Naturalism, Free Choices, and Conscious Experiences (2007)
As I understand them, Charles Taliaferro and Stewart Goetz (hereafter, TG) are best viewed as making two main objections to a certain kind of naturalism. In Section 1, I clarify what this kind of naturalism claims. In Section 2, I defend it against their first objection, and in Section 3 I defend it against their second objection.
1. The Kind of Naturalism in Question
The word “naturalism” has been used by philosophers, scientists, and others to denote many different views. But in this reply I’ll use it to denote just one view, the kind of naturalism that TG reject and I’ll defend. What does this view claim? TG write: “In brief, theism … gives a fundamental place to purposeful explanation, whereas naturalism gives the most central role of explanation to nonpurposeful and (typically) causal explanation.” I think I agree. Naturalism claims that nothing has a fundamental purposeful explanation. Something (e.g., some action) has a purposeful explanation when it occurs for some purpose. Naturalism can agree that some things have purposeful explanations. For example, naturalism can agree that I visit the gym for the sake of my health, i.e., for the purpose of maintaining my health. What naturalism denies is that anything has a purposeful explanation that’s fundamental. Naturalism says that whenever an occurrence has a purposeful explanation, it has that explanation in virtue of certain nonpurposeful (e.g., merely causal) facts. So, to continue my example, my visiting the gym really does have a purposeful explanation—I really do visit the gym for the sake of my health; but my visiting the gym for the sake of my health is simply the fact that my going to the gym is caused by my mental state of wanting to be healthy and my mental state of believing that visiting the gym promotes my health. A purposeful explanation isn’t some kind of noncausal explanation; it’s just a special kind of causal explanation—one that cites the wants (or purposes) and beliefs of an agent. (For the classic defense of this claim, see Davidson 1963.)
So naturalism implies that everything that has a purposeful explanation has it in virtue of nonpurposeful facts—presumably, facts about the interactions of fundamental physical particles in accordance with impersonal physical laws. But TG object that some things that have purposeful explanations don’t have them in virtue of nonpurposeful facts: the purposeful explanations they have are fundamental. Specifically, TG claim that free choices have fundamental purposeful explanations, and the fact that the universe contains conscious mental states such as pains has a fundamental purposeful explanation.
2. Free Choices
Naturalists can agree that free choices have purposeful explanations, since they can understand purposeful explanations simply as those causal explanations that cite the actor’s beliefs and desires (or purposes). But TG claim that free choices have fundamental purposeful explanations. If a free choice has a fundamental purposeful explanation, then the choice has no cause, but still has a purposeful explanation; and since the choice has no causal explanation for this purposeful explanation to be reduced to, the purposeful explanation must be fundamental. For example, if my free choice to write this paper has a fundamental purposeful explanation, then, although my choice to write this paper has no cause, it still has a purposeful explanation (perhaps I chose to write this paper in order to keep a promise) that must therefore be fundamental.
Accordingly, to persuade us that free choices have fundamental purposeful explanations, TG must first persuade us that free choices have no causes. Their argument is apparently that (1) we experience our free choices as uncaused, that (2) this experience is some evidence that our free choices really are uncaused, and that (3) there is no evidence that our free choices aren’t uncaused. Let me explain why I disagree with premises (1) and (3).
Premise (1) is far from obviously true. It’s certainly true that we don’t experience our free choices as caused, but that’s not the same thing as experiencing our free choices as uncaused. The distinction may sound pedantic, but it isn’t. Suppose that I’m standing in front of a London bus with a blindfold on. Then, because I can hear and smell the bus but not see it, I don’t experience the bus as red; but it doesn’t follow, and it’s not true, that I experience the bus as not red. Moreover, my not experiencing the bus as red is no evidence at all that the bus is not red; the reason why I don’t experience the bus as red is that my senses of hearing and of smell just don’t inform me about the colors of things, and neither do my other nonvisual senses. Likewise, if I introspect while making a free choice, I don’t experience the choice as caused, but it doesn’t follow that I experience it as uncaused. Moreover, for all we know, the reason why I don’t experience the free choice as uncaused is that introspection just doesn’t inform me about the causes of my choices; and if that’s so, then my not experiencing a free choice as caused is no evidence at all that the choice is uncaused.
