The Causal Closure Argument

The Causal Closure Argument

Print Friendly, PDF & Email


I begin by stating that I do not defend any form of materialism in this paper. Rather, I defend a commonsensical form of soul-body dualism in which souls make undetermined choices for purposes (reasons). I defend this commonsensical view of the world against an argument that is frequently used to undermine its truth. This is the argument from causal closure. Before setting forth and examining this argument, however, it behooves us to have a reasonably clear and concise commonsense sketch of how souls are causally related to their physical bodies on occasions when human beings make what I will assume are essentially undetermined choices (from here on, I will simply assume that choices are essentially undetermined). This picture is as follows: on certain occasions, we have reasons for performing incompatible actions. Because we cannot perform both actions, we must make a choice to do one or the other (or neither), and whichever choice we make, we make that choice for a reason or purpose, where that reason provides an ultimate and irreducible teleological explanation of that choice. The making of a choice is a mental event that occurs in a soul and either it, or some other mental event associated with it (e.g., an intention to act) directly causally produces an effect event in that soul’s physical body. In other words, there is mental-to-physical causation and its occurrence is ultimately and irreducibly explained teleologically by the reason that explains the making of the choice.

To put some flesh on the proverbial bones, consider the movements of my fingers right now on the keys of my keyboard as I work on this essay. If these movements occur because of a choice of mine to type, then these physical movements are ultimately and irreducibly explained teleologically in terms of the purpose for making my choice to write this essay, which, we can suppose, is that I make clear that there are no good scientific objections to the view that human beings are soul-body compounds and that those souls have free will (make choices for reasons). Hence, if the movements of my fingers are ultimately occurring because I made a choice to write this essay for a purpose, then a mental event involving me (a soul) must be causing those movements to occur as I write this essay for the purpose that I make clear that there are no good objections to the view that human beings have souls that make choices. In other words, if our commonsense view of a human being is correct, I, as a soul, cause events to occur in the physical world by making a choice to write this essay for a purpose.

From the example of my typing, it should be clear that the claim that there is causal interaction between a soul and its physical body is not a ‘God-of-the-gaps’ type of argument. In discussions about God’s existence, critics often argue that theists postulate God’s existence in light of an inability of science to provide a complete explanation for a physical datum (or data). This lack of a complete explanation is a gap in the scientific story. By analogy, a critic might argue that I am postulating my soul’s existence in light of an inability of science to provide a complete explanation for the movements of my fingers when I type this essay. But this argument would be mistaken. My claim is not that there are certain physical events (the movements of my fingers) for which a failure to find a complete physical causal story warrants appeal to the causal activity of a soul as their ultimate explanation. Rather, my claim is that our commonsense understanding of our purposeful activity entails that some physical events must occur whose ultimate causal explanation is not other physical events but non-physical mental events whose occurrences are explained teleologically by purposes.


What is wrong with this commonsense understanding of a human being? According to many philosophers, a serious problem for the view that souls make choices that causally produce events in physical bodies arises out of the practice of science.1 Richard Taylor puts forth a lengthy argument, the gist of which is as follows:

Consider some clear and simple case of what would . . . constitute the action of the mind upon the body. Suppose, for example, that I am dwelling in my thought upon high and precarious places, all the while knowing that I am really safely ensconced in my armchair. I imagine, perhaps, that I am picking my way along a precipice and visualize the destruction that awaits me far below in case I make the smallest slip. Soon, simply as the result of these thoughts and images, . . . perspiration appears on the palms of my hands. Now here is surely a case, if there is any, of something purely mental . . . and outside the realm of physical nature bringing about observable physical changes. . . . Here, . . . one wants to say, the mind acts upon the body, producing perspiration.

