause physical things to happen, reasons must be physical. Perhaps the most natural way to understand how reasons could be physical is to suppose that they are, like other natural things, subject to growth and decay. Plato’s Phaedo is mainly concerned with the question of the indestructibility of the soul—perhaps a solace for Socrates on the day of his execution—hence with an extended argument that reasons, and the soul, are not like other natural things. The argument is based on considerations about causation, explanation, and the soul.

The conversation begins with the distinction between the mind and the body and the distinction between thinking and sensing. Socrates claims that the soul reasons best when it is free of the senses and the body, and hence the best state for a soul is to be disembodied—in other words, dead. The fundamental distinction between thinking and sensing is invoked to argue that each of our souls must have existed before our bodies came into existence. Consider “sticks or stones or some other things that are equal” (Phaedo 74b, Plato 1977). We know they are roughly the same in size or length. But we don’t ever sense that they are perfectly equal in length, since our senses are too coarse (and the world is too varied) to detect perfect equality. How, then, do we know equality itself, since nothing in our experience is really and truly equal? Socrates’ answer is that the senses are not the only way to know things. The soul can know the Forms: eternal unchanging universals. Our very first acts of sensing anything presuppose knowledge of equality itself, since we can tell that things are unequal even when we are infants. So we must have had knowledge of the Forms from before birth.

Then, since “like is known by like” (De Anima 404b20, Aristotle 1941), the nature of the soul must be more like the nature of the Forms than like things that grow and decay. So, apparently, our souls are eternal and unchanging just like the Forms, and hence there is no reason to fear death.

Socrates’ interlocutors are not fully persuaded. What if, Simmias asks (Phaedo 85e–86d), the soul is just the body and is “scattered and dissipated by the winds” (84b) with the decomposition of the body? Even if it outlasts the body, Cebes asks (87b–88b), might it only outlast it for a time, and hence we still have reason to fear death?

Simmias points out that the arguments in favor of the separability of the soul—that the soul, unlike the body, is invisible, beautiful, partaking of the divine—should work equally well to show that the harmony of a lyre is separable from it. Harmonies, indeed, are special in the natural world in having distinctive mathematical properties, hence they too are more like the Forms than many other things subject to decay and destruction. Yet we know that the harmony is destroyed when the lyre is destroyed. Whatever its Formlike properties may be, lyres and harmonies are ordinary sensible objects. So maybe the soul is like that. Maybe the soul is just another item like the things we see every day, and maybe the characteristic features of the soul come from special ways that ordinary sensible objects are arranged: “We really do suppose the soul to be something of this kind; as the body is stretched and held together by the hot and the cold, the dry and the moist and other such things, and our soul is a mixture and harmony of those things when they are mixed with each other rightly and in due measure” (86b). Simmias’s picture of the relation between the soul and the body is very similar to the way the contemporary materialist sees it, particularly contemporary functionalists (chap. 4, sec. 4.3). The mind or soul is something that has a material nature. And the material nature of a thing—what it is made of and the arrangement or mixture of those elements—determines what it is like and how it behaves.

Socrates responds in three stages. The first stage makes specific objections to the idea that the soul is a harmony. The objections are not very convincing.

Souls themselves can be harmonious or disharmonious. A disharmonious soul would fail to be a harmony and hence would fail to be a soul, if the harmony theory were correct. But this objection overlooks the complexity of harmonies. Play a chord, say a C major chord on a piano or a guitar. Three strings will vibrate harmoniously together. Each string will vibrate at its primary or lowest frequency and also at other frequencies, the overtone series. Depending on what the strings are like, different strings will resonate more strongly at different places in the overtone series. Hence it is possible for three strings to resonate harmoniously together and yet at the same time resonate disharmoniously together.

Socrates, Cebes, and Simmias agree that the soul rules the body and can make it do things that otherwise it would not do. So—Socrates argues—the soul must be something different from the body. Again, this objection overlooks how complex a structured system can be. Steam engines, for instance, have governors: one part of the system controls how other parts are behaving.

