SOPHIA DOI 10.1007/s11841-014-0425-1
Evan Fales on the Possibility of Divine Causation Gregory E. Ganssle
# Springer Science+Business Media Dordrecht 2014
Abstract Evan Fales has argued that divine causation is not possible. His central argument involves an analysis of causation that requires that there has to be a mapping feature to guarantee that the particular effect follows the particular cause. He suggests that being related in space and time will provide the means to map the right effects onto their causes. In this paper, I argue that the spatial relation between cause and effect is not necessary to the causal relation. In cases of volition, it appears that the mapping of particular effects onto volitions is achieved by the intentional content of the volition. Therefore, spatial relations are not necessary to causation and the impossibility of divine causation has not been shown. Keywords Cause . Causation . God . Universals Traditional theists think that God acts in the world. He performs miracles from time to time, and he is causally related to the events that happen in accordance with natural laws. The claim that God is omnipotent is the claim that, if he exists, he can do anything logically or metaphysically possible. An omnipotent being, it seems, can act causally within the world. There, however, is something that is prima facie odd about a nonphysical entity causing physical effects. This oddness has been raised more often in criticism of substance dualism than about God’s causal activity, but it applies to the divine case as well. Both are instances of a general worry over non-physical–physical interaction. Evan Fales, in a recent book, charges theistic philosophers with providing very little by way of explanation of the mechanisms of divine causation. 1 Since the reality of God’s causal activity is central to any theistic view, this poverty of theoretical 1
Evan Fales, Divine Intervention: Metaphysical and Epistemological Puzzles. (New York: Routledge, 2010). Routledge Studies in the Philosophy of Religion. All citations to Divine Intervention will be parenthetical. In his preface, Fales announces that his book is ‘about the God of traditional monotheistic theologies, the theologies of Judaism, Christianity, and Islam.’ I will assume this concept of God when engaging Fales’ argument. G. E. Ganssle (*) Rivendell Institute, New Haven, CT, USA e-mail:
[email protected] G. E. Ganssle 15 Windsor Rd, Hamden, CT 06517, USA
G.E. Ganssle
explanation is a serious lacuna in the defense of theistic belief. Fales goes on to raise several arguments against the possibility of divine causation. In what follows, I shall address two of these objections. I shall conclude that Fales has not given us reason to worry about the possibility of divine causal activity.
Worries over Conservation of Energy Many purported cases of divine causal activity go beyond the kind of concurrence with the ordinary outcome of events that are described by the laws of nature. For example, if the miracles reported in the Scriptures really happened, then they require more than God’s general cooperation with natural laws. If Jesus actually walked on the water, something out of the ordinary happened. Fales makes the observation that these cases require either a violation of the laws of conservation of energy or a strategic exploitation of quantum indeterminacy. In either case, there is interaction between a nonphysical agent and a physical world. Fales himself does not think that appeals to quantum indeterminacy on behalf of divine action will be fruitful. In what follows, I shall follow his lead and leave this issue behind. Does a violation of conservation laws in a case of purported divine activity cause a problem? Fales discusses a distinction Robert Larmer makes between a weak conservation claim and a strong conservation claim. 2 The weak conservation claim is that energy and momentum in a system are conserved, as long as the system is closed. The weak claim allows for the possibility that the physical universe is not itself a closed system. The strong conservation claim is that energy and momentum are conserved, period. That is, the universe is immune to the introduction of new energy. Larmer points out that the strong claim entails the weak claim but the weak does not entail the strong. He goes on to emphasize the fact that all of our evidence supports the weak claim rather than the strong claim. Fales responds: But is Larmer right about this? I suggest that he is not. I suggest that we have evidence—abundant evidence—that the only sources of energy are natural ones. Our evidence is just this: whenever we are able to balance the books on the energy (and momentum) of a physical system, and find an increase or decrease, and we look hard enough for a physical explanation of that increase or decrease, we find one. There is no case in which given sufficient understanding of a system, we have failed to find such a physical explanation. (16) Fales’ response to Larmer’s distinction between strong and weak conservation claims is puzzling. Rather than posing a challenge to the employment of the distinction between the strong and the weak conservation claim, his response strongly supports Larmer’s use of it. It may be true that whenever we are able to balance the energy books we find a physical explanation for any change in energy and momentum. The only cases in which we come close to balancing the books, however, are carefully constructed to guarantee that the systems remain closed. These cases never involve, for 2
Fales is responding to Robert Larmer, Water into Wine: An Investigation of the Concept of Miracle (Montreal: McGill-Queens Press, 1988).