Premise (3) strikes me as definitely false; I say there is evidence that our free choices are caused. If TG are correct, and our free choices are uncaused, then those of our bodily movements that constitute actions must result from chains of neural events that begin in the brain, the first events in such chains being caused by choices that are themselves quite uncaused. But there’s ample evidence that this isn’t so, i.e., that those of our bodily movements that constitute actions result from chains of neural events that don’t begin in the brain. Part of this evidence is negative, i.e., the failure of neuroscientists over the past century to discover any chains of neural events that begin in the brain, lacking neural causes. Surely there would’ve been a Nobel Prize for such a discovery! But the most important part of the evidence is positive, i.e., positive evidence that all neural events are realized by (and hence are nothing over and above) physical events, and that every physical event has a physical cause. I’ve sketched such evidence, providing references to full treatments, in section four of my opening essay, “A Case for Physicalism about the Human Mind.” However, TG don’t consider any evidence of this sort, apparently because they assume that their opponent would have to be advancing a merely methodological claim.
Let me very briefly review this positive evidence. The many explanatory successes of biochemistry provide evidence that we can account for all the distinctive properties and behaviors of neural events on the assumption that they are realized by certain biochemical events. Similarly, the many explanatory successes of physical chemistry provide evidence that we can account for all the distinctive properties and behaviors of biochemical events on the assumption that they are realized in turn by certain physical events. Thus, biochemistry and physical chemistry strongly suggest that all neural events are realized by (and hence are nothing over and above) physical events. The evidence for thinking that every physical event has a physical cause (to the extent that it is caused at all) is simply the remarkably successful track record of physics in identifying physical causes of a wide variety of physical events. Of course, physicists still expect to have to change their theories to accommodate anomalous physical phenomena; but they expect to find such phenomena in, say, the large-scale structure of matter or high-energy particle colliders, not in human brains.
A further ground for thinking that free choices are caused is the peculiar consequences of supposing that they’re not caused. I’ll mention two such consequences. First, consider the fact that people very often freely choose to act in character; that is, they very often freely choose to act in ways that make sense given the beliefs, preferences, and moral principles that they profess at other times. For example, it’s a rarity if an accountant who’s hitherto been diffident and frowned upon ostentation chooses one morning to wear to the office a hot pink denim suit; it’s a rarity if a notoriously loud-mouthed sexist who thinks he’s automatically more intelligent than all women puts in for a transfer to a team with a female boss. Now, if free choices are uncaused by anything, and hence uncaused by the chooser’s beliefs, preferences, and moral principles, then there’s no obvious way to explain why, time and again, the chooser’s choices make sense in light of his or her beliefs, preferences, and moral principles; the supposition that free choices are entirely uncaused makes the fact that people generally choose to act in character an extraordinary coincidence. It might be replied that a person’s beliefs, preferences, and moral principles are defined by reference to their actual choices, and that’s why their choices mesh so smoothly with their beliefs, preferences, and moral principles. But this reply doesn’t explain the notable fact that over time a person by and large makes the same kind of free choices, i.e., those that mesh with a fairly stable set of his or her beliefs, preferences, and moral principles. If free choices were entirely uncaused, then you’d expect them not to conform to any pattern at all. For example, I freely choose about the same time every afternoon to drink a decaf mocha. Why so if each free choice is causally unconnected with my beliefs and preferences and indeed uncaused by anything?
Second, consider the evidence on the basis of which we’re prepared to say that someone else freely chose to act in a certain way. For example, nearly all viewers of the 2006 soccer world cup final between France and Italy will have concluded, very reasonably, that French star Zinedine Zidane quite freely chose to head-butt an Italian player in the chest. But what was their conclusion based on that made it reasonable? Well, the viewers saw that Zidane was subject to no force majeure, that he did not trip or fall, and that his head-butt was not a tic, or a spasm, or some kind of compulsive behavior. They also had no reason to suspect insanity or sudden brain damage, especially because both before and after the head-butt Zidane provided abundant evidence that his abilities to deliberate, to engage in practical reasoning, and to control his limbs were intact. However, if free choices were uncaused, then these bases for reasonably concluding that Zidane freely chose to head-butt an Italian player would be inadequate, since none of them makes it at all reasonable to think that some neural event in Zidane’s brain had no neural cause. The actual evidence we rely on in judging that someone else has chosen freely seems to fit poorly with the view that free choices are uncaused; it fits much better with the view that free choices are those (caused) choices that result when one’s mechanisms of deliberation, practical reasoning, and bodily control are functioning properly, and one’s beliefs and preferences get translated into choices in a biologically normal way.
