## H The Impossibility of a Synchronically Infinite Regress of Explanations

But what entitles Aquinas to deny the possibility of going on ad infinitum in such a causal series? The most fully satisfactory answer I know is the one developed by Rowe. Suppose that A is a dependent being whose existing right now is explained by B's current sustaining activity, and that B's sustaining of A is explained by reference to C. 'Can we now say', Rowe asks, 'that the explanation for the fact that the causal activity of causing A to exist is now going on might be found in B? It seems clear we cannot' (1975a: 26

26 In the passages I'm drawing on, Rowe is in fact developing an interpretation of Aquinas's attempt to block an infinite regress, but the attempt Rowe is focusing on is the one in the Second Way, which can be read along these lines only if the Second Way is interpreted as concerned not with coming into existence but rather with remaining in existence, an interpretation Rowe adopts, ascribing it to G. H. Joyce (Rowe 1975a: 27 n. 9). I think Joyce's line of interpretation is badly suited to the Second Way, but fits G6 well; so I think Rowe's explanation of the blocking of the infinite regress of sustaining causes is better suited to G6 than to the Second Way.

In keeping with a later medieval tradition, Rowe calls a causal series of this sort 'essentially ordered'. 'Now', he says, if C is causing B to be causing A to exist, then since we are operating within an essentially ordered series it also will be true that C is now causing A to exist. C, therefore, will be exhibiting that very sort of causal activity we are trying to explain. And if C is the first member of the series, we might be able to explain why the causal activity

causing-A-to-be-now-existing is now going on by reference to C.

27 Only the first two hyphens in the italicized phrase occur in Rowe's text; I've supplied the others.

However, if C is an intermediate cause, if some other thing is now causing C to be causing A to exist, then we cannot find the explanation for the fact that this activity is going on by reference to C. What then if the series progresses to infinity? Each member of the series will be right now exhibiting the causal activity we are trying to explain. It will be true that every member of the series is exhibiting the causal activity in question and also true that the fact that the causal activity is going on cannot be explained by any member of the series. For any member we select, it will be true that it is caused to exhibit the activity in question by some other member and, therefore, true that we cannot explain the fact that this sort of causal activity is going on in the universe by reference to that member. . . . [I]f the series proceeds to infinity there will be no explanation of the fact that a certain sort of causal activity [causing A to be now existing] ... is going on in the world. (Ibid. 34-5)

end p.106

And therefore, I would add, there could not in that case be a philosophically satisfactory, metaphysical explanation of the fact that A—or S—is now

continuing to exist.

28 Rowe carefully distinguishes between 'two different items: /'. the fact that A now exists, and //'. the fact that a certain sort of causal activity (causing A to exist) is now going on' (ibid. 33). His apparent reason for doing so is that 'Someone might argue that, even though B is not the first member, we can still explain item (i) by reference to B and B's causal activity vis-á-vis A. I do not wish to dispute this point. To say that we have not really "explained" the present existence of A until we explain why B is causing A to exist, tracing each step backward until we arrive at an ultimate first cause, may be nothing more than a confusion as to the nature of explanation' (ibid.; see also the sentence on pp. 34-5). But the situation Aquinas is concerned to characterize as no explanation at all is not one in which an ordinarily adequate sort of first-level explanation has been captiously rejected as insufficient. It is, instead, one in which the first-level explanation is in terms of something that is itself theoretically inexplicable. In such a situation no one with a philosophical interest in understanding A's presently existing could consider its being referred to B's causal activity to constitute any explanation at all.

A's—and, therefore, S's—existing now would be a brute fact, theoretically inexplicable, '//the essentially ordered series of causes resulting in A's present existence proceeds to infinity, lacks a first member' (ibid. 35-6).