Evan Fales on Divine Causation
example, the actions of intentional agents. They do not even involve living things. Systems that involve conscious agents seem to be too complicated and so defy our ability to balance the books. Fales’ argument here relies on the assumption that the cases in which we can balance the books are representative of all cases, even cases of intentional actions and free choices. The abundant evidence we have supports the claim that in carefully controlled systems—constructed to guarantee that they remain closed, energy and momentum is conserved. This claim is not very different from the weak conservation claim. It gives no support at all to the strong claim. The strong claim is the sort of universal claim that cannot be well supported with such a piecemeal testing process. It requires either a very strong inductive argument or a metaphysical argument to the effect that the universe is a closed system. The strong inductive argument requires the inclusion of enough sufficiently complex cases to render the sample relevant to the strong claim. The metaphysical argument amounts to an argument that no God who interacts with the world exists. The assumption that the universe is a closed system cannot, without circularity, be employed to defend the claim that divine causal activity is not possible.
Worries About the Nature of Causation Fales’ more important argument against the possibility of divine causation is built upon certain requirements that are part of his analysis of the nature of causation. It is clear that whether divine causation is possible depends on what the causal relation is. Once we have a position about what is required for causal relations to hold, we can see if this criterion can be met in the case of God’s purported activity. While philosophical analysis of the nature of causation is difficult, Fales argues that we can make enough progress to size up whether a God who interacts with the world exists. A comprehensive defense of divine causal activity, then, would need either to canvass all of the more plausible analyses of causation or provide a good argument that one analysis is better than the alternatives. In this short paper, I shall concentrate on the analysis of causation that Fales prefers, and I shall assume it is mostly correct. Fales claims that causal relations are relations between universals. He writes that ‘causation is a second-order relation, C, between (some) first order properties.’ (59) Many properties, though not all, stand in causal relation with other properties. Each such property has its causal relations essentially. That is, the property F would not be the same property it is if it had causal relations other than those it in fact has. Fales recognizes two implications of this position. The first is Platonism, in which these properties exist even if they are uninstantiated. The second is that all of the causal properties that exist are bound together in a ‘nomic web.’ Since the ‘essentially related to’ relation is transitive and symmetrical, each causal property is essentially related to every other causal property. 3 If any causal property that actually exists did not exist, none of the properties in the actual nomic web would exist. 3 One referee objected to each of these implications for what seem to me to be good reasons. Platonism about properties is highly contentious (the reviewer thought it was ‘almost certainly false’). The nomic web implies that my touching a key on my keyboard has a property that is essentially causally connected to a property of a particle of dust a hundred light years away. While this claim is hard to believe, I will leave these criticisms aside. My own critique of Fales’ position can grant these notions.
