blog.kennypearce.net

Pruss and Rasmussen on the Argument from Contingency

Pruss and Rasmussen's third chapter begins the book's main project, the examination of arguments for a necessary being. They describe the argument presented here as "classical," in contrast to the "newer, more sophisticated" arguments they will discuss later (p. 33). The argument they present is indeed pretty similar to versions that would be found in a typical survey of philosophy of religion. However, the discussion of the argument is careful and sophisticated, and it does show how the considerations about modality discussed in chapter 2 can help to improve our understanding of the argument, and in particular to answer some well-known objections associated with Hume and Kant.

Pruss and Rasmussen formulate their version of the argument as follows:

1. For any particular contingent concrete things, there is an explanation of the fact that those things exist.


2. Considering all the contingent concrete things that exist, if there is an explanation of the fact that those things exist, then there is a necessary concrete thing.


3. Therefore, there is a necessary concrete thing. (p. 34)

Regarding the validity of the argument, there are two issues, which the authors note. In the first place, it makes use of plural quantification. There are some tricky technical issues about this device. Second, for the argument to be valid we need to interpret premise 2 as implying that there are at least some concrete contingent things.

In the symbolic formulation of the argument the authors give, premise 2 is formulated in such a way as not to imply this, but on p. 35 the authors do make this assumption explicit.

Although the authors remind us (on p. 34) that 'concrete', in their usage, means 'causally capable', they do not note that this makes their assumption at least somewhat more controversial than it sounds. Philosophers often introduce the notion of concreteness by means of examples: tables, for instance, are concrete, while numbers are not. Now, of course my kitchen table exists, and of course it might not have existed, so of course there is at least one concrete, contingent thing. If you don't buy that, electrons and quarks and things are also concrete, but they didn't have to exist either, and so on.

But note that on Pruss and Rasmussen's definitions, this does not suffice, since a concrete entity is defined as one capable of causation. Hence, skeptics about the notion of causation—e.g., those who think it has been displaced in modern physics—will count as rejecting this assumption, even if they believe in contingently existing electrons. It would also be interesting to think further about how the argument would fare on various forms of anti-realism or reductionism about causation. Proponents of these views accept that some causal claims are true, but deny that they are getting at anything metaphysically 'deep'. So they would say that there are concrete entities (in Pruss and Rasmussen's sense), but they would deny that being concrete (in this sense) has any serious metaphysical significance. Depending on the strength of one's anti-realism, I think some of the arguments in this chapter might be undermined.

In any event, Pruss and Rasmussen expect the reader to grant that assumption without much fuss, which is presumably why they haven't bothered to make it a premise in the explicit formulation of the argument. Following a pattern used throughout the remainder of the book, they proceed to make a positive case for each of their premises, then consider a long list of objections.

I have just two comments to make about the rest of the chapter (for now), both of them to do with the defense of premise 1.

First, on pp. 37-38 (which is officially in the 'positive' section, not the reply to objections), the authors consider an objection which they attribute to Sean Carroll to the effect that they are assuming an objective, metaphysical notion of explanation, when in fact explanations are a subjective feature of how humans relate to the world.

This objection is of interest, I think, because it relates to the issue raised above, about the metaphysics of causation. If there are objective explanations (and I think there are), then it seems to me that they must be based on certain metaphysical relations that exist independently of, but are reported by, the sorts of explanations humans give. The sorts of explanations humans give are linguistic, or perhaps psychological, items, encoding information that removes mystery. (My talk of explanation as 'removal of mystery' is drawn from Pruss's earlier PSR book.) The believer in objective explanation thinks that, at least sometimes, these human explanations report objective "making it the case" relations in the world. So when I say that the earth is warming in part because of greenhouse gas emissions, I'm claiming that the greenhouse gas emissions are involved in making the earth warm, where the relation expressed by 'making' is something completely independent of our explanatory practices. In the example I've given, the relation we have in mind would appear to be causation. But now the whole picture works only if causation is completely independent of our explanatory practices. In other words, it seems that objective explanation requires some kind of anti-Humean metaphysical structure.

I was surprised, therefore, that Pruss and Rasmussen did not respond by defending some kind of anti-Humean picture here. Instead, they argue that standard scientific practice presupposes that "explanation-based reasoning can help us discover features of the objective world" (p. 38). Now, however, we are assuming scientific realism, i.e., the view that standard scientific practices "help us discover features of the objective world." There are scientific realists who are anti-realists or eliminativists about causation, so perhaps we are making some kind of progress. Certainly the points made on p. 38 are well-placed as directed against Carroll, who is definitely committed to the claim that scientific practices, including inferences to the best explanation, can reveal objective facts about matters remote from everyday life. (That's literally what he does all day.) However, some philosophers would want to raise a more radical objection that would be more difficult to answer, that is, a challenge to the strong form of realism the authors seem to presuppose. You can't defend everything at once, but this is a place I think the authors could have been more explicit about their undefended assumptions.

Second, on pp. 40-43, the authors discuss an argument to the effect that if it were possible for contingent concrete things to pop into existence without a cause, we'd probably observe such instances, but we don't. It seemed to me that the authors should have discussed pair-creation of particles in quantum field theory, since this is often described (in popular treatments) in precisely these terms (things popping into existence out of nothing). However, implicit in the book are (at least) two responses to this issue. In the first place, as the authors hint in their brief discussion of Laurence Krauss in the Introduction (pp. 3-4), this kind of talk seems to involve a peculiar use of the word 'nothing'. Surely the quantum vacuum is not nothing in the relevant sense. Second, as the authors note on p. 43, the view that is empirically supported here is, at best, that the laws of physics permit certain things to pop into existence in constrained ways under constrained circumstances. Now, those laws are either necessary or contingent. But assuming those laws are contingent we can just run the same argument again and wonder why laws don't randomly pop in and out of existence.

One view the authors don't discuss is a view on which the laws are necessary, but are not necessary beings in the authors' sense. There are at least two ways of doing this: one is to avoid reifying laws, and the other is to deny that laws are concrete, i.e., causally capable. Both strategies are very plausible for logical laws. The Law of Non-contradiction isn't a thing that exists in every possible world; it's just that every possible world is as the Law says it is. Similarly, the Law of Non-contradiction doesn't act in the world to prevent contradictions (it's not a cause), it's just that (as the Law tells us) no contradictions ever occur or ever could occur. If there are laws for the probabilistic production of concrete things that are similar to the Law of Non-contradiction in at least one of these two ways, then the argument fails. (Parfit's Probabilistic Selectors would do the trick.)

Regarding the chapter as a whole, most of the moves will be familiar to most specialists in philosophy of religion. (In a number of cases, they will be familiar from the authors' own previous work.) Perhaps they will not be so familiar to metaphysicians who are the primary audience here. In any event, the various moves are here put together in a careful and systematic way that will be helpful for both specialists and non-specialists, and the chapter does help to prepare the way for even better things to come. Stay tuned.