Here are two more tempting objections to the harmony theory (not considered by Socrates and his friends, though). Socrates hasn’t told us how a harmony amounts to a soul. Since we don’t have that structural story, we can’t see how a harmony would do it. Sometimes we know enough about some class of things that we are pretty confident we know what kind of explanation would work for it. (If you know roughly how circulation works in humans, you won’t find circulation mysterious in, say, aardvarks.) But sometimes we lack even that much insight into how things work: 150 years ago, it was so mysterious how life works that scientists could seriously maintain that it must be a nonphysical phenomenon. So, lacking even the beginnings of a theory of how a harmony could add up to a soul, one might, not unreasonably, think that a harmony couldn’t possibly add up to being a soul. (Compare the contemporary debate about consciousness between David Chalmers and the Churchlands; see Chalmers 1996; Churchland 1996.)

But we wouldn’t want to rest on this form of argument: just because we don’t know how the mind might be a harmony, we can’t conclude that the mind can’t be a harmony.

Second objection: if human beings were (merely) structured systems of ordinary physical stuff, wouldn’t they all act the same? Wouldn’t we just be a bunch of physical robots?

The answer is surely no. Our physical natures are vastly complex. There is more than enough complexity, and more than enough difference from one physical body to the next, to account for the differences among us.

The second stage (96a–101e) of Socrates’ response is more fundamental, an “investigation into the cause of generation and destruction.” Socrates reports, “When I was a young man I was wonderfully keen on that wisdom which they call natural science,” which investigates the causes for why things come to be and perish. It became clear to him that the natural science way of investigating things doesn’t work for mathematics. We might say that 1 and 1 are somehow parts of 2, but it doesn’t make sense to say that 2 is causally generated out of 1 and 1—“made out of 1 and 1”—or that 1 results from the destruction of 2. Socrates learned from Anaxagoras that there is an alternative to causal explanation: “It is Mind that directs and is the cause of everything … in the way that was best” (97b–c). This seems to be different from the kind of causation involved in coming-to-be and perishing, since Mind directs action in light of the goodness of the end result: in some peculiar way, what lies in the future is the cause of what Mind does in the present.

Anaxagoras himself, by Socrates’ report, didn’t develop this insight and continued to seek causes in the constituents of things. Socrates suggests that Anaxagoras might have given the following explanation for why he, Socrates, was now sitting in the jail cell:

Socrates thinks the explanation in terms of Mind would have been much better; in fact, the causal one is patently absurd: “For by the dog, I think these sinews and bones could long ago have been in Megara or among the Boeotians, taken there by my belief as to the best course, if I had not thought it more right and honorable to endure whatever penalty the city ordered rather than escape and run away” (98e–99a). Socrates diagnoses the error as a failure to distinguish “the real cause from that without which the cause would not be able to act as a cause” (99b).

What exactly Socrates means by “the real cause” he does not say. The third stage of the argument of the Phaedo develops a very different conception of explanation from what we would call causal explanation, different also from what we might have thought Anaxagoras was suggesting.

According to Socrates in this dialogue, the simplest possible answer to the question “why is that?” is the Form of whatever is indicated. I point to my dog and ask, “why is that?” (why is he a dog?), and the answer is, he participates in the Form of Dog. The next, more complex answer to a “why is that?” question supplies necessary connections among things that are guaranteed by the Forms. Dogs are animals, and animals move; hence he moves. Doing proofs in number theory is learning about the structure of the absolutely necessary connections among numbers. Equally, a proper answer to a “why?” question about, say, dogs should display the absolutely necessary connections between being a dog and the subject of the question. That is what Socrates says he came to think we really mean when we say we want to know the cause of things. We want it to become absolutely intelligible, rationally lucid, why things are the way they are. Knowing that is knowing the Forms of the particular things.

