June 18, 2019

This Post is Old!

The post you are reading is years old and may not represent my current views. I started blogging around the time I first began to study philosophy, age 17. In my view, the point of philosophy is to expose our beliefs to rational scrutiny so we can revise them and get better beliefs that are more likely to be true. That's what I've been up to all these years, and this blog has been part of that process. For my latest thoughts, please see the front page.

Pruss and Rasmussen's First Argument from Possible Causes

Pruss and Rasmussen's fourth chapter discusses what the authors variously describe as a "modal cosmological argument" or "argument from possible causes". Although this type of argument has received some discussion in the recent philosophy of religion literature, it is much less well known than the classical argument from contingency discussed in chapter three, and the dialectic of objections and replies is much less well-worn. The idea behind this kind of argument is that since the modal system S5 defended in chapter two validates the inference from possibly there is a necessary being to there is a necessary being, it suffices to show that the pattern of explanation envisioned by the traditional argument from contingency is possible. This can be done using weaker premises than those involved in the traditional argument from contingency.

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

  1. Causal Principle: for any positive state of affairs s that can begin to obtain, it is possible for there to be something external to s that causes s to obtain.

  2. A Possible Beginning: it is possible for there to be a beginning of the positive state of affairs of its being the case that there exist contingent concrete things.

  3. From a Possible Beginning to the Possibility of Necessary Existence: if (1) and (2) are true, then it is possible that there exists a necessary concrete thing.

  4. Therefore, it is possible that there is a necessary concrete thing.

  5. From Possibility to Actuality: if it is possible that there is a necessary concrete thing, then there is a necessary concrete thing.

  6. Therefore, there is a necessary concrete thing. (pp. 69-70)

Given a certain plausible interpretation of 'external' in premise (1), (3) will be a conceptual truth. (5), as just mentioned, is a consequence of the modal system S5 defended in a previous chapter. The premises in need of defense, then, are (1) and (2).

Two technical notes about these premises are in order. In the first place, this is not formulated as an argument about objects beginning to exist, but rather about states of affairs beginning to obtain. The authors make use of this feature of their formulation in answering Objection 4 (pp. 87-89), but as a reader I felt this point was getting lost from time to time, especially since many of the motivating intuitions seem to be about objects rather than states of affairs. It is, however, important that the argument does not rely on a principle about objects, since an external cause of the mereological sum (or whatever) of all of the actual contingent concrete things could be another concrete contingent thing. For instance, the view according to which our universe was 'born' from a 'parent universe' (as in some live models in physical cosmology) might be contingently false, and our physical universe might be all there actually is to contingent concrete reality. At the possible worlds where there's a parent universe, the happenings in that parent universe that cause (a duplicate of) our universe would be contingent, but they would be external to our universe. So even if our physical universe is the totality of concrete contingent reality, it might still be possible for it to have an external cause that is itself contingent. However, anything concrete and contingent would be internal to the state of affairs there being concrete contingent things so that state of affairs could not have an external concrete contingent cause.

Second, the authors are very careful in their definition of "beginning to obtain." Here's their formulation:

A state of affairs s begins to obtain if and only if (i) there is a time at which s obtains, (ii) there is a time or finite interval of time U, such that there is no time prior to U at which s obtains, and (iii) s would not obtain without time. (p. 71)

What is important about this definition is that it is consistent with (but does not require) the notion that time itself has a beginning, and that the state of affairs there being concrete contingent things began to exist when time itself began.

With those preliminaries out of the way, we turn to Pruss and Rasmussen's defense of premises (1) and (2).

The defense of (1) includes three arguments, given on pp. 73-75. The first is an argument from conceivability: it seems conceivable that the beginning of any state of affairs should have a cause. The second is an inductive argument: we've seen a lot of beginnings with causes, and things that actually have causes possibly have causes, so we've seen lots of beginnings that possibly have causes, so maybe they all do. The third is an argument from modal uniformity: there is no relevant difference among states of affairs that can begin that would explain why some would be causable and some not.

The first two arguments are weak. The third is somewhat stronger, though perhaps still unconvincing. The problem with the first argument is that we have to remember that we aren't talking about objects beginning to exist here, we're talking about states of affairs beginning to obtain. However, our intuitions about this are likely based on objects beginning to exist. I'm not sure I have any separate intuition about states of affairs beginning to obtain, but that's what's needed here. And, again, the reason for this is that we need to be able to formulate (1) and (2) in a way that validates (3), and that means we can't be talking about the coming to exist of some object like the mereological sum of the actual concrete contingent things.

The problem with the second argument is that we wouldn't be able to observe counterexamples, so we don't know how many there might be. That is, we can sometimes observe that a state of affairs is caused to begin to obtain, and actuality implies possibility, but we can't ever observe that a state of affairs can't be caused to begin to obtain.

The third argument, as I said, is strongest. The modal uniformity principle (which will come back in a later chapter) seems pretty plausible: things that differ in their modal profiles generally differ relevantly in their categorical properties, but among states of affairs that can begin to obtain there is no relevant difference that would allow some to be possibly caused and others not. However, the principle is again more plausible as applied to objects than states of affairs, and the examples the authors give to motivate it (pp. 74-75) are examples of objects.

