New APPS

35 responses to “Williamson on David Lewis”

1. 

   Sara L. Uckelman

   June 10, 2014

   If you’ll be reading Lewis, I’d like to commend an excellent little piece written by a previous colleague of mine, Janine Reinert, in Analysis 73, “Ontological Omniscience in Lewisian Modal Realism”, http://analysis.oxfordjournals.org/content/73/4/676.full.pdf+html?sid=d45388da-61b0-46fe-83b9-09b4c55980b3

   LikeLike

   Reply
2. 

   Jon Cogburn

   June 10, 2014

   Awesome. Thanks. I look forward to reading it.

   LikeLike

   Reply
3. 

   Michael Kremer

   June 10, 2014

   First, a little correction: the title of Lewis’s book is On the Plurality of Worlds [Oops, thanks. Fixed it. – Jon]
   Now for Williamson. I take the point to lie in the phrase “a theory of total spatiotemporal reality”. As I understand Lewis, he takes his alternative possible worlds to be universes which are as real (“concrete”, not “abstract”) as the one we inhabit; the actuality of our universe is merely indexical (like the here-ness of the place I happen to be writing this note, my apartment). So the total plurality of worlds is the total spatiotemporal reality. This means that the total spatiotemporal reality includes regions that are physically impossible (even if we allow a physical multiverse in which various physical constants have different values in different universes). This is supported on metaphysical grounds. I takt this to be what Williamson thinks physicists would deny.
   One way this argument might go is this: the very concept of space-time is tied up with a certain amount of physics, conceptually, so the idea of a total spatio-temporal reality divorced from all that we know of physics is incoherent. Or something like that. Lewis has to have an idea of “reality” which is extra-physical, to distinguish his theory from the “ersatzism” of the Plantingas and Stalnakers who think that possible worlds are merely abstract (representations of) ways things could have been, while at the same time claiming that there are physically impossible worlds. But it’s illegitimate to employ notions of spacetime, causality, etc, to individuate such “worlds” when they are cut off from physics.
   Anyway, though I have read Plurality I haven’t read Williamson’s book, yet, so I am speculating here…

   LikeLike

   Reply
4. 

   Chris

   June 10, 2014

   I think Williamson’s point is that when physicists make general claims they are supposed to be absolutely general. If special relativity holds, it holds in all possible worlds, not just the “nomologically possible” ones. Lewisian metaphysics requires us to give up the universality of physical laws, and insofar as it really true that physicists see themselves as finding the cosmological truths about any possible reality, which I think it is, the theories are in competition.

   LikeLike

   Reply
5. 

   Ross Cameron

   June 10, 2014

   I suspect that lurking behind this criticism of Lewis is the thought that, despite his protestations to the contrary, what Lewis is really doing is saying that actuality is much bigger than we ordinarily think. Since physicists are proposing theories about how spacetime and its contents actually behave – and not theories merely about how the part we know about actually behaves – any bit of spacetime must be governed by the laws physicists propose, or those physical theories are thereby false.
   This objection – that Lewis is merely postulating a bigger actuality – is wrong, and therefore so is any objection that relies on it. But I suspect that’s what’s going on (albeit implicitly).

   LikeLike

   Reply
6. 

   Jon Cogburn

   June 10, 2014

   This makes sense, but it still seems to me to be an awful argument. Why should it then be the case then that “On such matters, physicists may be felt to speak with more authority than metaphysicians”?
   As far as I understand (“Help me David Wallace” repeated in Princess Leia voice), there is debate within physics both about: (1) the extent to which our laws of nature are local or not with respect to this universe, and (2) how the set of physically possible universes should be constrained. When you add the necessary use of idealizations, and the necessity of making sense of the approximate truth of past theories (put that how you like) it’s daft to not at least think that physical impossibilia are at the very least epistemically possible.
   But then the understanding of epistemic versus metaphysical possibility is a philosophical debate. Why should a physicist care about this debate?
   I don’t know if this is relevant, but I would bet if you explained to a physicist the Churchill/Shoemaker view that all metaphysical possibility is physical possibility versus more robust views (with the understanding that the demarcation of the physically possible worlds is the purview of the physicist) most physicists would say that the issue is irrelevant to their concerns. But even if Williamson is right and they wouldn’t say that, it’s still not at all clear to me that they shouldn’t.

   LikeLike

   Reply
7. 