I very much doubt, then, that the choices we classify in everyday life as free are counterexamples to the naturalist claim that nothing has a fundamental purposeful explanation.
3. The Fact That the Universe Contains Conscious Mental States
But what about the fact that the universe contains conscious mental states? TG can be interpreted—fairly, I hope—as arguing thus:
(1) Naturalism can’t plausibly explain how it came about that the universe contains physically irreducible conscious occurrences such as experiences of pain and pleasure.
(2) Theism can plausibly explain how it came about that the universe contains physically irreducible conscious occurrences such as experiences of pain and pleasure.
Therefore, physically irreducible conscious occurrences such as experiences of pain and pleasure are evidence that the fact that the universe contains conscious mental states has a fundamental purposeful explanation.
Theism’s explanation of the universe is, of course, a purposeful explanation (i.e., God created the universe for a purpose) and presumably a fundamental one.
Both premises of this argument are questionable. Let me discuss (1) first, temporarily going along with the argument’s assumption that conscious occurrences really are irreducibly mental events, and that physicalism about the mind is therefore false. Even so, (1) seems to be false. Naturalism can easily explain how the universe came to contain physically irreducible conscious occurrences. It can do so by supposing that, among the fundamental laws governing the universe, there are some according to which, whenever such-and-such complex nonconscious occurrences occur, so-and-so conscious occurrences occur; perhaps such a law says that, whenever a human brain attains a certain kind and degree of complexity, a pain is experienced. Given such laws, the capacity for consciousness that some creatures enjoy, like the capacity for breathing, can be explained as having arisen through natural selection. Through mutation, some creature was born with a brain of the requisite kind and degree of complexity to generate conscious experiences; and then, because these experiences increased the creature’s fitness, such creatures were selected for. I see nothing implausible about such an explanation, so I think (1) is false.
Why, then, do TG think that (1) is true? The only argument I can find occurs in this passage:
If we restrict ourselves (as naturalists like Armstrong and Dennett want) to the explanatory framework of an ideal physics with mass and energy, it is hard to see how any configuration of the physical world can constitute let alone explain the emergence of consciousness.
Now I quite agree that, if the only fundamental laws are physical laws (i.e., laws linking physical events of one kind with physical events of another kind), then it’s impossible to explain the emergence of physically irreducible conscious events. But naturalists needn’t hold that the only fundamental laws are physical laws; they can also believe in fundamental laws connecting physical events with irreducibly conscious events. Of course, naturalists who believe in such laws are thereby rejecting physicalism about conscious events. But that’s no problem, since naturalism doesn’t logically require physicalism about conscious events: a universe in which nothing has a fundamental purposeful explanation might still be one in which consciousness is physically irreducible. It’s true that nearly all naturalists in fact endorse physicalism about conscious events as well; but that doesn’t show that naturalism logically requires physicalism about conscious events.
So far, I have been going along with the assumption of premise (1)—and indeed of premise (2) also—that conscious occurrences really are irreducibly mental events, and hence that physicalism about the mind is false. In fact, of course, I reject this assumption, for the reasons given in my opening essay, “A Case for Physicalism about the Human Mind.” But since TG argue for the assumption, I must explain why I find their argument (generally known as the Knowledge Argument) unpersuasive. According to this argument, we can coherently imagine a scientist of the future, Mary, who has never actually experienced pain herself, but who has come to know everything that a completed multidisciplinary science of pain (including psychology, neuroscience, and biochemistry) has to say about pain. But still she doesn’t know everything about pain. For we can imagine that, once her studies are over, Mary stubs her toe and thus experiences pain for the first time in her life. And, in doing so, she learns something: she comes to know that having a pain in the toe is like that. But since she already knew everything that a complete multidisciplinary science of pain has to say about pain, but didn’t know that having a pain in the toe is like that, the property of being like that must be a property that a complete multidisciplinary science of pain fails to mention. Such a property couldn’t therefore be a physical property, and so physicalism about the mind can’t be true.
