Many Universes and Possibilism
May 9, 2008 — 17:12

Author: Andrew Moon  Category: Atheism & Agnosticism Existence of God  Comments: 9

So there’s the theistic explanation for the fine-tuning of the universe and the many-universes atheistic explanation. One of the criticisms that theists have made of the many-universes atheistic explanation is that there is no independent evidence that such universes exist. I was wondering if anybody’s put forth the idea that all of David Lewis’ reasons for believing in concrete possible worlds are reasons to believe in many universes. With Lewis’ possibilism in place, we have a response to this criticism. (I’m sure that’s not the only response; anyway, this is the response I’m interested in.) Does anybody know if anyone has written on the connection between Lewis’ possible worlds and the many-universes hypothesis?

Comments:
  • Hey Andrew,
    One issue is Lewis’s requirement of causal isolation between worlds. I don’t think it’s essential to the many-worlds hypothesis that there are causal relations between various worlds but some (most?) formulations of the many-worlds hypothesis do posit causal relations between the various worlds. If the goal is to give empirical evidence for many worlds then Lewis’s possibilism won’t help.

    May 10, 2008 — 9:03
  • Andrew,
    The Lewisian model isn’t a multiverse, so it cannot be used in this way. In a multiverse you have a single world composed of many spatiotemporally isolated (island) universes. Given the way Lewis individuates worlds, such multiverses are not possible. Each spatiotemporally isolated universe is a distinct world. If you connect those universes in time or space (say, having one epoch after another) then a sort of quasi-multiverse is possible for him. In any case, I’m surprised you say this: “one of the criticisms that theists have made of the many-universes atheistic explanation is that there is no independent evidence that such universes exist”. Isn’t there? The many worlds interpretation of quantum theory has been around for a while and there are lots of many-world cosmologies around (“bubble theories” an such), so it seems to be doing some theoretical work.

    May 10, 2008 — 10:34
  • Andrew Moon

    Mike,
    Thanks, and you’re right that there is such evidence. I think I heard W.L. Craig say something like that, but now that I think about it, I think he wasn’t saying that there wasn’t any evidence, but that such evidence is very speculative (I’m thinking of the vacuum fluctuation models). I’ve been moved by Robbin Collins’ argument that vacuum fluctuation models would themselves have the same fine-tuning problem. But I have to admit that I often feel like I’m not really grasping what’s going on in these discussions (probably because I don’t know enough physics).
    Mike and Ted,
    Maybe I’m missing something here, but couldn’t all of the work that many universes are supposed to do in ‘many-universe explanations’ of the fine tuning be done by Lewis’ possible worlds? What difference does it make that they’re spatiotemporally isolated? I don’t see why that matters. (Actually, Ted, I think that’s what you were saying, right?)
    Couldn’t a possibilist think of things this way: there are an infinite number of concrete, spatio-temporally isolated worlds, and in some of these worlds, there is life. In a small portion of those worlds, there would be parameters of physics that permit life. Hence, it is not unlikely that a universe such as ours, that supports life, exists.
    Ted, I agree that Lewis’ reasons don’t count as empirical evidence, but they still count as reasons, which is what we should care about. Who cares whether its empirical or not. (Not that you were saying otherwise.)

    May 10, 2008 — 13:29
  • Mike Almeida

    Maybe I’m missing something here, but couldn’t all of the work that many universes are supposed to do in ‘many-universe explanations’ of the fine tuning be done by Lewis’ possible worlds?
    Andrew, this is a different question altogether. Here you don’t need any talk of multiverses. You can deny that there are any (as Lewis effectively does) and urge that there is no surprise that a world exists in which there are living things, since every possiblity exists. You can argue further that it is not surprising (or more exactly unlikely) that such a world is actual, since that amounts to being surprised at their being living things at the worlds that admit living things. I’ve actually given an argument on similar grounds for not not being surprised that evil exists or that this is not the best possible world (in ‘Theistic Modal Realism?’ in chp. 8, Metaphysics of Perfect Beings). I’ll remark that noting the shamelessness would be otiose and unmitigating. So presumably is the meta-noting. . .