   Jon Cogburn

   June 10, 2014

   This makes sense. Lewis himself discusses rigid and non-rigid uses of “space” and “time” in exactly this context. But what he says strikes me as imminently sensible. Physicists have to be interpreted as using the terms non-rigidly if we are to make sense of their practice. But if that’s the case (and if Williamson’s enthymeme is filled out the way you speculate), then Lewis could retort that Williamson is the one who is not appropriately deferential to the practice of scientists.

   LikeLike

   Reply
8. 

   Jon Cogburn

   June 10, 2014

   Ooh, that’s very interesting and helpful.
   It’s hard to see how the debate proceeds here. Lewis says that for him “actually” has an indexical role, so of course he’s not postulating a bigger actuality when taking other possible worlds to be real. His opponent who doesn’t want to use the word that way then has naturalistic reasons to not be committed (whatever that comes to) to physical impossibilia.
   I’m always tempted to make a distinction between what’s actually possible and what’s merely possible. What’s actually possible in some sense supervenes/is grounded on/etc. what’s actual [something inherently modal like causation (cf. Eric Hiddleston on counterfactuals) or, more fancifully something along the lines of Schopenhaurian “will”]. Since I’m not a naturalist, such a grounding thesis doesn’t preclude physical impossibilia from being actually possible. But a naturalist would make Williamson’s claim about what’s actually possible at least.
   In any case, it will be really interesting to read Williamson’s book with this issue in mind. So thanks tons for bringing it up.

   LikeLike

   Reply
9. 

   Justin

   June 10, 2014

   I agree with Ross’s suspicion but would add to it. This objection seems to me to just rely on some sort of question-begging assumption about the extent of actuality (as Ross mentioned), or about what domain physical theory is meant to characterize.
   Either the idea is that physics is an account of actuality, and actuality encompasses whatever there is (full stop), or the idea is that physics is an account of whatever there is (full stop), even if whatever there is outstrips what is actual.
   Lewis would deny both of these claims. The former, he thinks, embodies an incorrect account of actuality,* the latter embodies a confused conception of what physicists are really up to.**
   *See the actuality chapter in OPW.
   **For starters, Lewis thinks it is a priori that there are possibilities that violate the laws of physics; so it would hardly make sense for him to think of physicists as trying to describe such worlds. He also thinks that facts about what worlds there are are necessary. (See “How Can We Know?” in OPW.) But as he says in New Work, “physics is contingent and is known a posteriori.” He might also argue that physicists surely aren’t trying to characterize any part of reality — if there is such — that is causally or spatiotemporally isolated from us observers. (I can’t think of a place off the top of my head where he makes quite that point, though it’s an obvious one for him to make.) In general, a conception of physics that represents physicists as out to describe all of what there is will be one based on substantive philosophical assumptions with which Lewis will undoubtedly differ.

   LikeLike

   Reply
10. 

    Justin

    June 10, 2014

    One thing I should have added: Williamson might think that physics is supposed to be an account of whatever is spatiotemporal, even if this outstrips what is actual. (So physics wouldn’t have to account for non-spatiotemporal entities, if there are any.) But of course, this would rely on the same sort of question-begging assumptions as would the claim that physics is an account of whatever there is (full stop).

    LikeLike

    Reply
11. 

    Mike

    June 10, 2014

    This seems right. But it might be worth noting that Williamson exchanges the actualism-possibilism distinction for the necessitarianism-contingentism distinction. And there is no restriction of quantifiers to actually existing things (he doesn’t use the notion of ‘actuality’ at all). So he probably has Lewis committed to a view that he (DKL) would prefer not to hold. DKL has a view about what exists period (without the actually exists/possibly exists distinction), on this account, and the physicists have something to say about that.

    LikeLike

    Reply
12. 

    Jon Cogburn

    June 10, 2014

    I’ve just started the book.
    Is allowing quantifiers to range over merely possibly existing things the way he defends the Barcan formula? That seems pretty plausible to me, but I worry you would still be able to generate a problematic Barcan sentece for “actually existing”, i.e. It is possible that there is an actually existing unicorn, therefore there is an actually existing something such that it is possible that it is a unicorn.
    Again, I haven’t read it yet, so I realize I might be way off.

    LikeLike

    Reply
13. 