But what actually happens, alas, is not nearly so simple as this. To say that thoughts in the mind produce sweat on the hands is to simplify the situation so grossly as hardly to approximate any truth at all of what actually happens. . . . The perspiration . . . is secreted by tiny, complex glands in the skin. They are caused to secrete this substance, not by any mind acting on them, but by the contraction of little unstriated muscles. These tiny muscles are composed of numerous minute cells, wherein occur chemical reactions of the most baffling complexity. . . . These . . . connect eventually, and in the most dreadfully complicated way, with the hypothalamus, a delicate part of the brain that is centrally involved in the emotional reactions of the organism . . . . [B]ut it is not seriously considered by those who do know something about it that mental events must be included in the description of its operations. The hypothalamus, in turn, is closely connected with the cortex and subcortical areas of the brain, so that physical and chemical changes within these areas produce corresponding physical effects within the hypothalamus, which in turn, by a series of physical processes whose complexity has only barely been suggested, produces such remote effects as the secretion of perspiration on the surface of the hands.

Such, in the barest outline, is something of the chemistry and physics of emotional perspiration. . . . The important point, however, is that in describing it as best we can, there is no need, at any stage, to introduce mental or nonphysical substances or reactions.2

According to Taylor, while we are inclined to believe that certain physical events in our bodies are ultimately explained by mental events of non-physical substances, as a matter of fact there is no need at any point to step outside of the physical causal story to explain the occurrences of those physical events. Jaegwon Kim uses an example of a neuroscientist to make the same point:

You want [or choose] to raise your arm, and your arm goes up. Presumably, nerve impulses reaching appropriate muscles in your arm made those muscles contract, and that’s how the arm went up. And these nerve signals presumably originated in the activation of certain neurons in your brain. What caused those neurons to fire? We now have a quite detailed understanding of the process that leads to the firing of a neuron, in terms of complex electrochemical processes involving ions in the fluid inside and outside a neuron, differences in voltage across cell membranes, and so forth. All in all we seem to have a pretty good picture of the processes at this microlevel on the basis of the known laws of physics, chemistry, and biology. If the immaterial mind is going to cause a neuron to emit a signal (or prevent it from doing so), it must somehow intervene in these electrochemical processes. But how could that happen? At the very interface between the mental and the physical where direct and unmediated mind-body interaction takes place, the nonphysical mind must somehow influence the state of some molecules, perhaps by electrically charging them or nudging them this way or that way. Is this really conceivable? Surely the working neuroscientist does not believe that to have a complete understanding of these complex processes she needs to include in her account the workings of immaterial souls and how they influence the molecular processes involved. . . . Even if the idea of a soul’s influencing the motion of a molecule . . . were coherent, the postulation of such a causal agent would seem neither necessary nor helpful in understanding why and how our limbs move. . . . Most physicalists . . . accept the causal closure of the physical not only as a fundamental metaphysical doctrine but as an indispensable methodological presupposition of the physical sciences. . . . If the causal closure of the physical domain is to be respected, it seems prima facie that mental causation must be ruled out . . . .3

While Kim agrees with Taylor about the lack of a need on the part of a scientist to go outside the physical explanatory story, he introduces the stronger idea that to be successful the physical sciences need to make the methodological assumption of the causal closure of the physical world. Is he right about this? To insure clarity about what is at issue, consider one more example of movements of my body that according to common sense could only be adequately explained by mental causation of a soul whose choice is teleologically explained by a purpose or reason. Right now, I am tired and feel tight in my back after typing for several minutes, so I raise my arms in order to relax. Reference to my mental activity and my purposes for acting seems not only helpful but also necessary to explain both the movements of my fingers on the typewriter while I am typing and the subsequent motions of my arms when I relax. If we assume for the sake of discussion that I, as a soul, cause my fingers and arms to move by directly causing some neural events in the motor section of my brain, then when I move my fingers and raise my arms for purposes, I must directly cause initial neural events in my brain that ultimately lead to the movements of those extremities. In other words, in order to explain adequately (teleologically) the movements of my limbs, there must be causal openness or a causal gap in my brain. While Kim believes the commonsense view implies this causal openness, he also believes that it is because the commonsense view implies the existence of this causal gap that it must be mistaken. Because the neuroscientist methodologically assumes causal closure of the physical world, what she discovers as the explanation for what occurs in my brain and limbs when I type and relax must not and need not include reference to the mental causal activity of my soul and the ultimate and irreducible explanatory purpose for its choice to act. Given that the principle of causal closure entails the exclusion of a soul’s mental causation of a physical event and the ultimate and irreducible teleological explanation of that mental event and its effects by a purpose, it is imperative that we examine the argument from causal closure to see if it provides a good reason to believe that the movements of my fingers and arms when I am typing and stretching must be completely explicable in terms of neuroscience (or any other physical science), with the result that no reference to the causal activity of my soul and its purposes for typing and raising my arms is required.