Socrates then uses the necessary connections guaranteed by the Form of Soul to argue that the soul is immortal. Necessarily, the soul brings life to whatever has it (105c–d). And life is the opposite of death. So whatever has a soul necessarily excludes death and is hence deathless. Hence, Socrates concludes, when death approaches, only the body dies; the soul lives on.

This is a dazzling argument. It has the same core as the so-called ontological arguments for God’s existence in St. Anselm’s Proslogion and Descartes’s Fifth Meditation, a move from the nature of something to the existence of something. Socrates’ argument seems particularly unconvincing. One could in the same way argue for the immortality of anything that could be said to bring life to a thing. For instance, hearts bring life to things—consider the heart transplant—but hearts are not immortal.

The problem stems, I think, from confusing individuals and Forms. Forms are distinct from individuals, and that distinction needs to be rigidly enforced. Forms delimit how individual things may be, if they exist and share in the Form. It is true that anything that shares the Form of soul is alive, and so it is true that, necessarily, any living thing is not dead. So we could put Plato’s point this way: whatever has a soul is indestructible so long as it is alive. But that is most definitely not the same thing as immortality.

Despite this confusion, Plato is on to an important distinction in kinds of explanation. Take a particular action, such as waving your arm at a particular moment. We can ask, “why is that?” and get an answer in terms of reasons: you saw your friend, and you wish to greet her. We can ask Socrates’ “simple” question, “why is that?” and get his simple answer: it is an action of greeting, because it is an instance of the Form of actions of greeting. We may also ask, what makes that movement of your arm into an action of greeting: “why is that an action of greeting?” Here we could be asking why “that” (the arm moving) merits the description “an action of greeting.” Plato’s “necessary connections” strategy supplies an answer: “it was an action of greeting because it was caused and rationalized by your desire to greet someone.” Here we are exploiting the logically necessary structural connection between actions and reasons. As Plato pointed out, this kind of explanation is not the kind of causation involved in coming-to-be and perishing that apparently interested Anaxagoras, since it describes necessary connections among the natures of reasons, causes, and actions. This is true despite the fact that the explanation refers to the fact that the action was caused by your reason. (More on this theme below, in chapter 7.)

Socrates does not, however, put the point in terms of different kinds of explanation. He talks about what the real cause is. When he points out the absurdity of Anaxagoras’s bones-and-sinews explanation for why he, Socrates, is in jail, he says Anaxagoras fails to distinguish “the real cause from that without which the cause would not be able to act as a cause.” And when Socrates describes his own preferred kind of explanation, the one involving the Forms, once again he seems to think that here he has finally found what we really mean by “cause.”

Socrates does not explain why one kind of cause is the true cause while the other is not a cause. He also does not explain why it is not possible for both to be true causes.

But there really are several kinds of explanation. The auto mechanic’s cause has to do with masses and materials and forces. The biologist’s (functional or teleological) cause has to do with ends and goals and values. And the philosopher’s definitional cause has to do with how concepts are hooked together. Not distinguishing kinds of cause can lead to philosophical confusion: perhaps to Plato’s idea that the soul is immortal.

Whether the arguments succeed or not, Plato is clearly on the side of dualism: the view that the soul/the mind is fundamentally different from the body and indeed fundamentally different from any of the things that come to be and perish. Dualism is the view that the soul, or the mind, is something non-physical, something crucially not part of the world that we see and touch. Plato’s investigation into the Forms provides a model for what the soul might be like (or a “place” for a nonphysical soul): the soul (or the mind) is like the unchanging Forms.

This is a very difficult view to understand. Our minds are not unchanging; on the contrary, they change constantly. I turn my head, and my perception of the things around me changes steadily and completely (even if I fix my gaze on, say, the coffee cup before me, as my head moves, the place it appears in my visual field shifts, however slightly). As I wonder what to do next, memories and feelings and ideas come and go; finally, I form an intention, for instance, to pour a cup of coffee. Hence things in my mind come to be and perish. They do so under the influence of things in the world that come to be and perish. When I make things happen, I bring something to be that did not exist before, something that will itself perish sooner or later, for instance an action of greeting.