    David Wallace

    June 10, 2014

    I haven’t read the book, but I had a conversation about this with Tim a couple of months ago. I think he has something relatively specific in mind: that physicists, at present at least, are in the business of constructing theories (“multiverse” is the usual jargon term) in which many spatially and temporally disconnected chunks of reality coexist. So physics conceives of its task as to describe a physical reality consisting (perhaps) of lots of different disconnected parts. So is Lewis. So Lewis and the physicists are engaged in the same activity and their answers are in conflict.

    LikeLike

    Reply
14. 

    Mike

    June 10, 2014

    You can defend Barcan (both directions) assuming actualism, as Zalta & Linsky do, so you don’t need the somewhat wilder claims of Williamson to get those. They are part of the Simplest Quantified Modal Logic, as they call it. But they’re also part of Hughes & Cresswell’s introductory QML; that’s not often mentioned. All you need, one way or another, is a single domain of objects for all worlds. I do think Zalta is committed to there existing “non-concrete” objects that are possibly unicorns, though, if that worries you. I’m assuming here that unicorns are possible. For Zalta & Linsky, a contingent object is one that is concrete in some worlds and not in others, which is also not quite standard.
    Incidentally, on your previous comment, there are lots of people who distinguish between what’s possible and what’s possibly actual. Plantinga doesn’t, but Adams and Fine do, and so do most everyone else who think there are true singular propositions of the sort, ‘possibly, Adams never existed’. I’m also inclined to distinguish what’s possible and what’s actualizable (i.e., what can be brought about). I have no argument but hope its true that what is possible just turns out to be what is actualizable.

    LikeLike

    Reply
15. 

    H. Teichman

    June 10, 2014

    Is anyone aware of any literature that attempts to make a tight conceptual connection between metaphysical possibility and (more or less Aristotelian) practical reasoning? E.g. when ‘deliberating’ I may say to myself ‘Now if I had done X instead of Y, then Z would have happened. So here and now, if I do X1, Z1 will probably result. Ergo: I’ll do X1.’
    And, whenever I hear contemporary modal world-talk, I wonder whether anyone ever took seriously Kripke’s remark in N&N that possible worlds are ‘stipulated’, not discovered.

    LikeLike

    Reply
16. 

    Eric Winsberg

    June 10, 2014

    That’s how I read the free-standing paragraph. But that also strikes me as in implausible claim. You can think of physics, I would say, as consisting in three parts: a conception of what the mathematically describable states of the world are, what the actual world is like, and what other of the mathematically describable states are physically (nomically) possible. The normal reading of all three of these claims is that they all describe empirically contingent facts. The world might have been Newtonian and so the answer to the first question might have been different that what we now expect it to be. Or Parmenides might have been right and a single point might have done the job. Physicists are not even in the business of dreaming up a conceptual or mathematical structure for representing the co-existence of those sorts of possibilities, let alone discussing what their ontological status is.

    LikeLike

    Reply
17. 

    Mike

    June 10, 2014

    Except that Lewis regards spatio-temporally isolated universes as non-actual. Presumably the physicists regard them as actual (if any actually think that there are genuinely causally isolated universes). So, they’re talking past one another, if they’re supposed to be arguing about what ‘the world’ is like, rather than what it could be like.

    LikeLike

    Reply
18. 

    Justin

    June 10, 2014

    Here’s something I didn’t recognize originally but that David Wallace’s comment brought to my mind:
    It doesn’t seem open to a Lewisian to interpret a physical theory of the kind Wallace mentions as only entailing that there actually are spatiotemporally isolated universes. On that sort of interpretation, such a theory would come out false in any possible context on Lewis’s view, since he maintains that no world has spatiotemporally disconnected parts. Surely that’s a substantive worry. Lewis ought to at least be able to say that in some context in some possible world the theory is true. If he can’t say this, it seems he has to say something in the ballpark of “physical theories that posit multiverses are necessarily and a priori false,” which would constitute a pretty implausible challenge to physics.
    Considerations like this, I think, help motivate Phil Bricker’s version of modal realism, which accepts worlds with spatiotemporally isolated parts.

    LikeLike

    Reply
19. 