width=390Contrary to what Kim maintains, there is good reason to think that the argument from causal closure is unsound.4 To understand where it goes wrong, let us distinguish between a neuroscientist as an ordinary human being and a neuroscientist as a physical scientist. Surely a neuroscientist as an ordinary human being who is trying to understand how and why my fingers move and arms go up while I am typing must and would refer to me and my reasons (purposes) for acting in a complete account of why my limbs move.5 Must she, however, as a physical scientist, avoid making such a reference? Kim claims that she must avoid such a reference because as a physical scientist she must make a methodological assumption about the causal closure of the physical world. Is Kim right about this and, if he is, is such a commitment compatible with a commitment on the part of a physical scientist as an ordinary human being to causal openness? Or must a neuroscientist, who as a physical scientist assumes causal closure, also assume, if he is consistent, that as an ordinary human being his mention of choices and their teleological explanations is no more than an explanatory heuristic device that is necessary because of an epistemic gap in his knowledge concerning the physical causes of human behavior?

In order to answer these questions, it is necessary to consider what it is about physical entities that a physical scientist such as a neuroscientist is often trying to discover in his experimental work. What is the purpose of a neuroscientist’s inquiry? In the case of Kim’s neuroscientist, what she is trying to discover as a physical scientist are the capacities of particles or micro-physical entities such as neurons to be causally affected by exercised causal powers of other physical entities, including other neurons. For example, in his pioneering work on the brain Wilder Penfield produced movements in the limbs of patients by stimulating their cortical motor areas with an electrode.6 As Penfield observed the neural impulses that resulted from stimulation by the electrode, he had to assume during his experiments that the areas of the brains of his patients on whom he was doing his scientific work were causally closed to other causal influences. Without this methodological assumption, he could not conclude both that it was the electrode (as opposed, say, to something ‘behind the scene’ such as an empirically undetectable human soul, either that of the patient or someone else, or God) that causally affected the capacities of the neurons to conduct electrical impulses, and that it was the causal impulses of those neurons that causally affected the same capacities of other neurons further down the causal chains to produce the movements of the limbs. There is no reason, however, to think that because Penfield’s investigation of the brain required the methodological assumption of causal closure of the areas of the brains he was studying during his experiments that he also had to be committed as a physical scientist to the assumption that the physical world is universally (in every context) causally closed, where universal causal closure entails that the relevant brain (neural) events can only be causally produced by events of other physical entities and not instead by mental events of immaterial souls alone when they indeterministically choose and intend (plan) to act for purposes. That is, there is no reason to think that because a neuroscientist like Penfield must assume causal closure of a delimited area of the brain in the context of his experimental work in order to discover how physical entities causally interact with each other that he must also be committed as a scientist to the universal explanatory exclusion of mental events of souls that on certain occasions cause the occurrence of events in the physical world. All that the neuroscientist as a physical scientist must assume is that during his experiments souls (either the patients themselves or others) are not causally producing the relevant events in the micro-physical entities in the areas of the brain that he is studying. If the neuroscientist makes the universal assumption that in any context events in micro-physical entities can only have other physical events as causes and can never be causally explained by mental events of souls and their purposes, then he does so not as a scientist but as a naturalist, where a naturalist is a person who believes that the occurrence of physical events can only be explained in terms of the occurrence of other physical events and without any reference to ultimate and irreducible purposes of souls.7

