Tuesday, November 15, 2005

Actual Ambiguities

Consider the term: "the actual world". It means, transparently enough, something along the lines of "that possible world which is actual". As with many definite descriptions, this displays a sort of de dicto/de re ambiguity. On the one hand, we might take the term 'the actual world' as a straightforward description (or a kind of modal indexical), so that any possible world can refer to itself by using the term, if we conceive of that world as "counteractual" rather than merely counterfactual, i.e. when we conceive of that world as having been actualized in place of our own world. This is analogous to how the de dicto reading of "the U.S. president" allows it to refer to John Kerry in some possible worlds, even though Kerry isn't president in our world. Much like Kerry could have been president in place of Bush, so, perhaps, some other possible world could have been actual in place of our own. (I'm actually not convinced that's a meaningful suggestion. But I'll ignore such worries for now.)

Alternatively, we might take the term to be a sort of name or rigid designator, the description picking out an actual object, but thereafter holding the reference fixed throughout all possible worlds. So, for example, if we rigidly designate X = "the [actual] U.S. president", then 'X' is just another name for George W. Bush, and in counterfactual worlds where Kerry is president, it would be false to say "X is the U.S. President", because Bush wouldn't be. Applying this to our present problem, we might take "the actual world" to be a name rigidly designating our own particular world. On this reading, if my counterpart on a frozen possible world were to say "I'm actually cold" (in the philosophical sense of "I'm cold in the actual world"), then he would be saying something false, since in the actual world - i.e. our one - I'm not cold. On the de dicto reading previously mentioned, however, he might well be saying something true. It would be true just in case he was cold, on his world.

Okay, hopefully that's all clear. Now, I bring this up because it's relevant to a couple of recent posts over at Fake Barn Country. In one, Jonathan argues that we cannot believe metaphysical contingencies, because a belief that p just is a belief that actually p, but the truth value of the latter belief's content is metaphysically necessary (if true, it's necessarily true; if false, it's necessarily false). But this argument equivocates between the two senses of 'actually' described above.

A belief that p is only equivalent to a belief that actually p in the de dicto sense, i.e. where 'actually p' just means 'p in this world (conceived as actual)', that is, just plain 'p'. And there's nothing metaphysically necessary about the truth value of that. The stronger version doesn't work, because my frozen counterpart believes he is cold (i.e. "RC is cold"), but he doesn't necessarily believe that I, his actualized counterpart, am cold. He doesn't know anything about my situation. His beliefs are about his possible world, not mine. And a belief about his world is clearly not the same as a belief about my world. So, these counterfactual people show that a belief that p is not the same as a belief that actually p -- not in the strong (de re) sense that Jonathan requires.

In an earlier post, Jonathan made the following claim:

"Claim: To imagine a scenario is to imagine that the scenario obtains in the actual world."

I suggested in comments that this is only true in the de dicto sense. A better way to put it might be that we imagine scenarios as counter-actual, rather than actual. We don't imagine them taking place in this very world [the de re actual world], where I'm warm and Bush is president. Rather, it seems to me, we imagine scenarios (fictions especially) as taking place in some other world, a world which we then conceive of as being actual in place of our world.

None of the FBC guys seemed especially taken by the idea, which is odd, because it strikes me as obviously correct. (Then again, Jonathan thought his claim was obviously true too, which I strongly disagree with, so perhaps 'obvious' is a dangerous word to be throwing around in this discussion!) What do you guys think?


  1. I'm more sympathetic to your view, although I'm not convinced that possible worlds talks elucidates anything about the way we imagine. When I imagine p, I just imagine p, and am not committing myself to anything about what world p is in.

    I'm also inclined to reject the move from 2 to 3 and from 4 to 5. I think actuality of worlds needs to be kept distinct from actuality of something in a world. [Actually p] can only be metaphysically necessary if it is true in all possible worlds; but it is manifestly false that [actually p] is true in all possible worlds. It is true in this possible world, which happens to be actual, but in other possible worlds [actually p] is false, because whereas worlds may be actual or possible simpliciter, any p in a world can only be actual or possible in that world. (Molinist and Leibnizian versions of possible worlds analysis -- that is, the old, seminal possible worlds analyses sometimes implicit in the works of people like Molina and Leibniz, easily capture the distinction between the two. Contemporary possible worlds analysis always seems to me to have strayed from plausibility.) Suppose that p is false in this world; nonetheless there may be some possible world in which [actually p] is true. But that doesn't commit us to saying that that possible world is an actual world, because '[actually p] is true' works just like 'p is true'; just as we can say that p is true in some nonactual possible world, so can we say that [actually p] is true in some nonactual possible world.