    Jon Cogburn

    June 10, 2014

    But then everything hinges on what the interlocutors mean by “actual.” If a physicist wants to say that spatio-temporally isolated universes are actual, then one asks whether actual is a blanket term for everything or whether it is indexical in Lewis’ sense. If the former sense, then the view isn’t inconsistent with a trivial reformulation of Lewis’, as it’s just saying that spatio-temporally isolated universes are part of everything (part of Lewis’ point).
    On the other hand, if the physicist wants to use the term in an indexical sense, contrasting the isolated bits of the actual universe with possible, non-actual universes, you get something like Bricker’s view that Justin talks about in 18. But I think that Williamson takes his criticism to be against such a view too, or else why point out that the Genuine Realist’s worlds following different laws than the physicists?
    I do think that Lewis would say something along these lines. His response to Skyrms’ regress in footnote 1 of Chapter 2 of On the Plurality of Worlds with respect to the term ‘reality’ is very similar, and the three premises in the footnote are analogs to the three premises in the text about ‘actuality.’

    LikeLike

    Reply
20. 

    Justin

    June 10, 2014

    It may be—as seems likely—that during the formulation of the physical theory the distinction between actual and possible existence never arose. It might also be that physicists would shrug at such an arcane “philosophical” distinction, or that their opinions wouldn’t seem to require unquestioning deference (pace Williamson). In such a case, it would seem like whether it is better for the Lewisian to interpret the theory as characterizing all the universes there are (full stop) or only some of them (=the actual ones) would depend on details about the theory.
    The former sort of interpretation wouldn’t be appropriate if the theory is such that the existence of any (what a Lewisian would consider) merely possible universes would either (1) entail, in conjunction with the theory, known falsehoods/absurdities about the observable universe(s?) or (2) be incompatible with the theory’s truth. If the theory had either of these characteristics, it would seem that the only interpretation that would make sense for the Lewisian would be one taking the theory to characterize actuality (understood as a restriction on what there is). But then I think you run into the problem I suggested above. So as far as I can tell, it would be a kind of dilemma.
    Lewis has principled reasons for rejecting Bricker’s view; the two views are definitely incompatible. So it seems to me that at least if the physical theory does have the features just mentioned (and don’t ask me!), it would be right to say that the physical theory is in conflict with Lewis’s MR.

    LikeLike

    Reply
21. 

    Justin

    June 10, 2014

    Also, now it seems to me, even if the existence of merely possible universes wouldn’t screw up the theory’s predictions or be incompatible with the theory itself, interpreting the theory as making some claim about the space of all worlds still seems potentially problematic.
    In OPW, Lewis says claims about the what worlds there are, e.g., that there is a world where a donkey talks, are necessary if true and impossible if false. So, presumably, taking the physical theory to characterize the space of all worlds, i.e., as saying “there are worlds that satisfy such and such criteria” where this is compatible with there being additional worlds that don’t satisfy the criteria, would mean interpreting it as being either necessary or impossible.

    LikeLike

    Reply
22. 

    Charles Pigden

    June 11, 2014

    ‘The effect of Lewis’s influential and ingenious system-building was to keep centre stage a view that imposed Quine’s puritan standards on modality long after Quine’s own eliminativist application of those standards have been marginalized (Williamson 2013, xii)
    Surely there is nothing particularly problematic about Williamson’s remark. Lewis is in a sense an eliminativist about modality in that claims about what is necessary or possible are to be analyzed by quantifying one non-actual entities in other possible worlds (which are carefully specified in the reducing theory so as not to rely on the concept of possibility). Lewis retains a modality-free Quinean ideology but inflates his ontology in order to make sense of putatively modal facts. It seems to me that an infinity of concrete possible worlds is an absurdly high price to pay for the dubious benefit of doing away with modal primitives, but that was David’s motivation.

    LikeLike

    Reply
23. 

    Jon Cogburn

    June 11, 2014

    Ooh thanks. I think that’s clearly what he meant now.
    While I’m with Williamson on this I still think it’s a little bit unfair, because Quine’s puritanism is pretty badly motivated (taste for desert landscapes plus nonsense about canonical notation plus not listening to Ruth Barcan Marcus).
    On the other hand, Lewis himself cites Hume in the relevant contexts, saying that he thinks Hume’s arguments about causality (as many people interpret them) apply to necessity itself. Then his Genuine Realism plus counterfactual analysis of causality gives him an analysis of causality consistent with (how many people interpret) Hume’s phenomenology.
    This being said, I think Williamson’s analysis of what’s gone wrong (as you present it) is largely correct, and also gets right the important motivation for some variety of actualism. On the other hand, I think Lewis is correct in arguing that extant actualists are always in danger of presenting the actual and fictional worlds as too text-like.
    What’s needed is an account of how a non Lewisian view can explain: (1) how the actually possible is in some sense at least grounded in the actual (in a way independent of commitments to naturalism), (2) how the epistemicly and merely possible admit Lewisian alien properties, without (3) treating possible worlds in too linguaform a matter as standard actualists do, while (4) making sense of modal logic.
    Again, I’m significantly behind on this literature now. If anyone reading this knows of people who try to fulfill these desiderata (and I know there’s recent stuff on modal realism without possible worlds of any sort, even the type that actualists put forward), I’d much appreciate it. I think Eric Hiddleston’s piece about causality grounding counterfactuals is probably a good place to start, and some of the German Idealists and maybe Deleuze believed something that could start to justify (1)-(3) (world as will, etc. coming out of deep thinking about problems with Kant’s response to Hume).

