Philosophical Thoughts: Aquinas and Swinburne

St. Thomas Aquinas does not often find kindred spirits among contemporary analytic philosophers. Among those not particularly sympathetic is the imminent Richard Swinburne. It is well known among his readers that Swinburne carries no brief for classical theism, the concept of God attributed to St. Thomas. What is regrettable about this situation is how Swinburne stands athwart or ignores Aquinas where the Angelic Doctor could potentially be helpful. In this post, I try to offer a case in point.

In chapter five of The Existence of God (Rev ed. pp.70-89). Swinburne goes through a treatise on complete explanation. The primary thrust of this idea is there are theories that, if you were to try going beyond it, would not yield any further explanatory power. Swinburne distinguishes between full, complete, ultimate, and absolute explanations.

A full explanation explains by itself why something happened, regardless of whether there is any further explanation of it. X can be a full explanation of Y whether or not there is an explanation of X. If we had to pursue explanations all the way down, we would never be able to explain anything without appeal to the origin of the universe. Full explanations can get us quite far. Far enough to make good arguments about causes and explanations.

A complete explanation is a special kind of full explanation. It takes into account other factors in an occurrence for which there is no explanation for their existence or operation in terms of other factors operative at the time of their existence. A complete explanation of X by Y would, it seems to me, take into account factors of Y.

An ultimate explanation is a special kind of complete explanation. This type of explanation is where we pursue explanations and factors of an occurrence all the way down until we get to ultimate brute facts. If X explains Y, and W explains X, and so forth we eventually hit a brute fact for which there is no further explanation. Perhaps this is V. V explains W, W explains X, and X explains Y. But V is a brute fact. 

An absolute explanation is one where brute facts are not cited; all factors in an occurrence are either self-explanatory or logically necessary. Swinburne does not think there are absolute explanations of logically contingent phenomena, nothing explains itself (76). For Swinburne, arguments for the existence of God are arguments to a complete explanation of phenomena.

The issue Swinburne has is that he believes Aquinas is arguing on a priori grounds that any full explanation has a complete explanation. Swinburne thinks that in some cases it is probable that phenomena with full explanations have complete explanations. In the appendix to chapter five, Swinburne takes up some thoughts on causation from Scotus and Aquinas.

When looking at Aquinas’ causal reasoning, Swinburne uses the common example of the train cars in discussions of essentially ordered causes. There is an infinitely long train, each car pulls the other one. For Swinburne, the car X pulling car Y is a full explanation of Y’s movement along the track. There is no need to appeal to anything further in explaining the motion of Y, even if there are other things going on somewhere. Earlier, Swinburne referred to one form of opposition to this idea as the “completist fallacy.” He believes Hume commits this in Dialogues when Philo raises the “who designed the designer” rebuttal to the design argument. Swinburne thinks Hume is wrong because, presumably, the Designer is a full explanation of the order of the world. Perhaps it is not a complete explanation, but it is full or at least sufficient to explain the phenomena in question such that no other theory or hypothesis would yield greater explanatory power. Swinburne thinks Aquinas commits the same fallacy as Hume because Aquinas believes that there is no explanation or cause of Y unless there is an engine (‘first cause’ in metaphysical terminology) that moves the entire train. For Aquinas, X is not a full explanation for the movement of Y.

Swinburne is at least correct in that Aquinas would agree that X is not a full explanation of Y. But Aquinas would hold that X certainly had a causal role in this series, just not an essential causal role. X would not be sufficient to explain the movement of Y, for X, in this case, does not actually have any causal power on its own to effect movement. X is only an accidental cause of Y, and ultimately cannot fully explain the movement. Further, X only explains Y insofar as they are connected. But X could be removed. Y would then be connected to W, and still in motion.  The question would remain; what then accounts for or sufficiently explains the motion of Y? If X could in principle be removed, and Y connected to W with the motion of Y continuing, then X cannot be a full explanation of Y. There is not enough explanatory power in isolating X to explain the movement of Y, given that we know in this example that X does not have motive power of its own. X must get its motive power from another source, as is the case for Y.