Clearly my body is an ordinary thing that is part of the natural world. The natural world, by way of my senses, causally sculpts my conscious experience: I see and hear and smell the things that are actually here before me. My will sculpts the path of my body through the natural world. As I discussed at the beginning of this chapter, Plato thinks of action in causal terms: when someone acts akratically, somehow the wrong part of the soul gains control over the body and makes something happen in the world around us. How could my soul do these things if it is radically different in kind from anything in the natural world?

Socrates holds that the aim of philosophy is the separation of the soul from the distractions of the body. But even when Plato describes what happens to the soul after death (Phaedo 110b–115a and Republic 614b–621d) the description is of a physical world. He remarks (Phaedo 114c): “Those who have purified themselves sufficiently by philosophy live in the future altogether without a body; they make their way to even more beautiful dwelling places which it is hard to describe clearly, nor do we now have the time to do so.” But Socrates needs more than time to “describe clearly” this beautiful place. What would it mean to “dwell” in such a place if one didn’t have a body? How could one open or close the door or sit by the fire? Assuming that the most beautiful places have dogs, how could one scratch one’s dog?

If dualism were true, if the soul/the mind were a different kind of thing from the body, then it may survive the death of the body. It is exceedingly painful, both intellectually and emotionally, to think how the person is no longer here when the body is evidently dead. Yet intellectually it is far more difficult to see how the person could ever do anything—make anything happen—if she isn’t physical at all, if she doesn’t have a body. Physical things have physical causes; how can something that isn’t part of the physical world at all make any difference to it? This problem becomes acute with the rise of modern materialism, as I’ll explore further in chapter 3, below.

In two famously obscure passages (De Anima III.5, 430a10–26, and Metaphysics XII), Aristotle argues that souls (or a certain kind of soul) must be able to exist without matter and must be immortal. But in most of the rest of his writings he argues that the soul is not something distinct from the natural world. The soul is the form of a living body. Hence Aristotle’s view does not have dualism’s difficulty explaining mental causation. But his view does lead quite directly to several distinctive problems about mental causation, problems also experienced by contemporary accounts of the mind-body relation.

Aristotle develops his view of the soul against a metaphysical framework that is much more suitable for giving explanations of natural phenomena than is Plato’s. The two most important elements of the framework are his hylomorphism and his doctrine of the four causes. (Unless otherwise noted, I quote the texts of Aristotle in Aristotle 1996.)

Every man-made thing—every artifact—has three basic features: (a) what it is made of, its material; (b) its form (this might simply be a matter of shape, or it might be a dynamic disposition to do various things in various circumstances); and (c) its function, what it is for. The form and the matter together explain the characteristic behavior of an artifact and hence explain how the artifact performs its function (Physics II.1, 192b8–193b21).

Everything else in the world, according to Aristotle, is also a hylomorphic entity—some hylê, matter, in a certain form, morphê. Cats, for an example of a living thing, (a) have flesh of various kinds (fur, bones, sinews, and so forth), (b) arranged in a shape that is disposed to do various things, which (c) contribute to the function of the cat (primarily, to make more cats and, secondarily, to do the various things—hunting, eating—that enable it to perform that primary function). Nonliving things also have matter and a form. A stone, say, a piece of limestone, has a certain kind of matter arranged in a certain form and (according to Aristotle) has a function.

The form/matter distinction has levels. The parts of an animal are organs. Each organ has a form and a matter. The matter (tissues, cells) in turn has its own form and a matter. And so on all the way down. Aristotle did not endorse the atomic conception of matter (although he knew of the basic idea, from Democritus) but rather thought of all things as composed of the four elements earth, air, fire, and water.

Aristotle and Plato share the idea that a cause is an explanation: something you could give in answer to the question “why is that?” Thinking of objects as hylomorphic entities generates different kinds of answers to “why” questions. Ask “why is that?” pointing to my cat Grushenka, and you get different answers as you focus on different aspects of that one cat: (a) she’s made of cat flesh of various kinds, (b) she has the typical form of a cat, (c) she is for making more cats, and (d) this cat got set into motion by her cat parents.