It is relevant to note in this context that Penfield himself was not a naturalist. Rather, he was a soul-body dualist.8 One can surmise, then, that were Penfield to have been presented with the argument from causal closure, he would have found it wanting. And for good reason. In seeking to understand how events of different physical entities affect the capacities of micro-entities such as neurons, a neuroscientist such as Penfield is seeking to learn about properties of physical entities that are essentially conditional or iffy in nature. A property that is conditional in nature is a property that is specified in terms such as ‘If such-and-such is done to object O (e.g., a cause C is exerted on O), then so-and-so will occur to O (e.g., O will move at rate R). As the Nobel physicist Richard Feynman says, scientific questions are “questions that you can put this way: ‘if I do this, what will happen?’ . . . And so the question ‘If I do it what will happen?’ is a typically scientific question.”9 The following description by David Chalmers of basic particles that are studied by physicists nicely captures their iffy nature:

Basic particles . . . are largely characterized in terms of their propensity to interact with other particles. Their mass and charge is specified, to be sure, but all that a specification of mass ultimately comes to is a propensity to be accelerated in certain ways [moved at certain rates] by forces, and so on. . . . Reference to the proton is fixed as the thing that causes interactions of a certain kind that combines in certain ways with other entities, and so on . . . .10

What Chalmers describes as a ‘propensity’ of a particle to be accelerated is a capacity of it to be moved which is such that if it is actualized (triggered) by an exercised causal power of another entity (whether physical or non-physical in nature), the particle will be necessitated to behave in a certain way. There is nothing, however, in the nature of the propensity or capacity of that particle that entails that it can only be actualized by the exercised power of a physical entity. That is, there is nothing in the nature of that propensity or capacity that entails that it cannot be actualized by persons making undetermined choices for reasons. Hence, the actualization of a micro-particle’s capacity to behave in a certain way by a person on an occasion when the latter makes a choice for a reason is not excluded by anything that is discovered in a scientific study of that capacity. And it is precisely on occasions like those noted by Kim, when finger and arm movements occur seemingly for purposes, that a neuroscientist will reasonably believe that the originative micro-physical movements are traceable to the causal activity of a soul that is choosing to act for a purpose. If a neuroscientist makes the presupposition that micro-physical entities can have their capacities actualized only by other physical entities and never by choices made by souls for purposes, then he does so as a naturalist and not as a scientist.

My response to the causal closure argument assumes Feynman’s and Chalmers’ iffy picture of micro-entities that, in addition to being iffy, is also deterministic in the sense that no effect will occur in any micro-entity unless some causal event determines or necessitates that effect to take place. Might there not, however, be random (non-deterministic) changes in the system of micro-entities as well as the deterministic ones? In other words, while sometimes a neuron fires because it gets deterministic causal input from the neurons with which it is connected, at other times it fires at random (without any deterministic cause), perhaps as a result of random quantum fluctuations in a chaotic system that are magnified at the neuronal level.

If we assume for the sake of discussion that neurons do sometimes fire randomly, is it possible to distinguish sharply between those firings that occur randomly and those that occur as the result of being causally determined by a mental event of a soul? After all, the two kinds of firings are alike to the extent that neither has a physically deterministic cause. I believe that it is possible to make this sharp distinction between the two kinds of firings. The way to make the distinction is in terms of contexts that are known, in the case of ourselves, through first-person experience and, in the case of others, through third-person observation. All one need do is ask how plausible it is to maintain that every time a person purposefully chooses to do something such as move his fingers to type, an initial neuron just happens to fire at random (as a result of quantum fluctuations, etc.) with the result that finger movements occur that perfectly mesh with or map onto those that are intended by that person. Because such repeated coincidences would literally be, dare I say, miraculous, the only plausible view is that the neuron must not be firing randomly but because of the causal input from a soul choosing to act for a purpose.