Swinburne refers to Aquinas as seeking an a priori argument for his position. Swinburne finds there is no argument on these grounds against an infinite regress of essentially ordered causes. From my perspective, it seems Swinburne means there is nothing bound up in the terms at issue, viz. essential cause, accidental cause, effect, and so forth that preclude one from holding without contradiction there could be an infinite regress. That is, there is nothing incoherent in supposing there could be an infinitely long train in which each car simultaneously moves another. It is problematic to attribute a priori reasoning to Aquinas since he believed we got our knowledge of things and their causes from direct interaction with the world. Leaving this aside, for now, Swinburne may be right as far as it is stated. There may be nothing a priori incoherent or contradictory in making this assertion. But perhaps this could be challenged, given what we know about train cars, motion, and so forth. Maybe even as bound up in their terms or definitions. For example, we know train cars have no motive power of their own. We know that things without motive power cannot move themselves. We know that no matter how many train cars there might be, perhaps an infinite number, there must be something with motive power that is moving them if they are in fact moving. When Swinburne says “each truck [train car] by its own motion simultaneously makes the other truck move…” (89) I think the major disagreement here is the reference to “…by its own motion…”. Aquinas would not agree that the train cars (again, just using this as the example of essentially ordered versus accidentally ordered causes) have their own motion. Since train cars do not have their own motion, they must get it from another. And thus, it is not possible that the train cars could do anything by their own motion. X does not have its own motion, per se. Neither does W. Y is not moved by the motion of X per se, because X does not have any motion or motive power per se. It is only in virtue of an engine that any motion is conveyed or available to X, which then passes through X to Y. At best, X is only a partial explanation of the movement of Y. Only as an accidental cause of the motion of Y is X an explanation of Y’s movement.

The reason why I think adopting Aquinas’ reasoning could help is this. Swinburne refers to causal powers of bodies when comparing ways of scientific thinking (42-44). He argues that the ‘ancient’ way of thinking that bodies have powers and liabilities (what Aquinas might say are act and potency) are a better way of thinking about scientific explanation. Swinburne rightly seizes on the modern move to think of laws and events independently of bodies (or substances). I think this breakdown helps Swinburne’s account of agent causation, which is that such causation is not reducible to events or laws, and is (for Swinburne) not reducible to ‘bodies’ as such since agent causation is not species of liability (i.e. potentiality) of a body.

If he recognizes that bodies (substances) have powers and liabilities (e.g. they are combinations in some way of act and potency), then it follows that some bodies do not have within themselves the powers to do certain things. The stone does not have within itself the power of locomotion. Neither does the stick. It is the hand of the person moving the stick, which moves the stone. The stone has within itself the power or liability to be moved. So does the stick. But they cannot bring these about without the simultaneous action of the hand of the person (agent) moving the stick. It is not possible that the stick is a full explanation of the movement of the stone. But this is what Swinburne seems to be claiming in his critique of Aquinas. He seems to be saying that the motion of the stick, which does not have in itself the power to move anything based on what it means to be a stick, is all that we need to sufficiently explain the movement of the stone. On this point, I believe Aquinas would disagree. And I think Aquinas has a much stronger argument. Swinburne’s approach on this point seems to cut off the causal explanation process right when it starts to get interesting. This could be due to the very scientifically focused methodology, where the arguments start to spill over into metaphysics with reasoning tools that are made for not wholly adequate for that application. If he were to take Aquinas’ notion of essentially ordered causes, Swinburne would have a more robust account of explanation, which would help his overall case. It would represent great consistency, especially with the appeal to the natural scientific reasoning about bodies that he referenced earlier. 

Because...Why? Making Causal Inquiry Interesting

Causation has long engendered deep and passionate discussion. This springs from our innate desire to know. We want to know why and how things are the way we observe them. Why do unimpeded objects always fall back toward earth when we throw them up in the air? Why does the universe exist? What explains variation and complexity in biological species? Why does my child think it a crime against humanity to sleep at night? Countless other questions could be raised as examples. But is causation interesting in itself? Metaphysicians certainly think so. Few others, including philosophers, take much notice. I think most people give perhaps only a passing thought to causes because they unreflectively accept a very narrow, and highly problematic, view of causation. They take temporally sequenced events as a de facto explanation of causes. Not only is this position untenable, it is (worse) completely uninteresting. In what follows, I will offer a few brief thoughts on why this is the case.

When most people think of A causing B, they think that A moved B from point 1 to point 2. Or that A made B from a collection of raw materials, so A caused B. Or, if B regularly follows A, then A is the most likely cause of B. Causation is typically bound up with some type of temporal succession of events. Ceteris paribus, if C is always observed to follow B, and B follows A, A is thought of as the probable cause of C.