G.E. Ganssle
Let us consider an ordinary causal episode. To take an example familiar to those of us who are former teenage vandals, suppose a rock hits and breaks a window. The rock hitting the window (call it event 1) 4 causes the window to break (event 2). Event 1 is a token of the type being a rock hitting a window in a particular way (event type A). Event 2 is a token of being a window breaking in a particular way (event type B). Being of type A and being of type B are the properties that stand in the relation C to each other. So, causation in this case is the relation C that holds between the properties being an event of type A and being an event of type B. Even if no event A actually occurs (that is, no rocks ever hit windows in that particular way) the property being an event of type A (that is, being an event of the type of rock hitting window) stands in the causal relation to the property of being an event of type B. Standing in this relation is essential both to being an event of type A and to being an event of type B. Furthermore, it is the causal relations that serve to individuate these properties from any other properties. The fact that these properties necessarily stand in the particular C-relations to one another that they do, Fales argues, provides the metaphysical ground for laws of nature. What is more fundamental than the laws of nature is the existence of physical objects and their properties (along with the relations that are necessary to those properties). So, a law of nature that describes the fact that a window will break if a rock hits it is grounded on the fact that the property of being an event of type A (a rock hitting a window in the right sort of way) necessarily stands in the causal relation to the property of being an event of type B. With this analysis in mind, we can turn to Fales’ argument that God cannot stand in causal relations. In short, Fales’ argument is that causal relata must be both spatial and temporal. God is not spatial and many philosophers think he is not temporal. 5 Therefore, God cannot stand in any causal relation. The claim that causation requires spatiotemporal relation is not directly a consequence of Fales’ analysis of causation, however. It is supported by another requirement of his position. Fales recognizes that the position that properties of physical things stand in causal relations with properties of other physical things is not sufficient to explain why a particular token event caused the particular token event it did rather than another token event of the same kind. He explains: Suppose the law is this: that every A-event is correlated with a B-event. But which B-event? It cannot be just any B-event; this A-event is presumably the cause of some determinate B-event, and not other such events. At the level of the metaphysical ground of the law, we have (so far) simply a C-relation between 4 Not every case of a rock hitting a window will cause the window to break, of course. It must be a case of the rock hitting the window with sufficient force and there being in place whatever other necessary conditions are relevant for the window to break when a rock hits it in this way. When I refer to the event of a rock hitting the window in a particular way, I shall assume all of these other conditions are satisfied. For simplicity, I will continue to write as if what stands in causal relations are events. I have become convinced that this is not the case. Substances stand in causal relations. See E J Lowe, Personal Agency: The Metaphysics of Mind and Action (Oxford: Oxford University Press, 2008). 5 Most philosophers writing on the subject today hold that God is in some sense temporal. The traditional view is that God is not temporal. For contemporary discussions, see the essays in Gregory E. Ganssle ed., God and Time: Four Views (Downers Grove: Inter Varsity Press, 2001), and in Gregory E. Ganssle and David M. Woodruff ed., God and Time: Essays on the Divine Nature (New York: Oxford University Press, 2002).
Evan Fales on Divine Causation
the universals A and B. But instances of the relation—individual A-events causing individual B-events‐ must generate a selective mapping that assigns particular Bs to particular As. What is the ontological explanation for this mapping? A natural answer is: the C-relation itself. But the C-relation is a universal; what ties a given A to a given B must be a given instance of the C-relation. Which instance? It will be the one that is tied to the relevant instance of A, of course. But in what manner is it ‘directed toward’ a given instance of B? (62 italics in original) Fales sees that his analysis of causation requires an answer to the last question of this paragraph. That is, how is an instance of event A directed toward a given instance of B? Holding causation to be a relation between universals leaves one with what we could call the problem of particularity. It is not just that rocks hitting windows cause windows to break, but that this particular event of the rock hitting the window (this rock, this window, this place, this time) causes this particular window to break (same window, same place, same time). There must be some feature of the relation between these events such that the particular effect, and not another effect of the same type, follows the particular cause. Fales refers to this requirement as a ‘selective mapping that assigns particular Bs to particular As.’ A solution to the problem of particularity is what Fales’ account of causation requires. Fortunately, there is a ready solution at hand: ‘There is a natural answer to this question, namely that there is “built in” to the C-relation the requirement that the relevant B-event stand in some spatiotemporal relation to its cause.’ (62) What solves the problem of particularity, on Fales’ view, is the spatiotemporal relation between the cause and the effect. The reason this particular rock hitting the window causes this particular window to break is that the particular window is located at the appropriate place and time, exactly where the rock makes contact. The spatiotemporal relation between these token events guarantees that the particular effect is caused rather than another. The problem of particularity is solved. This solution is perfectly adequate for many cases of causation. That is, if there is the right spatiotemporal relation between the particular cause and particular effect, the problem of particularity may be solved. Fales’ position is stronger than the claim that the spatiotemporal relation is sufficient to solve the problem of particularity, however. His claim is that the spatiotemporal relation is a necessary condition for any case of causation. He holds that the spatiotemporal relation is built in to the causal relation. Causes must be spatiotemporally related to their effects. If he is right about the spatiotemporal requirement being built in to the nature of causation, then he is correct in his claim that every cause must be spatial, and God cannot be a causal agent. Fales seems to reason from the observation that spatiotemporal relation will be sufficient in many cases to the view that it must be the only solution to the problem of particularity possible. He then uses this conclusion to argue that there can be no divine causal activity. Of course, the first inference assumes that all causal relations are relations between the properties of physical particulars. This assumption cannot serve as a premise in an argument that there can be no physical–non-physical causal interaction without circularity. It may be possible to achieve the mapping of particular effects to particular causes (and solve the problem of particularity) without relying on spatiotemporal relations. If so, then Fales’ claim that spatiotemporal relations are necessary to causation will be
G.E. Ganssle
shown to be false. We will grant that a necessary part of the causal relation is some kind of built-in directionality—a mapping feature. Spatiotemporal relations work for those causal scenarios whose relata are wholly spatial and temporal. What Fales does not consider in this part of his argument is that there may be other means of accommodating the need for the cause to be directed toward the effect. In any discussion of non-physical causation, it is difficult to provide examples that will not be rejected out of hand. Physicalists will not find any example plausible. Yet the philosopher defending the possibility of non-physical causation has to be able to use some example or other in order to discuss how non-physical causation might work. I shall use mind–body dualism as an example. Even those who do not find dualism plausible understand its basic claims. Thus, we can look to see how a dualist might solve the problem of particularity. 6 The insights we gain, then, may be applied to the divine case. If Fales is right about causation, it should be impossible for a dualist to solve the problem of particularity. But it seems that a dualist can offer a solution. A dualist will analyze intentional choices in a way that allows there to be sufficient mapping of token effects to token causes. She will appeal not to spatiotemporal relations but to intentional content. Let us say I wish to read the mystery, The Cat Who Sang for the Birds. 7 In front of me are two copies of the book side by side. Each stands in similar enough spatiotemporal relation to my volition. I choose to pull the left one off the shelf rather than the right. What accounts for this particular effect following my volition rather than the other effect? There is a perfectly adequate explanation at hand. I chose to pull the book on the left. I formed a volition that had as part of its content the book on the left rather than the book on the right. When a human being acts volitionally, there is a built-in directionality to the causal episode, yet the built in directionality is not fully explained, it seems, by whatever spatial relation holds between the relata. The cause is directed toward the effect because of the content of my volitional state. My will has an intentionality that is directed toward the one book rather than the other. 8 To take another case, suppose I think of my wife and then I remember that she asked me to start cooking dinner. These thoughts are not physical things, again supposing dualism to be true, yet the first caused the second to happen. What solves the problem of particularity in this case is the content of my first thought. That content causes the content of what Jeanie asked me to do to come into my consciousness. Of course, I am not arguing that dualism is true. The point is that we can grant that our analysis of causation needs some kind of mapping mechanism to solve the problem of particularity without holding that this mechanism must involve a spatiotemporal relation. In the case of my own agency, it is the intentional content of my volition that specifies the outcome. Even if dualism is false, it seems that it is the intentional content Jaegwon Kim has introduced what he calls the ‘pairing problem’ as a challenge to dualism. His pairing problem is close to Fales’ problem of particularity, although Fales does not refer to or discuss Kim. See Jaegwon Kim, Physicalism, or Something Near Enough (Princeton: Princeton University Press, 2005): 78–88. See also ‘Lonely Souls: Causality and Substance Dualism,’ in Soul, Body, and Survival: Essays on the Metaphysics of Human Persons ed., Kevin Corcoran (Ithaca: Cornell University Press, 2001): 30–43. 7 Lillian Jackson Braun, The Cat Who Sang for the Birds (New York: G. P. Putnam, 1998). 8 Even if dualism is true, the human being is embodied and thus it makes sense to describe one book as being on the left and the other as being on the right. Since God is not embodied, he will pick out the books without reference to the spatial description of being on the left or being on the right. 6
Evan Fales on Divine Causation
that explains why the particular effect follows from the cause, however that content is represented or embedded or located in the brain. It is not that Fales has ignored this kind of agency in his book. He has not discussed it in the context of the need for mapping particular effects to causes, however. Prior to his discussion of the particularity problem, he raised what he thinks are decisive objections to thinking of human agency as relevant to the investigation of causation. If this objection can be shown to be wanting, as I shall argue it is, then the way is open to recognize that the intentional content of an agent’s volition can be brought to solve the particularity problem.
