On the Infinite Regress Objection

Does it undermine cosmological reasoning?

The infinite regress objection aims to undermine the fundamental claim of cosmological reasoning that there is a primary cause of contingent things that is itself non-contingent. As the putative objector might say, why can’t we explain contingency with further contingency, ad infinitum? This contingent thing is explained by that contingent thing, and so on and so forth. Turtles all the way down, so to speak.

There are two strategies available to the cosmological reasoner when it comes to dealing with the infinite regress objection. The first is to argue that an infinite regress is irrelevant, the second is to argue that an infinite regress is impossible.

So, for example, one might formulate a cosmological argument seeking to explain why there is anything of the type contingent. In this case, the objector claiming to explain contingency through an infinite regress of contingent things is presuming the thing to explained, since any particular contingent thing or infinite string thereof is logically posterior to the fact that there are contingent things. And what presumes the explanatory target cannot explain it. Thus, whether there are finite or infinite contingent things, is irrelevant in this case. What we seek is an explanation of all contingent things (plural), however many and however arranged.

Leibniz took an approach like this with his example of Geometry books, an infinite number of them, with each one copied from the one previous. Leibniz contends that, even if such a scenario were possible, we would obviously still seek an explanation of the entire infinite chain of Geometry books — that is, why there are any books at all — and, moreover, why the subject is Geometry and not biochemistry, or whatever. Frankly, I find this illustration so intuitive to be demonstrative — for me, anyway. (Naturally, your mileage may vary).

What about the other option? That is, to argue that an infinite regress is impossible. There are two ways of going about this. First is to argue, in general, that any effect cannot be preceded by an infinite causal chain. That is the thesis of causal finitism, proposed by thinkers like Alexander Pruss. Obviously, if this bolder thesis can be sustained, then just about any (stage 1) cosmological argument can be put through, including the Kalam. However, the more modest — and to my mind, more certain — approach is to suggest only certain types of causal series cannot regress infinitely and thus are necessarily terminating. Famously, this is where Aquinas argues that causal series per se cannot regress infinitely, whereas causal series per accidens can. The basic idea is that some causal series display the feature of derived or borrowed causal power, which is to say, certain members in the causal chain do not have the causal property under consideration in virtue of themselves, but rather are borrowing or deriving or transmitting that causal property from something else. Stock example: suppose a lamp is powered and plugged into an extension chord. Of course, we know that an extension cord only transmits electricity, it has no inherent capacity to produce it. Would an infinite number of extension cords answer the question of how the lamp is powered? No. Would plugging an extension cord into itself solve the problem? No. What we need in this situation is something that actually has the power to produce electricity, otherwise the effect would not occur. That’s the idea.

Anyway, the infinite regress objection is considered at further length in my forthcoming book, The Best Argument for God.

Related Posts