  2. What is the "actuality of something in a world"? Do you just mean anything that is true in that world, so that [actually p] is perfectly synonymous with just plain [p]?

    That's fairly close to my "de dicto" version. But it's not a very useful term then, if the "actually" is entirely redundant.

    I think the standard (contemporary) approach goes for what I labelled the "de re" interpretation. That way, it is no longer "manifestly false that [actually p] is true in all possible worlds."

    On the standard picture, as noted in my old post on modal indexicals, "modal statements don't say anything about the specific world they are indexed to. Rather, they make claims about the entire collection of possible worlds." Let's pick out the actual world, and name it '@'. Suppose p is true at @. Then from any possible world you like, it is true that [p is true at @]. (There is only one @, and p is true at it. So there is no possible world from which p is false at the world @. Just like there is no other country from which Wellington is not the capital of New Zealand. Once the index is fixed, the truth of an indexical statement is likewise fixed, no matter where you might "ask" it from.)

    That's assuming an S5 modal logic. You might be thinking of something weaker. But I'm not really sure how to conceive of the metaphysics behind those other modal logics.

    What does it mean to say something is necessarily possible, if not simply that it is possible? More generally, what would it mean for a modal statement to be only contingently true? Is the modal multiverse somehow 'dynamic' and radically relative to each individual possible world? (So that other possible worlds differ depending on which world you consider them from? It sounds bizarre. Surely possible worlds should be independent of each other, yielding a nice static S5 multiverse!) I don't get it at all. But if anyone can explain this in a way that makes sense, I'd be very interested to hear about it...

  3. Richard,

    I agree with your way of phrasing it: the 'actually' in 'actually p' is redundant when p is true, and it's not a very useful term (as opposed to the usefulness of distinguishing actual worlds from possible ones, with which it is often confused). What you are calling the de re interpretation is, as far as I can see, simply incoherent. At least, the only way we could accept it as even plausible is if the actual world is the only possible world -- in which case we might as well stop talking about possible worlds and find a simpler modal discourse.

    The modal logic issue doesn't help the de re advocate any. p is true at possible world @ may be true for all possible worlds; but it doesn't follow from this that if @ is the actual world that [actually p] is true for all possible worlds (unless we already grant that [actually p] is a necessary truth). This would be to confuse actuality of a world and the truth of [actually p] in a world. It would be the same mistake if we were to say that ''p is true in @' is true for all possible worlds' implies 'p is true in every possible world', i.e., 'p is necessary'. This just parallels the move from ''[Actually p] is true in @' is true for all possible worlds' to '[Actually p] is true in every possible world'. Again, the only way I can see this being plausible is if the every possible world is the same, which, given that there is an actual world, would imply that the actual world is the only possible world. And again, if that's where we're going we might as well drop the possible worlds jargon.

    So if we call into question the conflation of 'actually p in possible world @' with 'p in actual world @', I don't think the argument moves forward any.

    Perhaps, though, it is due to thinking that 'actually' in a possible worlds analysis should operate like other modal operators. I don't see any reason why this should be the case; 'actually' is picking out something in contrast to 'not actually' -- and I don't see how either can be said to make claims about the entire set of possible worlds. Precisely what they do is pick out some possible worlds in contrast to others. That's what makes them useful in the first place.

  4. Doesn't [actually p] just mean [p is true at the actual world]? That is, on the 'de re' reading, [p is true at @], which you've granted to be a necessary truth. Where's the incoherence?

    (BTW, I've just published a new post which delves a bit deeper into some of these issues. I'd love to hear your thoughts.)

  5. Well, it's entirely possible that I'm confusing things. But while [p is true at @] is true in all possible things, it doesn't follow from this that [p is true at the actual world] is true for all possible worlds; unless, of course, @ is necessarily the actual world, i.e., unless it is true for every possible world that @ is the actual world. In other words: unless the actual world is the only possible world.

  6. Reading your other post I see we've spilled over into the natural question to ask in response to my comment. So my part in the discussion continues over there.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.