Fales’ Objections to the Relevance of Agency The main objection Fales raises to agency is that there is a relevant difference between our experience of the everyday physical pushes and pulls on our bodies and our experience of our volitions and their effects. Our experience of physical causation is sufficient for us to have a concept of causation while our experience of our own volitions is not. Furthermore, he implies that we have an immediate experience of the pushes and pulls on our own body while our experience of our volitional acts is mediated or indirect. The distinctions Fales makes between our experience of our volitions and our experience of physical pushes and pulls are drawn from the observation that we are unaware of the many steps in the chain between our volitions and our actions in the world. In the case of physical causation, we have, I believe, at least an acquaintance with the basic process of pushes and pulls which, I have elsewhere argued, we actually experience. If that is correct, then we possess at least a concept answering to the notion of physical force—a concept, so I would argue—that is more modally robust than Hume’s proposed constant conjunction. David Armstrong has suggested that we have similar acquaintance in experience of psychophysical causation, as when we engage in voluntary and intentional bodily activity. Of this I am less convinced. Hume’s own observation that when I voluntarily raise my arm, I am in no sense aware of the causal process, involving the nerves of the brain and arm, by means of which the volition effects the arm motion, urges caution. If that is so, then psychophysical causation in humans remains a deep mystery; and so, a fortiori, does theo-mundane causation. (43 italics in the original) Elsewhere, he argues that the causal connection between our volitions and our actions is not immediate, although they seem to be so: Hume has what seems to me to be a decisive objection: the causal connection between volitions and overt bodily action seems to us to be immediate, but in fact it is not; it is mediated by a complex chain of events in motor nerves that activate the relevant muscles. Yet, we are not aware of these neural processes at all—and hence not directly aware of the causal facts in this case. (165, note 10.)
G.E. Ganssle
Fales’ objection to our drawing upon our own volitions for our understanding of causation, then, is that we do not have the right sort of experience of how our volitions cause the movements in our bodies. The implication is that we do have the right sort of experience of causation between physical things and our bodies. He refers to Hume who emphasizes that although it appears that our willing is a direct cause of our action, the mechanism is actually quite complicated. Furthermore, the complicated sequence of events in our nerves and muscles is something of which we have no awareness. It is instructive to recognize, first of all, that Hume’s comments are given in a significantly different context than the concerns of Fales. Hume is attempting to discover from which impressions we get the idea of ‘power, force, energy or necessary connexion.’ 9 He thought that each idea must be either traceable to a particular impression or built out of ideas, each of which can be traced to impressions. As Fales mentions, Hume argued that it is impossible to trace the idea of power or necessary connection to our impressions of our own agency because the causal chains are too long and involved. We learn from anatomy, that the immediate object of power in voluntary motion, is not the member itself which is moved, but certain muscles, and nerves, and animal spirits, and, perhaps, something still more minute and more unknown, through which the motion is successively propagated, ere it reach the member itself whose motion is the immediate object of volition. Can there be a more certain proof, that the power, by which this whole operation is performed, so far from being directly and fully known by an inward sentiment or consciousness, is to the last degree, mysterious and unintelligible? (Hume, 66) Hume’s conclusion is that our idea of cause and effect cannot come from the inner impression of the power of our volition. This claim, however, comes only after he has concluded that ‘external objects as they appear to the sense, give us no idea of power or necessary connexion…’ (Hume, 64) For Hume, neither our experience of physical things in the world nor our experience of our own volitions is sufficient to give us the concept of power or necessary connection. The conclusion Fales draws from the distinction between grasping causation through our experience of our own agency and grasping it through our experience of the world is not one with which Hume would be happy. Secondly, what we learn from anatomy about the complicated neural processes that accompany our willing to move something in the world applies just as much to our experience of the pushes and pulls of physical things against our body. For me to be aware that my cat is brushing up against my leg involves the complicated translation of the physical stimulus into neuro-electric impulses that travel from my leg into my brain and are processed by my brain in such a way that I am conscious of the pressure—but I am conscious of it not as happening in my brain—where the processing occurs, but as happening in my leg where the cat is. It can be argued that my knowledge of the cat brushing my leg is mediated by as complicated a series of events as my forming the volition to pick the book on the left and my doing so. In this case, the distinction Fales 9
David Hume, Enquiries Concerning Human Understanding. Third Edition ed., PH Nidditch (Oxford University Press: Clarendon Press, 1975) 62. All references to this book will be cited parenthetically.