I conclude, then, that premise (1) has considerable intuitive plausibility, but the authors' arguments do little to increase its plausibility.

Regarding premise (2), the authors list some reasons why some philosophers are likely to find it plausible, but develop only one argument in detail. This argument begins by asking what maintains Tibbles the cat in existence from moment to moment. I suspect that many non-theists will want to get off the boat right away here, because they may think that the very idea that Tibbles needs to be maintained in existence (if that means something metaphysical, and isn't just a question about how biological organisms maintain homeostasis) imports some theistic metaphysics about God's activity in sustaining the universe. More generally, one might appeal to the notion (ably defended in a classic paper by Grünbaum) that what stands in need of explanation is relative to a theoretical framework, and the idea that Tibbles' continued existence needs explanation relies on an outmoded theoretical framework.

Pruss and Rasmussen offer some response to this objection in their consideration of 'existential inertia' on pp. 77-78. However, the discussion is quite brief and not terribly convincing.

In my opinion, it would have been better for the authors to rely on what they call Reason 1 (p. 75), the argument from physics. Current physics suggests that physical reality as a whole likely began to exist (in Pruss and Rasmussen's sense, which is compatible with the view that time itself is part of physical reality). This is a good reason for thinking that it is metaphysically possible that physical reality began to exist. Further, most opponents would admit that it is metaphysically possible that physical reality should be all there is to contingent concrete reality. (Indeed, most non-theistic analytic philosophers think this is actually the case!) Hence, it is metaphysically possible that the state of affairs there being contingent concrete things began to obtain.

Alternatively, you could argue as follows: there are solutions to the equations of General Relativity that (on the intended physical interpretation of the math) imply that physical reality began to exist. Therefore it is physically possible that physical reality began to exist. But what is physically possible is metaphysically possible. Therefore, it is metaphysically possible that physical reality began to exist. But it's also metaphysically possible that physical reality be the totality of concrete contingent reality. Therefore, it is metaphysically possible that concrete contingent reality began to exist.

Both arguments contain one questionable step, relating to the recombination of possibilities: it is not in general true that if p is possible and q is possible then p&q is possible, but I assume that in this particular case if it is possible that physical reality began to exist and also possible that physical reality is all of contingent concrete reality, then both of these are possible together. Although the general principle is not valid, the particular case looks pretty plausible. The second argument contains a second questionable step: from the claim that a certain solution to the fundamental equations of a current fundamental physical theory exists to the claim that the physical interpretation of that solution describes a physical possibility. Many philosophers of physics and physicists think that some solutions to the GR equations are 'unphysical' (i.e., as the philosophers say, physically impossible). However, I think most would agree that standard Big Bang models represent physical possibilities. So in both of these steps, we could either treat the general rule as allowing prima facie justification, or we could try to formulate a more precise rule that would handle the exceptions.

In any event, I think both (1) and (2) are probably true but not for the reasons Rasmussen and Pruss emphasize. I note that this is not necessarily a shortcoming of the book: part of its approach is to give lots of different arguments coming from lots of different premises. It is not part of the authors' intention that any one philosopher should accept all the arguments or accept them for precisely the reasons given. Rather, the aim is to show that an enormous range of philosophical positions turn out to converge on the common conclusion that there is a necessary being. So the fact that I find the premises plausible but not for the reasons they give could be seen as strengthening their overall case.

I want to conclude with one quick observation about the answers to objections. I found Objection One (pp. 79-82) confusing. This is an objection from essentiality of origins: anything that's caused is necessarily caused. But this principle is extremely implausible as applied to states of affairs, for reasons the authors give in response to Objection 4 (pp. 87-89): states of affairs are general, not particular, so it's certainly not the case that there's only one causal chain that can lead to the obtaining of a given state of affairs. Again, without this assumption the authors' argument doesn't work: unless a strong form of essentiality of origins is true, there's no reason the totality of actual contingent concrete things couldn't have a contingent concrete cause at some other possible world. However, on further reflection, I don't think the authors have made any logical or philosophical error here: the only problem is that a crucial step in the imagined objector's line of reasoning is not made explicit, and I didn't successfully fill it in on the first reading. Here's how I think the objection is supposed to go. Let 'A' be the proper name of any possible particular object. Suppose that the state of affairs A's existing possibly begins to obtain. If A's causal history (or lack thereof) is essential to A, then to say that A's existing is possibly caused to begin to obtain is to say that necessarily, if it obtains at all, it is caused to begin to obtain. This is the objector's basis for thinking that the weak causal principle is not so weak. But once we make this aspect of the objection explicit, then another possible response is open to us: we could restrict the principle to states of affairs containing no rigid designators (or expressible without rigid designators, or whatever we need to say depending on what we think states of affairs are). I'm not sure whether or not this approach is better than the authors' own strategy, which is to say that it's possible that a duplicate of that state of affairs obtain. (Actually, when we spell out the details, the two may turn out to be equivalent.)

Posted by Kenny at June 18, 2019 5:55 PM
TrackBack URL for this entry: https://blog.kennypearce.net/admin/mt-tb.cgi/859

Post a comment

Return to blog.kennypearce.net