Thus Aristotle’s four causes (Physics II.3, 194b17–195a27): the material cause is the matter of which a hylomorphic entity is composed, the formal cause is the form, the final cause is the end or goal of the thing, and the efficient cause is what sets a thing in motion.

Aristotle is very clear that all four answers are perfectly legitimate (Physics II.3, 195a5–27). There is no reason to think that one is more nearly the right answer than any of the others. They are different causes, but they are not in competition.

Aristotle does see a kind of competition elsewhere. We can specify a particular kind of cause (one from the four) in more or less detail or precision. “A man, for example, is building because he is a builder, and he is a builder insofar as he has the building craft; his building craft, then, is the prior cause, and the same is true in all cases” (Physics II.3, 195b23). Aristotle says we must always seek the “most precise” cause, the explanation that picks out the cause in the most complete detail. That cause is prior. But he does not say that the prior cause is the only cause, even within a category of causes. That leaves open the possibility that Aristotle could permit more than one cause within a category of causes. He could say that the causes of the man’s building are (a) that he is a builder and (b) that he has the building craft: the first is a less prior, the latter a more prior efficient cause.

Aristotle’s theory of the soul is expressed in the following formula: “the soul is the first actuality of a natural body that is potentially alive” (De Anima 412a27). Consider a ten-day-old fetus. It has the potential to grow eyes. It does not have eyes now, but if it develops normally, it will have eyes. Once it has eyes, much of the time the child will not see (when it closes its eyes or sleeps). When the child actually sees things, the potential to see is fully actualized: that is the second actuality of vision. When our eyes are closed or we are asleep, our eyes only have the first actuality of vision. The fetus has what we might call the “zero-th” actuality of vision, the mere potential to develop eyes.

So the first actuality of a thing is the state in which it has the ability to perform a characteristic activity but isn’t now performing it. The second actuality is the thing actually doing what it is able to do. The second actuality of a living body is the body actually doing the things characteristic of living. The first actuality of a living body is its power to do those things: what it is about the body that permits it to act in the manner characteristic of living things. The soul, according to Aristotle’s account of the tradition (De Anima book I), is what explains the characteristic activities of living things (notably, motion and knowledge). Hence the soul is the first actuality of a living body. Finally, the form (the arrangement of the parts) of the living body is what explains how the body is able to do these things: hence the soul is the form of a living body.

Since the form is simply the arrangement of the parts and the fact that they interact in a certain way, the soul is not a thing in addition to the body. As Aristotle puts it: “Hence we need not ask whether the soul and the body are one, any more than we need to ask this about the wax and the seal or, in general, about the matter and the thing of which it is the matter. For while one and being are spoken of in several ways, the actuality <and what it actualizes> are fully one” (412b6–9). Since the soul is what explains the characteristic activities of any living thing, Aristotle is perfectly comfortable with the idea that dogs and goldfish and geranium plants all have souls. His use of the concept of soul is thus even broader than Plato’s. Since all the activities of what we would call the mind are also characteristic activities of the living human being, Aristotle’s conception of the soul is broad enough to encompass our conception of the mind. Consequently, he needs to explain the mind in hylomorphic terms. I will look at his explanation of sensation and his explanation of belief, desire, intention, and similar states.

When we sense things (when we see, hear, smell, etc.), we get information about those things, information about properties or characteristics of those things. Aristotle understands the properties of things to be the forms of those things.

Aristotle proposes (De Anima II.5, 416b33–418a7) that sensation is the sense organ coming to have the very same form as the perceived object (but of course with its own matter). A “pin art” impression toy similarly takes on the form of an object but in a different matter.