    LikeLike

    Reply
24. 

    Patrick

    June 11, 2014

    You might be interested in Lance and White’s “Stereoscopic Vision.” It’s not strictly speaking about metaphysical possibility, but rather about persons and freedom, but the topics are not unrelated, and they try to draw very tight conceptual links indeed between the sort of reasoning you discuss and freedom, at least.

    LikeLike

    Reply
25. 

    Charles Pigden

    June 12, 2014

    I have no positive solutions to offer, though I think that the right way to go is to accept that reality is deep down modal. This suggests an eliminativist program. Rather than explaining necessity and possibility in terms of possible worlds you explain possible worlds talk in terms of the primitive modalities.
    However you may be interested in an article that I coauthored with Rebecca Entwisle (an ex-student of David’s) ‘Spread Worlds, Plentitude and Modal Realism: a Problem for David Lewis’ in which we argue
    1) that David’s system is threatened by paradox,
    and
    2) that David himself can only avoid these paradoxes by either a) accepting modal primitives (in which case his system loses its raison d’être) or b) by mimicking the sleazy set theorist who deals with Russell’s paradox by denying that there’s a set corresponding to EVERY condition whilst prudently refusing to say anything systematic about the exceptions.
    Incidentally our paradox-generating device exploits the fact that S5 is implausibly strong (as indeed it is). But if you reject S5 and go for a weaker modal logic, you are stuck with accessibility relations which are, to put it mildly, difficult to make metaphysical sense of. (Hughes and Cresswell resort to parlour-games)
    I have classes to prepare so I must stop.

    LikeLike

    Reply
26. 

    Jon Cogburn

    June 12, 2014

    Charles,
    This sounds fantastic. I’ll read your paper soon. My intuitions are already with your conclusion, and what you describe sounds like an important part of getting clear about what is being affirmed when we affirm that reality is deep down modal. German Idealists, process philosophers, and Deleuzians have been making equivalent claims for a long time and it’s really exciting to me that people with decent facility with modal logic are picking up the charge.
    Does your and Entwisle’s paradox relate to Kaplan’s paradox (and Russell’s version for actualists)? I haven’t thought about how weaker modal logics might block these kinds of paradoxes. Most people gesture at the sleazy set theorist with respect to them (yuck). Do you have the citation for the Hughes and Cresswell?
    Bueno, Menzel, and Zalta have an interesting take which is on-line at http://mally.stanford.edu/Papers/paradox.pdf . I’ve spent some time with the paper, but given my not stellar logic skills it’s the kind of thing that would take me a couple of weeks of dedicated work to digest, and I haven’t had that time yet. But I think it might be interesting to connect up what they are doing with what you and Entwisle have done.

    LikeLike

    Reply
27. 

    Jon Cogburn

    June 12, 2014

    Patrick,
    The Lance/White paper has been very near the top of my reading list for the last two months. Greg Restall has attempted to formalize it in Sequent Calculus at http://consequently.org/papers/cfss2dml.pdf . Unfortunately it would take me a few weeks to get optimally clear about what’s going on. But I understand it enough to think that accommodating Lance/White/Restall is about as important as figuring out something sensible to say about the Barcan formulas.
    There’s also going to be a discussion of two dimensionalism in the proof theory textbook that Restall’s writing, which I think will be more accessible (loosely speaking) than the paper. He’s putting successive drafts of the book at http://consequently.org/papers/ptp.pdf .
    Restall also has a work in progress (http://consequently.org/papers/invention.pdf ) that’s central to a lot of the comments raised above. He explains the genesis of necessity and possibility without taking either as primitive or starting with possible worlds. It’s great stuff.

    LikeLike

    Reply
28. 