I think the Knowledge Argument fails, because it relies on a general principle that’s false. According to the Knowledge Argument, since Mary knows all the science of pain, but then learns something new (i.e., that pain is like that), it must be that the subject matter of the new knowledge (i.e., pain’s being like that) is a feature of reality that the science of pain didn’t refer to or talk about at all. The general principle underlying this reasoning seems to be this: if someone knows a certain body of knowledge, but then learns something new, it must be that the subject matter of the new knowledge is a feature of reality that the old body of knowledge didn’t refer to or talk about at all. However, this general principle seems false; it holds true of very many but not all situations. Consider the following story. Because of a blow to the head I suffer terrible amnesia and forget who I am. But I read in the newspaper that tomorrow, for reasons that don’t matter, one Andrew Melnyk will be publicly flogged. “Bad news for this Melnyk fellow,” I think to myself, but I soon return to my quest to find out who I am. Later, however, I discover that I am Andrew Melnyk, and then, of course, I realize, to my horror, that tomorrow I will be publicly flogged! The point of the story is that, when I learn that tomorrow I will be publicly flogged, I thereby gain vitally important new knowledge. According to the general principle on which the Knowledge Argument relies, the subject matter of my new knowledge must be a feature of reality that my old body of knowledge didn’t refer to or talk about at all. But the subject matter of my new knowledge isn’t some feature of reality that my old body of knowledge didn’t refer to or talk about at all. On the contrary, the subject matter of my new knowledge (that tomorrow I will be publicly flogged) is exactly the same feature of reality as the subject matter of my old knowledge (that tomorrow Andrew Melnyk will be publicly flogged)—simply because my being publicly flogged tomorrow just is Andrew Melnyk’s being publicly flogged tomorrow. So the general principle on which the Knowledge Argument relies is false, and the Knowledge Argument fails to prove its intended conclusion.
Of course, you might well ask what is the difference between my new knowledge (that tomorrow I will be publicly flogged) and my old knowledge (that tomorrow Andrew Melnyk will be publicly flogged) if the subject matter in each case is the same. In broad strokes, the answer is that two items of knowledge can represent the very same feature of reality, but still differ from one another in representational format (think, for example, of how a photo of someone differs from a verbal description of the same person) and in how they are processed (for example, “Mary loves John” is easier to process than “John is loved by Mary”) and perhaps in other ways too. So, for all that the Knowledge Argument shows, pain’s being like that might be a straightforward physical property of the brain that, before she stubs her toe, Mary can represent using concepts drawn from the science of pain, but that afterwards she can represent by using special representational formats that can only be used by someone who has actually experienced pain.
Let me now turn to premise (2). How do TG suppose that theism can plausibly explain how it came about that the universe contains physically irreducible conscious occurrences? They write this:
The existence of minds like ours is not a mystery in a theistic universe, however, because theism takes the most fundamental fact of reality to be a conscious and purposive mind. In other words, the conscious, purposive reality, God, did not emerge from anywhere but was there all along as a necessary existent, and God created nonphysical minds such as us for the purpose of experiencing complete or perfect happiness through, at least in part, the adoration of God.
Premise (2) presumably implies that such an explanation of why the universe contains physically irreducible conscious occurrences is plausible. But is it?