The preceding description of causality is precisely what David Hume had in his gunsights. Such events, loose and separate, as it were, cannot really tell us anything. We cannot empirically observe “causal powers” and we have no logical contradiction in B not following A (no matter how regularly it does), so causality is committed to the flames. Sure, we could say that A caused B, but there is no state of affairs where A is causing anything, whether it is B, bicycles, or bananas.

If we buy into the strictly temporal/sequential view of causality, then it is very difficult to demonstrate why Hume is wrong. Sure, his conclusions might be intuitively absurd. But, as Hume would no doubt rebut, “so what?” Intuition does not falsify his position nor establish causality (so understood). Other responses to Hume, such as appealing to possible worlds, counterfactuals, and so forth do not, it seems to me, really get at the heart of Hume’s argument. In the end, I think it is very difficult to respond to the portly Scotsman without appealing to an Aristotelian conception of nature and causes. Most contemporary philosophers will not make this move and will largely leave causation neglected in the corner of a dark room. The Stagirite looms outside, not being invited to the party at all.

The most interesting questions, such as why and how, are answerable on Aristotle’s view. And satisfyingly so. We seek explanations and we find them. Take a very simple example; my tan floor tile. Not much interesting about that, right? But what if we started with a few questions that we could ask about anything in our daily experience? Like why is the floor tile here? Am I the cause of the floor tile under my foot? After all, six months ago, I bought it from the store, mixed the mortar, cut, and laid the tile on the floor. Yet, on an Aristotelian view of causation, it really cannot be wholly said that what I did previously is the cause of the tile under my foot.

We could perhaps say that I was the cause of the tile getting where it is at a time in the past. But why is it here right now? Why does it keep the qualities it has, like color, size, weight? If we really think about it, there are myriad other factors involved, such as the subfloor, house foundation, earth, and all the other laws of physics that must hold constant for the floor tile to be under my foot at any given moment. The reason the tile is under my foot right now is not explainable in any interesting way by the fact that I transported and laid it. Nor does the metaphysical interest lie with the tile manufacturer. When we are concerned about causality, the less relevant thing is the preceding event in time. This is because the prior temporal event does not really answer the basic metaphysical questions we are after. We want to know why this and not something else. The floor tile could, in principle, not be under my foot right now (even if it was in the prior second). We even might put something like this back to Hume. In any event, the temporal sequence might be metaphysically interesting in some way to us, such as understanding the process by which one might lay floor tile, or how floor tile is made and successfully transported. Yet there would be underlying metaphysical questions for each of these events that would necessarily force us into a deeper inquiry.

What if we just stopped at the manufacturing of the floor tile? Is there anything about this that assures it will stay the same shape and color on the way to my house? Here, again, we could resort quickly to various laws of physics and chemistry. But, what about those? Surely these physical laws and chemical compounds held constant during each moment of my journey. At any moment of investigation along the way, I would have the same question. Why this, and not that? Aristotle gives us answers. Formal, material, efficient, and final causes each represent actualized potentials. Thus, we can understand the cause(s) of the floor tile in a multifaceted way. We need not stop with a sequence of events. Such a sequence is accidental to the tile itself as it stands (or is stood upon) right now.

If we are only focused on causality in terms of temporal events, then we will ultimately lack sufficient explanation for anything. To take another example, as I look out into my front yard, I see my son’s Big Wheel. I could say the cause of the Big Wheel in my front yard is that my son left it there yesterday afternoon. Maybe I could press on and ask how the Big Wheel got to my son. And then we could talk about the plant, factory, workers, shipping, etc. Ultimately, we would just be explaining each thing by appealing to the prior event ad infinitum. However, this does not really tell us much. Why not ask why the Big Wheel exists as it does? Why does the Big Wheel not spontaneously melt or turn into a pumpkin? Why not ask why the Big Wheel does not collapse into non-being? These are metaphysically interesting questions. Such questions help us understand reality.

We can explain the existence of the floor tile or the Big Wheel in Aristotelian causal terms in a layered analysis, going so far as we desire in hierarchically ordered causes until terminating in Pure Act. Or, we can stop with a more basic, coherent, and robust understanding of the physical object; existing as a substance with various accidents, having a certain a form, material, efficient, and final cause(s). In any event, we can have a deep metaphysical conversation starting with the simplest objects. I am not aware of another system from which such a fruitful discussion can spring forth. This is especially true of efficient causality, which should be conceived as much more than the mere bringing forth of something from one state to another as most artifactual examples entail. This is, of course, one of the many benefits Aquinas brings in his Aristotelian-grounded philosophy and not an incidental reason why one should consider moving beyond thinking of causes merely in temporally sequenced terms.