    May 10, 2008 — 14:30
  • James Beebe

    Mike,
    I don’t think it’s correct to say that the Lewisian model of possible worlds isn’t a multiverse. Granted, the claim that they constitute a multiverse isn’t compatible with Lewis’s stipulation that spatiotemporal contiguity is the criterion for individuating worlds/universes. But this criterion isn’t sacrosanct. To insist upon using his criterion seems to amount simply to insist upon using the terms ‘universe,’ ‘world’ or ‘multiverse’ in a certain way.
    But Andrew is right that, according to common ways of understanding multiverses, Lewis’s model should count as a multiverse. Even if, as Ted notes, many multiverse models hypothesize spatiotemporal continuity and causal relations between universes, there are models that do not.
    James

    May 11, 2008 — 19:38
  • Andrew Moon

    Mike,
    Thanks for the references (and I don’t mind at all that they’re yours!). You also said that Lewis (effectively) denies that there are multiverses? Where are you referring to? (Also, to anybody out there, has there been any literature specifically relating possibilism to the fine-tuning discussions? I’m interested in writing a paper on this, so that’s part of why I’m asking.)
    Also, it might help me if you define ‘multiverse’. Is that a term of art? Or is that a term that physicists use in some specific way? Or is it a term that phil. religion people are using in some specific way? (Also, thanks James, for the comment.)

    May 11, 2008 — 20:41
  • Lewis’s stipulation that spatiotemporal contiguity is the criterion for individuating worlds/universes. But this criterion isn’t sacrosanct.
    I certainly didn’t intend to suggest that it is sacrosanct. It does not matter to me how anyone decides to use the world ‘multiverse’. The discussion began with the assumption that we were working within a Lewsian view. So I made my comments within a Lewisian view. That was the idea. My last suggestion was that you need not move beyond those assumptions to accomodate Andrew’s intuitions on how such a modal metaphysics might be put to use. So the whole multiverse question is moot, given his purposes.
    But you add,
    But Andrew is right that, according to common ways of understanding multiverses, Lewis’s model should count as a multiverse.
    I just don’t think there is a common way to use the word ‘multiverse’. Or else, I’m not aware of any common use.
    You say,
    To insist upon using his criterion seems to amount simply to insist upon using the terms ‘universe,’ ‘world’ or ‘multiverse’ in a certain way.
    I’m not insisting on it; I’m playing by what I thought were the ground rules. In any case, it’s not a suggestion for linguistic usage. It’s a metaphysical view. In the absence of the Lewisian criterion, we need an alternative criterion for individuating worlds (I say this a few notes above as well). There are other ways to do it, sure, but it’s tricky to do, and it takes some metaphysical work (I have in mind Kris McDaniel’s, ‘Modal Realism with Overlap’, and Philip Bricker, ‘Island Universes and the Analysis of Modality’).

    May 11, 2008 — 21:04
  • You also said that Lewis (effectively) denies that there are multiverses? Where are you referring to?
    He discusses it in On the Plurality of Worlds, but I don’t have a copy handy, so I can’t offer a page reference. I take a multiverse to be a single world composed of many distinct universes. Typically those universes are island universes–spatiotemporally isolated from one another. There are presumably other possibilities that will serve other purposes.

    May 12, 2008 — 12:39
  • This was an interesting discussion. I think that we might view the relationship between the multiverses of theoretical physicists and the metaphysical modal realism of philosophers in this way: the multiverses are an expansion of what we had viewed as nomological possibilities (with varied physical laws, constants, dimensionality, etc.). If we think about expanding this project toward its limit we approach the full expanse of metaphysical possibility discussed by philosophers.

    May 15, 2008 — 11:44