To illustrate: Your eye is able to see white. When you aren’t seeing white—when your eyes are closed or you are looking at something nonwhite—the form of your eye is the first actuality of a body able to see white. The same goes for the white thing: unless it is being seen, it isn’t producing any visual perceptions, so its form is the first actuality of a body able to be seen as white. When your eye turns to the white thing in good light, and you are attending to what it looks at, both first actualities go to their second actuality states: the white thing is being seen, and the eye is seeing. (In fact, Aristotle held [425b26] that the second actuality of the object of perception and the second actuality of the sense are the very same thing.) That is: you see white.

It is not so clear how to extend this idea to belief and desire. Take the belief that the New York Yankees won the World Series in 1999. For someone to have this belief is for the cognizing part of the soul to take on the form of the fact that the Yankees won the World Series in 1999. The matter of the cognition and the matter of the actual fact are different, but they share a form. But what is the form of the fact? and what would it be for a soul to take on that form?

The form of something can be dynamic—that is, involve characteristic activities (in addition to static shape). We can in parallel think of believing—thought taking on the form of a state of affairs without the matter of that state of affairs—in terms of characteristic activities. Having the form of something in one’s soul might be having the ability or disposition to engage in various activities that are related in special ways to that thing. If a person has a belief that somehow involves the number 1999, then that person should be able to count. If a person believes things about the Yankees then she should be able to engage in baseball-related activities: she should be caused to form baseball-related beliefs by baseball-involving events, and her beliefs should be able to motivate baseball-related actions. (For example, if she sees the words “Go Mets!” she forms the belief that a Mets baseball fan wrote those words, and if she’s offered a ticket to a Yankees game, she is grateful.) My suggestion, then, is that taking on the form of a fact, without the matter, is taking on a disposition to behave in a certain way, in particular, to behave in various ways in relation to that fact. (I intend for this suggestion to echo recent work on psychosemantics, or causal accounts of the content of mental representations, as, for instance, in Dretske 1988 or Fodor 1987. These theories begin with causally maintained isomorphisms between the way something in the world behaves and the way something in the organism behaves.)

Aristotle thus believes that the soul, and perceptions, and thoughts are all completely physical. Hence he does not have Plato’s problem of explaining how a nonphysical soul can cause things. Yet Aristotle’s materialism has its own difficulties explaining mental causation. In this section, I will look at three ways Aristotle might respond to Socrates’ puzzle, in the Phaedo, of whether our thoughts or our bones and sinews are the cause of our motions through the world. The first of the three is the core of the solution I will develop in chapter 9 to the contemporary mental causation problem. The second and third are tempting but ultimately unworkable.

Socrates thinks the real cause of his sitting in the cell is his conviction that it is best to remain. The bones and sinews are “that without which the cause would not be able to act as a cause” (Phaedo 99b). Socrates appears to be invoking mind/body dualism at this point. The mind, and the conviction, are things that are distinct from the bones and sinews and could exist without them, but without them would not be able to act as causes.

Aristotle lampoons the idea that the soul is something distinct from the body: “They speak as though it were possible, as in the Pythagorean stories, for just any old soul to be inserted into any old body, whereas in fact each body seems to have its own distinctive form and shape. What they say is like saying that carpentry gets inserted into flutes; in fact a craft must use suitable instruments, and equally the soul must use a suitable body” (De Anima 407b22–26, Aristotle 1941). The precise character of your soul and the precise character of your body are two sides of the same coin.

But the question remains, what causes the body—a moving arrangement of parts—to sit in the cell? One explanation is an efficient-cause explanation: the way the body was moving earlier is what causes the body to move now. According to the hylomorphic account of particular states of mind offered above, Socrates’ conviction that it is best to remain in the cell is part of the structure (the form) of his body, since it is a part of his soul, and his soul is the form of his living body. So we can think of his conviction as some part of his body (nowadays we think it is part of his brain) that has the right kind of form to cause the behaviors characteristic of having such a conviction. Hence when he acts on this conviction, this part of his body—a certain portion of matter, with a certain structure—is the efficient cause of what the body now does.