Aquinas doesn't take the brute-fact alternative seriously, whether in G6 or anywhere else. As Rowe quite rightly observes (ibid. 36-7), that fact about Aquinas shows that he assumes or considers self-evident some form of'the Principle of Sufficient Reason [PSR], a principle that in its strongest form maintains that no thing can exist and no fact can obtain without there being an explanation for that thing's existence or for that fact's obtaining' (ibid. 37). Rowe argues convincingly that PSR is untenable in its strongest form (Rowe 1975£>), and I agree. But I also agree with his claim that 'no one has put forth any convincing argument for the falsity of PSR2', this weaker form of PSR: 'Every existing thing has a reason for its existence either in the necessity of its own nature or in the causal efficacy of some other beings' (Rowe 1975a: 261). I subscribe to PSR2, interpreting the expression 'a reason for its existence' in the sense of a reason for its presently existing. Not only the history of science, but even a fundamentally rational attitude towards ordinary reality, presupposes PSR2. And since there is no ordinary existing thing about which we could tolerate the blithe announcement that there simply is no reason for its existence, rationality forbids our abandoning the principle when the existing thing in question is extraordinary or all-pervasive—a thing such as the universe, or matter.

end p.107

It may already be apparent, but it will become clearer, that the form of PSR presupposed in G6 is PSR2. Even at this point it should be clear at least that in G6 Aquinas is assuming that1Every existing thing' that is related indifferently to existing and not existing 1has a reason for its existence ... in the causal efficacy of some other beings.'

It seems to me, then, that argument G6 is acceptable as far as the sentence ending in line 8. In any essentially ordered series of causes invoked to explain the presently continuing existence of any and every dependent being, there must be something that serves as a first, non-instrumental, independently operating cause.

Could that something be the natural laws themselves? Their persistence is a necessary condition common to the existing of all the dependent beings we've been considering, but the persistence of the laws—or, more precisely, of the natural dispositions or governing regularities represented in them—certainly isn't self-explanatory. The necessity that has sometimes been ascribed to them isn't logical necessity, but rather a kind of conditional necessity. Nor do the laws themselves, even sublimated and unified in the Theory Of Everything, or the Final Theory, seem to constitute a plausible candidate for the role of first, non-instrumental, independently operating

cause.

29 A paradigm of the distinction between a necessary condition and the sustaining cause that supplies the condition is (a) nourishment as a necessary condition of life and (b) the source of the nourishment as the cause sustaining life.

Anything that could count as Alpha would, obviously, have to have some

intimate sort of relationship with natural laws, but identity goes too far.

30 Cf. Davies 1983: 45: The God who is outside time is regarded as "creating" the universe in the more powerful sense of "holding it in being at every instant". Instead of God simply starting the universe off (a belief known as deism rather than theism), a timeless God acts at all moments. The remote cosmic creator is thus given a greater sense of immediacy—he is acting here and now—but at the expense of some obscurity, for the idea of God being above time is a subtle one.

The alternative roles of God in time, causing the creation, and a timeless God holding the universe (including time) in being, are sometimes illustrated schematically in the following way. [End-note citing Swinburne 1979: ch. 7, q.v.] Imagine a sequence of events, each one causally dependent on the preceding one. They can be denoted as a series . . . E3, E2, Ei, stretching back in time. Thus, Ei is caused by E2, which in turn is caused by E3 and so on. This causal chain can be denoted as follows:

where the "L's remind us that one event causes the next through the operation of the laws of physics, L.

The concept of a causal God . . . can then be illustrated by making God, denoted G, the first member of this series of causes:

By contrast, if God is outside time, then he cannot belong to this causal chain at all. Instead, he is above the chain, sustaining it at every link:

and this picture could apply equally well whether the chain of causes has a first member (i.e. a beginning in time) or not (as in an infinitely old universe). With this picture in mind, we may say that God is not so much a cause of the universe as an explanation.' See also Braine 1988.

### Pointing to the laws end p.108

counts as indicating part of the answer to the big question—Why is there this sort of world rather than another sort, or nothing at all?—but only the part that has to do with there being this sort of world rather than another, not at all the part that has to do with there being something rather than nothing.