    Jon Cogburn

    June 12, 2014

    Mike,
    I just wanted to thank you for comment 14. That’s tremendously helpful. Do you have a citation for Adams and Fine on this stuff? There’s so much literature it’s an uphill battle to get competent.
    My first exposure to modal logic was in linguistics, where formal semantics types tend to use unified domains and use an existence predicate separate from the existential quantifier. There are lots of reasons they do this. Among others, recursively hooking up a (lambdified) tense logic with a fragment of natural language is vastly easier if you do.
    But when you have a unified domain across all temporal and world indexes (as noted above), and don’t take the existential quantifier as an existence predicate, the Barcan formulas don’t seem to be a big deal. I hope this is plausible without assuming actualism and without getting too wild.
    I don’t know if the distinction between Williamson and the linguists is more than merely terminological though. Friends of mine who’ve finished the book told me that Williamson takes the existential quantifier as an existence predicate. But if he’s willing to countenance Richard Nixon as currently existing as a possibilia and the same entity existing in the past as an actual being, I don’t see any substantive difference (from the perspective of doing semantics at least) between that and the linguist who says the same object no longer exists but is eternally in the domain of quantification at every world and time index. I mean, he’d wrongly predict that a competent speaker would say “Richard Nixon still exists.” But that’s not really a linguistically relevant distributional judgment, so it’s fair game for the philosopher to say that the competent speaker is mistaken about that.
    I’ll wait until I’ve finished the Williamson to inveigh about this any more. I won’t know if that’s exactly what Williamson has in mind for another few weeks. I think that the oddness of his views about vagueness might have led me to think he was saying something more controversial when talking with my friend last semester.
    As far as the unified domain plus existence predicate thing, what’s weird is that sixteen years ago David Dowty told me that Dana Scott convinced the formal semantics community in linguistics that this was kosher, and that it ended up being really helpful. But linguists read Cresswell (if I’m remembering right, his structured propositions stuff is on the map) so maybe it comes from him?

    LikeLike

    Reply
29. 

    H. Teichman

    June 12, 2014

    Much appreciated. I’ll look this up.

    LikeLike

    Reply
30. 

    Mike

    June 12, 2014

    You don’t need Lewis’s ontology to get the reduction, fwiw. Sider gets the reduction too assuming ersatz worlds in his “The Ersatz Pluriverse” Journal of Philosophy (2002). Maybe the larger question is why one would want the modal reduction in any case. Can’t be that the naturalistic worldview does not easily include such odd properties: there are indispensible dispositional properties in science, etc, for example. It might be just the theoretically appealing feature of doing more with less.

    LikeLike

    Reply
31. 

    Mike

    June 12, 2014

    Lewis is actually of two minds on the matter. In OPW, he says that he see advantages to including worlds that are multiverses of the sort Bricker describes, and he sounds like he could be persuaded to do so. But he settles for things that are similar to Bricker worlds. He gives four of five examples of worlds that are possible, on his account, that look quite a bit like Bricker worlds, and that we (i.e., we who think there are Bricker worlds) might be confusing for Bricker worlds. See OPW, 72 ff., Isolation. On the other hand, he also says that he takes his use of the word ‘actual’ to be a mere matter of meaning, where ‘actual’ means, in his utterances, ‘this-worldly’. So physical theories that posit Bricker-style multiverses are not a priori impossible, they just would not properly be called ‘worlds’.

    LikeLike

    Reply
32. 

    Mike

    June 12, 2014

    Hi Jon,
    The Adam’s, Fine (and Prior) papers include the ones that follow. But Chris Menzel is excellent and really clear on this stuff. So, you might want to take a look at his work. If you have trouble finding the Prior/Fine & Prior stuff, I have copies of most of it. Happy to send it along.
    Adams, Robert Merrihew: 1979, ‘Primitive Thisness and Primitive Identity,’ The
    Journal of Philosophy 76, 5-26.
    Adams, R. (1981): ‘Actualism and Thisness’, Synthese 49, 3–41.
    Fine, Kit: 1977, ‘Postscript’ in A. N. Prior and Kit Fine, Worlds, Times and Selves,
    —-“Plantinga on the Reduction of Possibilist Discourse,” in J.E.Tomberlin and P. van Inwagen (eds.), Alvin Plantinga (Dordrecht: Reidel, 1985)
    —–Modality and Tense—Philosophical Papers (Oxford University Press, 2005)
    —-“Necessity and non-existence,” in Fine (2005)
    —-“Tense and Reality,” in Fine (2005) University of Massachusetts Press, Amherst.
    Prior, A. N.: 1957, Time and Modality, Clarendon Press, Oxford.
    Prior, A. N.: 1960, ‘Identifiable Individuals,’ Review of Metaphysics 13, 684-696.