The argument from causal closure is a methodological argument about what scientists must supposedly assume in order to practice science. However, a naturalist might have a different methodological concern that is expressed something like this: “Soul-body dualism assumes a fundamental distinction between what is mental and what is physical. Why suppose for methodological purposes that the apparent distinctness of the mental and physical necessarily reveals anything fundamental about their true nature?”11

width=320I believe that this concern about methodology gets things backwards. Thus, the correct question at this juncture is this: “Why suppose that the apparent distinction of the mental and physical does not reveal the true nature of things? Why not stick with ordinary appearances as guides to reality, unless we have a reason to doubt those appearances?” As a methodological point, naturalists usually respond that we cannot stick with ordinary appearances as guides to reality because of the methodological commitment of scientists to causal closure. In other words, naturalists do give an argument for not sticking with ordinary appearances. What I have done in this brief paper is examine that argument and found it wanting. The conclusion, then, is that the argument from causal closure does not give us a good reason to doubt our initial conviction that the natural world is causally open. If naturalists were to respond at this point by saying that the soul-body dualists’ assumption of causal openness in the brain assumes that reasons and purposeful behavior cannot possibly be materially instantiated (that those reasons and purposeful behavior cannot find a sufficient substrate in neural mechanisms), then they (naturalists) would once again get things argumentatively backwards. Given our ordinary soul-body dualist understanding of ourselves wherein we are souls that make undetermined choices for reasons, the burden of argument is on naturalists to give a reason why we must think that reasons and purposeful behavior have to be materially instantiated or realized. Again, naturalists have given an argument—the argument from causal closure, to which I have developed a response. But the burden at the outset is not on the soul-body dualist to explain why reasons and purposeful behavior cannot be materially instantiated. The burden is on naturalists to explain why they have to be materially instantiated, if they are to be taken as real. If the argument from causal closure doesn’t work, and I have argued that it doesn’t, then soul-body dualists await some other reason from members of the naturalist camp that would explain why reasons and purposeful behavior, if they are real, must be materially instantiated.

Finally, I want to raise and respond very briefly to a related methodological concern. A naturalist might respond to my evaluation of the causal closure argument by maintaining that there are viable accounts of choices, purposes, and other mental events or states that do not invoke anything like souls and indeterministic human freedom. Thus, for the naturalist who assumes causal closure, mental events can be physically realized or instantiated and thereby play a role in explaining human behavior (along these lines, see endnote 3). Once again, however, this response is argumentatively wide of the mark. I have in no way argued, and no soul-body dualist need argue, that a naturalist is not free to try to develop an account of human beings that identifies choices, purposes, and other mental events with physical events. How satisfactory such an account might be will of course be a matter for discussion (again, see endnote 3). The question at issue methodologically, however, is whether retaining an ordinary soul-body dualist conception of ourselves in some way or other undermines the framework and practice of science. The causal closure argument is put forth in support of the position that it does. I have argued that this argument fails. What is methodologically appropriate now is for a naturalist to respond to my argument. What is not methodologically appropriate at this juncture is for a naturalist to develop a naturalistic view of ourselves in which what is mental gets identified with or instantiated in what is physical.


I conclude that the causal closure argument fails. If it does, then one of the main reasons for thinking that souls do not exist and cannot make choices that causally influence events in the physical world is undermined. Of course, this conclusion does not establish that souls that make such choices do exist. The reasons one might have for thinking that such souls exist is a topic that is beyond the scope of this essay and one to be pursued on another day.


1 The argument of this section is taken from Stewart Goetz and Charles Taliaferro, Naturalism (Grand Rapids: Eerdmans, 2008), Chapter 2, where we discuss an additional form of the argument that I present in this paper.