Evan Fales on Divine Causation
appeals to in order to rule out the role our volitions might have in our analysis of causation disappears. Thirdly, the question of what counts as a direct or indirect object in a case of knowledge or of volition is very different than in the case of a causal sequence. Both Hume and Fales conflate these domains. In a causal sequence, only one item is the proximate cause of an effect. 10 All others are non-proximate. We could say that the proximate cause is the direct cause of the effect and the rest of the events in the causal sequence are indirect. The indirect causes contribute to the effect by causing other items in the sequence that ultimately produce the effect. The proximate cause produces the effect without the mediation of any other item in the causal chain. In the case of knowledge, the direct object is what is known without being mediated by any other objects of knowledge. An indirect object of knowledge is one that is known in virtue of our knowing another object directly. If I see a game on TV, the television is the direct object of my vision. I see the game indirectly, in virtue of my observing the television set. Although the television is the direct object of my perception, my visual experience involves a complicated causal series of events. The presence of this causal story does not render my perception of the television indirect. The television is known directly, not in virtue of any other object of knowledge. The case of volition has some similarities to the case of knowledge. The direct object of a volition to act might be the movement of the body one aims to employ in order to accomplish the task. So, when I will to turn off a light, the direct object of my willing might be the movement of my hand. There is a complicated causal story between my volition and my hand’s motion. What has my volition as a proximate cause is most likely an event in my brain. But the object of my willing is not this event. There is a big difference, then, between the direct and indirect objects of knowledge or volition on the one hand and the proximate and secondary causes on the other. The important point here is that to know something directly, it is not required that the object of knowledge be the proximate cause of the knowing. In the same way, one can will directly to act without the volition being the proximate cause of the effect. In each case, there are complicated causal sequences about which we are unconscious as we are willing or knowing. The fact that there is a complex causal story in the working out of our volitions, therefore, does not render our agency irrelevant to our investigation of the nature of causation. Fales’ objection to bringing our experience of our own volitions into our investigation of causation is not successful. As a result, we can draw upon the resources of human volition in solving the problem of particularity. 11 We can grant that there must be a mapping feature built into the causal relation to guarantee that the particular effect B maps onto a particular cause A. What this feature is can vary with the kinds of causes under discussion. For some physical causes such as a rock hitting and breaking a window, the spatiotemporal relation is sufficient to map the cause to a particular effect. It is possible that when a human agent acts, it is the intentional content of an agent’s 10 There is only one proximate cause, assuming for simplicity’s sake, that there is no causal overdetermination at work. In cases of causal overdetermination, there are comparatively few proximate causes. 11 If there are other objections to appealing to intentionality in order to solve the problem of particularity, these will have to be met. Fales raises no other objections, so as far as his case is concerned, the way is open to appeal to intentionality.
G.E. Ganssle
volition that maps the cause to the intended effect. The problem of particularity is solved in each case. Each of the two challenges to the possibility of divine causal activity, then, can be met. First, Fales does not provide the kind of grounds he thinks he does to prefer the strong conservation claim to the weak conservation claim. Therefore, the theist can hold that divine agency introduces new causal chains into the universe, even if the introduction of these new causal chains requires that the universe be an open system. Second, the intrinsic intentionality of the volition of an agent provides all of the directionality that is required to solve the particularity problem. Therefore, our confidence in the possibility of divine agency is not shaken by Fales’ arguments to the contrary. Acknowledgements I benefitted greatly from the discussion of this paper at the Notre Dame Philosophy Club. Especially helpful were comments there by Cameron Cortens, Katie Finley, Ryan Solava and Luke Potter. I also presented this work at the Society of Christian Philosophers northeast regional conference at Fordham University in Manhattan, March 18-19, 2011, the regional meeting of the Evangelical Theological Society, April 9, 2011, and the national meeting of the Evangelical Philosophical Society in November of that year. I presented much ofthis material as the Rivendell Integration Lecture on January 23, 2012. I am grateful for the helpful discussion at these events as well. Anonymous reviewers for this journal were especially helpful.