    LikeLike

    Reply
33. 

    Justin

    June 12, 2014

    In a relevant passage in the “Isolation” chapter of OPW, Lewis considers an argument that goes: (1) a world might consist of disconnected spacetimes, (2) any way a world might be is a way some world is, (C) worlds aren’t demarcated by spatiotemporal unity. In response to the argument, Lewis rejects “the alleged possibility of disconnected spacetimes within a single world.” (p. 72) He says this claim seems to him to be “no central part of our modal thinking, and not a consequence of any interesting general principle about what is possible.” (p. 71) Granted, he says he’s “reluctant” to do this, and he mentions ways of doing “passable substitute” Bricker worlds using his spatiotemporally unified worlds, but there’s no unclarity as to whether Lewis’s view is incompatible with Bricker’s. (Forgive me if you weren’t suggesting that there was such an unclarity — I couldn’t tell. In any case, all I’ve said is that Lewis’s view and Bricker’s are in definite disagreement, which I take it, they are.)
    Now, I’m not sure what Lewis would say about the modal status of the sentence “There are disconnected spacetimes.”
    Would he say that the sentence is impossible, since no one world contains disconnected spacetimes? Would he say that it’s necessary, because it is a truth about the pluriverse that there are disconnected spacetimes and truths about the pluriverse are necessary (see “How Can We Know”)? Beats me.
    One thing that seems clear is that he can’t say that the sentence is contingent.
    But that’s a potential problem for Lewis, since physical theories might posit or entail that there are disconnected spacetimes, and physical theories are supposed to be contingent.
    In the comment to which (I think) you were replying, I was considering what would happen if a physical theory didn’t entail that there are disconnected spacetimes, but rather, entailed that there are actually disconnected spacetimes. I was wondering how a Lewisian might interpret that entailed statement. As you mention, ‘actually’ for a Lewisian is an indexical. But since there is no world at which there are disconnected spacetimes, there is no context (within any world) relative to which it will be true that there are disconnected spacetimes.
    I wasn’t sure precisely what to make of that, since it amounts to saying that there is no possible context in which it will be true to assertively utter “There are actually disconnected spacetimes.”
    Whatever precisely a Lewisian should say about it, it seems it will have to be in the ballpark of saying that disconnected spacetimes are impossible, or that it is a priori that there are no disconnected spacetimes, or both.
    In any case, it seems like a problem for a Lewisian who would like to interpret these pluriverse-happy physical theories as, in effect, embedding their talk of disconnected spacetimes under the “actually” operator.

    LikeLike

    Reply
34. 

    TParent

    June 15, 2014

    Warning: Self-promotion.
    I argue (fwiw) that Lewis’ dilemma is: Are Lewisian possibilia in the proper domain of physics or not? Since physics aims to account for everything that exists, it seems so. But then, any Lewisian possibilium is presumably physical.
    I am unsure how a Lewisian would reply to this (assuming we philosophers are not able to dictate the proper domain of physics.) If anyone has any thoughts, I’d be very grateful to hear them. (Btw, there’s a draft on my website.)

    LikeLike

    Reply
35. 

    Bryan Pickel

    June 16, 2014

    Surely some aspect of Williamson’s argument is as follows: (1) Lewis’s possible worlds, if they exist, are unified by being maximal spatiotemporally related. (2) If Lewis’s worlds exist, then some of them contain violations of the laws of physics. (3) The laws of physics as specified by the phycists are meant to apply to any spatiotemporarlly related objects. So, Lewis’s possible worlds conflict with physics.
    It’s noteworthy that Lewis himself addresses the possibility that all spatiotemporal relation must exhibit certain features. His response is that possible worlds need not be unified by spatiotemporal relations: “It would be nice to suppose that all worlds are related by the very same relations, namely the ones that we call ‘spatiotemporal’, despite whatever behavioural differences there may be. I do not reject this suppositin. But I am unwilling to rely on it.// What I need to say is that each world is interrelated (and is maximal with respect to such interrelation) by a system of relations which if they are not the spatiotemporal relations rightly so called, are at any rate analogous to them.” (PW, 75)

    LikeLike

    Reply