2 Richard Taylor, Metaphysics, Fourth Edition (Englewood Cliffs, NJ: Prentice Hall, 1992), pp. 20-22.

3 Jaegwon Kim, Philosophy of Mind (Boulder, CO: Westview Press, 1996), pp. 131-132, 147-148.

4 In fairness to Kim, it is important to note that he too recognizes the counterintuitive nature of the conclusion of the argument from causal closure, which is that our mental lives have no explanatory role to play in accounting for events in the physical world (our mental lives are explanatorily epiphenomenal). Hence, in order to preserve an explanatory role for the mental, he believes that we should be committed to a reduction of the mental to the physical:

Mind-to-body causation is fundamental if our mentality is to make a difference to what goes on in the world. If I want to have the slightest causal influence on anything outside me—to change a light bulb or start a war—I must first move my limbs or other parts of my body; somehow, my beliefs and desires must cause the muscles in my arms and legs to contract, or cause my vocal cords to vibrate. Mental causation is fundamental to our conception of mentality, and to our view of ourselves as agents . . . ; any theory of mind that is not able to accommodate mental causation must be considered inadequate, or at best incomplete. . . . Does this mean that we are committed willy-nilly to reductionism? The answer is no: what we have established . . . is a conditional thesis, ‘If mentality is to have any causal efficacy at all—it must be physically reducible.’ Those of us who believe in mental causation should hope for a successful reduction. Physicalism, or Something Near Enough (Princeton: Princeton University Press, 2005), pp. 152-153, 161.

According to Kim, then, physical reduction (reduction of the mental to the physical) enables us to preserve our belief that mentality makes a causal explanatory difference. Notice however the price that must be paid to embrace this ‘solution’ to the problem of causal closure. Mentality can make such an explanatory difference, only if we give up both the idea that mental actions are ultimately and irreducibly explained by purposes and the view that we have libertarian free will. While Kim is correct when he insists that none of us wants to give up on the idea of mental causation, some of us also do not want to give up the idea that mental causation itself occurs only because mental events such as choices are indeterministic events that are ultimately and irreducibly explained by purposes. Given the high price that must be paid to endorse Kim’s ‘solution’ to the problem of causal closure, it is imperative to examine whether there is a good reason to believe in the principle of the causal closure of the physical world.

5 In maintaining that a neuroscientist as an ordinary human being would surely refer to my purposes in an explanation of my typing, I am not claiming that we are always right when we provide a teleological explanation of another person’s behavior. We might sometimes be wrong. But it is a huge step to conclude from ‘some teleological explanations of behavior are false’ that ‘there is ultimately no role for teleological explanations of behavior.’ Moreover, while I might be mistaken about whether your behavior was merely reflexive or purposeful, I am in a far better epistemic position to know whether my own behavior was purposeful or not. The following words of Alfred North Whitehead are apropos: “Scientists [and, I would add, philosophers] animated by the purpose of proving that they are purposeless constitute an interesting subject for study.” The Function of Reason (Boston: Beacon Press, 1958), p. 16.

6 Wilder Penfield, The Mystery of the Mind (Princeton: Princeton University Press, 1975).

7 Karl Popper says that “the physicalist principle of the closedness of the physical [world] . . . is of decisive importance, and I take it as the characteristic principle of physicalism or materialism.” Karl R. Popper and John C. Eccles, The Self and Its Brain (New York: Routledge, 1977), p. 51. Popper adds, and then argues for the view, that “there is no reason to reject our prima facie view [that the physical world is open to mental, purposeful explanations]; a view that is inconsistent with the physicalist principle.” Ibid.

8 Penfield, The Mystery of the Mind, pp. 76, 80.

9 Richard Feynman, The Meaning of It All (Reading, MA: Perseus Books, 1998), pp. 16, 45.

10 David Chalmers, The Conscious Mind: In Search of a Fundamental Theory (New York: Oxford University Press), p. 153.

11 The methodological objections considered in this section are raised by the naturalist Tom Clark at his website in response to Taliaferro’s and my book Naturalism.