Suppose Aristotle had argued that the matter of the conviction—its parts—is the efficient cause of the present motion of the body, rather than the conviction itself. Then there is a short argument that the only efficient causes are the four elements. The matter of large things like people is typically more hylomorphic things. Bones are a certain structure of the matter of bones. So the argument that undercuts the claim that the conviction has to be the cause of the action would also undercut the bones as a cause and in fact any hylomorphic entity. So the only causes are the things that are not further hylomorphic entities: the four elements.

But he need not get on to this slippery slope. He can instead simply say that both the conviction and the parts of which it is made (in their structured arrangement) are efficient causes of the body sitting in the cell. In chapter 9, I will show how to flesh out this idea.

Aristotle pays careful attention to the differences among the fundamental categories of things and, in this case, the differences among the four kinds of causes. Since they are so different, and so fundamental, it’s clear that getting the category of a thing wrong leads to real philosophical confusion. Teleological explanation only looks problematic, for instance, if one thinks that the final cause is an efficient cause (since, typically, efficient causes occur before and final causes occur after things done for an end).

So Aristotle could, apparently, say that there is a category confusion (Ryle 1949) involved in asking whether both the bones and sinews and Socrates’ conviction cause him to remain in the cell. He could say: The bones and sinews are efficient causes of the movement of the body. The fact that Socrates has this conviction is the fact that these elements (this matter) are arranged in a certain way. But—he can say—it is not true that the conviction is another efficient cause of motion. Rather, the conviction is a form, not matter. Just as the soul is not another thing of the same kind as the matter that is arranged in the form of the soul, the conviction is not another thing of the same kind as the matter that composes it.

The problem with this line of thinking about the causes of action is that Aristotle does think that Socrates’ conviction is an efficient cause of action. He has to, since (as Plato showed) our everyday experience of making things happen is experience of our reasons as efficient causes of action. The bones and sinews are material causes of the conviction as well as efficient causes of his action. And the conviction too is an efficient cause of the action. Hence the category confusion move won’t help Aristotle stay off the slippery slope of the last section: the temptation to say that the bones and sinews are the real efficient causes or, rather, that only the four elements are the ultimate real causes.

Aristotle points out that there are different “sciences” of aspects of the soul: “The student of nature and the dialectician would give different definitions of each of these affections—of anger, for instance. The dialectician would define it as a desire to inflict pain in return for pain, or something of that sort, whereas the student of nature would define it as a boiling of the blood and of the hot <element> around the heart. The student of nature describes the matter, whereas the dialectician describes the form and the account: for desire, for instance, is the form of the thing, but its existence requires this sort of matter” (403a29–403b5). The student of nature studies things like the boiling of blood and the hot regardless of their connection with psychology. The psychologist is the scientist who studies the “account,” or definition, of psychological things. The Aristotelian psychologist’s account of reasons might look like this: A particular person’s belief that the Yankees won the World Series in 1999 is the matter of that person taking on a certain dynamic form, namely, the disposition to behave in the familiar ways that people do when they have that belief. This behavior involves a causal pattern: the belief has certain typical causes (perception, testimony) and certain typical effects (speech behavior, action). This account of belief defines belief in terms of its causal connections to other things (just as the dialectician defines anger as a desire to inflict pain in return for pain). Hence to have a belief one must have a state that stands in the causal relations called for by the definition. In other words, beliefs are second-order states: the state of being in some (first-order) state or other that is capable of standing in certain causal relations.

Now, if being a belief is by definition being a state that is caused in a certain way and typically causes certain things, then, it would seem, Aristotle has a straightforward solution to the problem. The conviction, just like the belief, must by definition be a cause of actions based on that conviction. So it has to be that the conviction causes Socrates to remain seated in the cell. (And, presumably, so do the bones and sinews.)

The trouble is that when we define belief (or conviction) in this way, belief turns out to be something that has certain causes and effects. But it’s the “something” here that is doing the causing, not the thing that (by definition) is the pattern of causing and being caused. In other words, all the causing is done by the first-order states (in this case, the bones and sinews, that is, the body), not by the second-order states. The second-order states constitute a sort of office, a job description; the first-order states are the holders of the office, the things that actually do the job.