Theists can always explain why so-and-so exists by saying that God decided to create so-and-so, and that whatever God decides to create comes into existence. But if they leave the explanation at that, there’s no explanatory gain. For we can always ask why God decided to create so-and-so, and explaining why God decided to create so-and-so (rather than such-and-such) looks no easier than explaining the phenomenon we started with—why so-and-so exists. Another way to see the same point is to notice that, if this is the theist’s game, then naturalists can play it too. Just as theists can always explain why so-and-so exists by saying that God decided to create so-and-so, naturalists can always explain why so-and-so exists by saying that something quite mindless, a new kind of particle perhaps, once existed whose nature it is, in accordance with some hitherto unknown fundamental law, to cause so-and-so. Of course, this naturalistic explanation is worthless; but if so, then so is its theistic analog, from which it differs in no relevant respect. In principle, however, theists can do better—by exhibiting God’s decision to create so-and-so as one manifestation of some standing desire or characteristic that has, or can have, other manifestations too. And presumably this standing characteristic would be God’s moral perfection or, perhaps, his perfect love.
Bearing in mind these points, let’s evaluate TG’s specific suggestion that “God created nonphysical minds such as us for the purpose of experiencing complete or perfect happiness through, at least in part, the adoration of God.” Clearly TG aren’t just saying that nonphysical minds exist because God created them; commendably, they’re also trying to explain why he created them. Their explanation assumes that God desires to create other minds that experience complete or perfect happiness and that we are such minds. However, this assumption seems inconsistent with the fact that many, perhaps even all, human minds are less than completely or perfectly happy. If God desires to create other minds that experience complete or perfect happiness, and we are such minds, and if God is omnipotent and hence able to fulfill any of his desires, then why aren’t we all a lot happier than we are? Evidently TG have some explaining to do. Now I’m not saying they couldn’t do this explaining; several possible explanations are imaginable, though they might only lead to further difficulties. My point is merely that, until TG actually do this explaining, and successfully, their hypothesis concerning why the universe contains nonphysical minds can’t be called plausible—contrary to their premise (2).
TG argue that the fact that the universe contains conscious mental states has a fundamental purposeful explanation. However, if what I’ve been arguing is right, the fundamental purposeful explanation that they offer isn’t—yet—a plausible one; and there’s a rival naturalistic explanation that at least isn’t implausible.
Davidson, D. 1963. “Actions, Reasons, and Causes.” Journal of Philosophy 60: 685-700.
Hobart, R.E. 1934. “Free Will as Involving Determination and Inconceivable Without It.” Mind 43: 1-27.
Jackson, F. 1982. “Epiphenomenal Qualia.” Philosophical Quarterly 32: 127-136.
Jackson, F. 1986. “What Mary Didn’t Know.” Journal of Philosophy 83: 291-295.
Ludlow, P., Y. Nagasawa, and D. Stoljar. 2004. There’s Something about Mary: Essays on Phenomenal Consciousness and Jackson’s Knowledge Argument. Cambridge, MA: The MIT Press.
Papineau, D. 2002. Thinking about Consciousness. New York, NY: Oxford University Press.
Stampe, D. and M. Gibson. 1992. “Of One’s Own Free Will.” Philosophy and Phenomenological Research 52: 529-556.
 My arguments in this paragraph and its successor are distinct from but inspired by those in Hobart 1934.
 Furthermore, Zidane’s remarks to the press several days later made it clear that he did not regret his action, and that it was a deliberate response to a grave insult; it had not resulted from an insuperable surge of anger.
 One version of this sort of understanding of free choice can be found in Stampe and Gibson 1992.
 The argument was first advanced by Frank Jackson (see his 1982 and 1986), though he has now abandoned it. It was and continues to be the subject of a large philosophical literature; see, most recently, Ludlow, Nagasawa, and Stoljar 2004, where Jackson’s recantations can also be found.
 For an account of what these special representational formats might be, see, for example, Papineau 2002, ch. 2.
 That God desires to create other minds that experience complete or perfect happiness looks at first sight like a good candidate to follow from God’s perfect love or moral perfection. On closer examination, however, I think the matter is not so clear. Perfect love requires benevolent treatment of minds that already exist, but doesn’t seem to require bringing new minds into existence. Must a morally perfect being desire to create other minds that experience complete or perfect happiness? Perhaps; but surely the question raises difficult questions in moral philosophy.
 Thanks to Paul Draper for helpful comments on an earlier draft.
Copyright ©2007 Andrew Melnyk. The electronic version is copyright ©2007 by Internet Infidels, Inc. with the written permission of Andrew Melnyk. All rights reserved.