blog.kennypearce.net

Pruss and Rasmussen on Modal Uniformity

Pruss and Rasmussen's sixth chapter is entitled "From Modal Uniformity." Based on the general format of the book, one might have expected a new argument for a necessary being from modal uniformity, but that is not exactly what happens in this chapter. Rather, a principle of modal uniformity is offered in support of the possibility premises employed by the previous arguments.

The general idea behind principles of modal uniformity is that certain kinds of differences in propositions look modally irrelevant. That is, we don't expect these differences to lead to a difference in modal status. Pruss and Rasmussen focus on quantitative differences. Thus, for instance, if it's possible that there be exactly 1000 daffodils then certainly it's also possible that there be 999 or 1001. But of course there's nothing special about this narrow number range, so we'd expect just any number to be possible.

The authors again go looking for the weakest/safest plausible principle that will meet the needs of the argument. They arrive at the following:

normally, if a proposition p differs from a proposition q by a mere quantity, and if p and q are (i) logically consistent, (ii) expressible without twin-earthable terms, and (iii) not about necessary things, then p is possibly true iff q is possibly true. (p. 116)

By way of explanation of the principle, Pruss and Rasmussen offer the suggestion that "possibility differs from impossibility in a basic, categorical way, and (ii) basic categorical differences don't easily turn on mere differences of degree" (p. 117). I'm not sure exactly what "basic categorical" means here. Perhaps "basic" is supposed to mean something like a perfectly natural property, or an intrinsic fundamental property, or something, and perhaps "categorical" is just supposed to mean not coming in degrees.

In any event, the principle looks pretty plausible. Furthermore, as a defeasible inference, the examples the authors use don't look too bad.

The first line of thought goes like this. We know that lots of contingent concrete things have a cause. So we have very good reason to believe that it's possible that there be finitely many contingent concrete things and that, say, 60% of them have a cause. But this differs only quantitatively from the propositions that there be finitely many contingent concrete things and all of them have a cause, so the latter should be regarded as possible as well. As a defeasible inference, this looks good to me: an opponent should explain why it's not possible (in a world with finitely many contingent concrete things) for 100% of contingent concrete things to be caused. Of course, many philosophers have tried to do this. A lot of those arguments are answered in chapter 9.

The second line of thought goes like this. It's possible that "there be a caused beginning of there being exactly three contingent concrete things" (p. 121), and so for various numbers. So we should think it's possible for there to be a caused beginning of there being exactly one contingent concrete thing. Again, as a defeasible inference this looks good; anyone who rejects it should explain why.

I found the application of the first conclusion convincing, but not the application of the second.

The first conclusion, we are told, supports the traditional Argument from Contingent Existence, but that's not really right. It's actually being used to support a form of modal cosmological argument. (The traditional Argument from Contingent Existence doesn't have a possibility premise.) But the claim that possibly there are finitely many concrete contingent things and they all have a cause does (given a non-circularity constraint) imply that there is a necessary concrete thing. (Recall that 'concrete' means 'causally capable'.) So far so good.

The second conclusion is supposed to support both of the arguments from possible beginnings, but this, I think, does not work. The reason is simple: if there are more than n concrete contingent things, then you can cause it to begin to be the case that there are exactly n by annihilating some. But this means that there being a caused beginning of there being exactly 1 concrete contingent thing does not require there to be any concrete non-contingent things. Suppose there were two concrete contingent things, and Thing 1 causally annihilated Thing 2. Then Thing 1 would cause it to begin to be the case that there was exactly 1 concrete contingent thing. Note also that this is different from the objection they consider on p. 122 which is that Thing 1 might prevent Thing 2 from coming into existence.

Perhaps Pruss and Rasmussen will object that the state of affairs in question would not be completely or totally caused (so that it doesn't count unless the thing causes both there being at least one thing and there being no more than one thing), but that wasn't in their statement, so a stronger premise would be required. Or perhaps "there being exactly 1 concrete contingent thing" is meant to be interpreted as a tenseless, but then it's hard to figure out how to interpret "beginning to be the case". So I'm not sure how to make this argument work.