This problem for Aristotle—below, in chapter 7, I will call it “the new logical connections argument”—is serious, and it infects contemporary philosophy of mind, particularly the behaviorism of Ryle and contemporary functionalist accounts of the mind. The problem is bad enough to justify rejecting certain versions of functionalism altogether (chapter 7, sec. 7.6).

Plato thinks of the kind of causation involved in coming-to-be and perishing as a species of the kind of explanation we seek in mathematics and in the analysis of meanings and concepts. He uses that model of explanation to argue that the soul is indestructible. This account of the soul has two serious difficulties from the perspective of mental causation. First, souls (minds) change constantly as they perceive and think different things, but if they are more like Forms than like vegetables (changeable natural things), they cannot change. Second, perception and action demand ordinary causation between things in the natural world and thoughts. But if souls are fundamentally different in kind from natural things, then they can’t have causal relations with natural things—hence no mental causation.

Aristotle keeps as much as he can in the natural world. The soul is the form of a living body, hence it is an aspect of a thoroughly natural object. There are various kinds of causal relations that natural objects enter into, but they are all relations in the natural world. Aristotle thus appears better situated to explain how mental causation is possible than is Plato. Still, there are difficulties. The parts of a structured object, working together, cause the same things as the whole structured object, hence there is a reason to think that only the unstructured parts of hylomorphic entities ever cause anything. And there is a danger in defining the soul, and defining particular mental states, as a dynamic form, that is, as standing in causal relationships with other things. The danger is that the dynamic form picks out a pattern of things interacting causally. Then the causing is done by the elements of the pattern rather than the whole pattern.

Neither Plato nor Aristotle considered the precise version of the mental causation problem I described in chapter 1, section 1.2: that for anything you or I do (anything we make happen), there is an explanation in terms of the physical properties of our bodies and the physical laws. Neither Aristotle nor Plato (especially Plato) thought that physical matter explains very much. Both knew of the atomist theories of the world of Democritus and Leucippus. Both thought that the order and purposefulness of the world could not possibly arise out of the unguided interactions of unthinking matter (see particularly Aristotle, Physics 2.8).

Since the seventeenth century and the rise of modern physics beginning with Galileo and Newton, it has become ever more plausible that the order and purposefulness of the world could be produced simply by the interactions of blind physical matter (blind to the order and purposefulness that emerges). It wasn’t unreasonable for Plato and Aristotle to doubt this, given that the atomism of their contemporaries had very little systematic empirical support. (Equally, it wasn’t unreasonable to think that matter-in-motion could produce all the phenomena there are: Lucretius, three hundred years later, a Democritean materialist who had no more empirical support than did Democritus, proposed exactly that; see Lucretius 1969.) Only with the depth and power of contemporary physical theory are we compelled into the mental causation problem in its present form: since physics, in principle at least, explains everything, there is no room for the mind to make anything happen.

Christopher Bobonich’s “Akrasia and Agency in Plato’s Laws and Republic” (1994) describes Plato’s evolving account of weakness of will. Some useful articles on the Phaedo and causes and the soul include Gregory Vlastos’s “Reasons and Causes in the Phaedo” (1969), Christopher Byrne’s “Forms and Causes in Plato’s Phaedo” (1989), Gareth B. Matthews and Thomas A. Blackson’s “Causes in the Phaedo” (1989), and Sean Kelsey’s “Causation in the Phaedo” (2004).

A good source of articles on Aristotle’s De Anima is Essays on Aristotle’s De Anima (Nussbaum and Rorty 1992). In that collection, Burnyeat’s “Is an Aristotelian Philosophy of Mind Still Credible?” and the article by Martha Nussbaum and Hilary Putnam, “Changing Aristotle’s Mind,” deal with the question of how close Aristotle’s view is to contemporary functionalism.