Sunday, October 23, 2011

Lewis on Fictional Worlds

My friends assure me that D.K. Lewis was completely serious about his belief in the real existence of possible worlds. For me they're a useful heuristic, but I guess it helps to go whole hog, since to do that commits you to a more robust view of truth: it really is true in a possible world that something possible happened although it didn't happen here.  Lewis's idea makes truth more robust by shifting the adjective possible from modifying truth to modifying a set of worlds.  Then not only are possible things true in some worlds, but it is true, pure and simple, that in all worlds taken together every possible thing is true.

Which can be a depressing idea, depending on the vividness of your imagination, and depending also on your views of counterparts and zombies and the like.  (I myself would want instead to assume that all my counterparts are zombies, and that I escape death at every moment by threading the maze of worlds where my counterparts, are, at every moment, dying.  I could believe in a kind of asymptotic immortality this way, or a kind of perpetual standing wave of reincarnation into some still-living version of myself.  Or, really, a kind of packet-switching of the entity or identity consisting of myself plus all my zombie counterparts; me, I'd always exist in one of the packets getting through to the future, and not in the packets that weren't.  This suggests, by the way, a reason to be a one-boxer: I want to live in one of the one-box worlds.  Lewis, a two-boxer, himself admits that the one-boxers would do better than he would, as a matter of fact.   He'd still pick one box though, since that couldn't causally change his outcome.  But to resume the thread of my discourse....)

If Lewis is serious, and if he's right, he would justify, and also give a spooky and discomfiting authority to, a kind of literary fiction (and fan fiction too) which it's harder to justify if there's just one world, "la vie, la seule qu'on a." Here's how Lewis says we should think about fictional worlds:
The worlds we should consider, I suggest, are the worlds where the fiction is told, but as known fact rather than fiction.  The act of story-telling occurs, just as it does here at our world; but there it is what here it falsely purports to be: truth-telling about matters whereof the teller has knowledge.
Lewis doesn't say why the fiction has to be told, perhaps because he thinks it's obvious: the world of the fiction is one in which the events occur and in which the fictional narrative also exists, since it is, at least implicitly, part of the fictional narrative that it exists.  Now, I myself don't think that it is an implicit part of a narrative that it exists in the world it purports to describe, for reasons I've mentioned in previous posts, but I don't think this actually matters very much for Lewis's view.  I'll note though that he does seem to worry a similar issue at the very end of the essay, when he raises a question about how to treat a fictional character who in the fiction is telling a fiction. But I think all we need to accept is that any consistent fictional world, as long as it is possible, really exists somewhere. Some of those worlds will indeed include true accounts of the events in the world that are counterparts of the fictional accounts of those events in our world.  But I don't see that they need to. (In fact if they did need to, it would seem that there might be an infinite regress to worry about, since we'd have to add to any narrative that it is a true account of the world, which would change it and require another such assertion, etc. Lewis seems to prefer to go the way of collapsing iterations rather than allowing for this infinite regress, but sees the collapse as problematic too, since now it would be hard to keep separate fiction and the things that are fictional within that fiction.)

So every consistent and possible fictional world really exists, somewhere.  Then the work of the writer of "realistic" fiction (where realistic means something like "possible") is a kind of discovery of truth, the truths consistent with the writer's possible and therefore true premises.  It's like math, or like chess, or like their intersection: given a possible state of affairs, here is an array of incidents, events, characters, interactions, and outcomes consistent with it.  Since possible states of affairs are all of them true somewhere, so are all the incidents, event, characters, interactions, and outcomes consistent with them.  The fiction writer discovers truths which we might be glad to know, or be horrified to know.  But they are true, somewhere (unless you go my zombie route, in which case their truth won't matter as much: they'll just really be avatars -- in the gaming sense, not the strongly counterpartial sense in which they'd have something of the meaning they have as reincarnations - coincarnations, perhaps, of real souls.)

So reading fiction is reading about something that's true somewhere.  If there are inconsistencies, then your reading about multiple truths, all of which may obtain in different places, even if not simultaneously.  And the kind of literary criticism so popular a century or so ago, where critics hazard and argue for strange back-stories: all that's true too.  Hamlet was a woman (in many many worlds, in all the worlds in which she existed); was Claudius's son; was his daughter; was old Hamlet's murderer.  Desdemona did betray Othello with Cassio (in many, many worlds), and Emilia did sleep with Othello in many of them.  Fan fiction consistent with the authoritative fiction it riffs on is all true too, somewhere.

Yes, all consistent fan fiction is true, and all consistent back-story mongering is true, but I am making a stronger claim than that.  It can be, is somewhere, true of our Hamlet, our Desdemona, that is to say, Shakespeare's Hamlet and Desdemona, if it is (possibly) consistent with what we know about them.  But if it is consistent with what we know about them, well then these are things that are true, and true in possible worlds as close as you could want them to be to our own. ("Closeness" of possible worlds is how Lewis handles overall resemblance or divergence between worlds.  The worlds in which Hamlet has three eyes and is green are farther from us than the worlds in which he looks like Jude Law.)

Maybe literary critical arguments are another way of arguing for or against closeness of possible worlds. "Only an idiot would think Claudius was innocent" really means "The possible worlds in which Claudius is innocent are each much farther away than its almost perfect replica where he's guilty."  Maybe.  But that world still exists, and that story, that solution to the chess problem the play presents to its critics, is still true.

If Lewis is right, just imagine the Sartrean seriousness of the responsibility of the critic, or the fan-fictionist, or any audience member guessing at what happens after the story ends.  It's not that we cause those worlds to exist by imagining them.  It's that we come to know that they exist, and whatever claim those events have on our emotions is a claim we now can't dodge.

Writing would be like writing on an Etch-a-Sketch or frosted window: we expose fragments of what's behind it in our writing, and then can't unlearn what we've learned.  As pure epistemological beings, we can know in theory that all possible things are true.  But as empirical humans, the true things that we learn explicitly about have a very great claim on our attention and care, our Sorge. Which suggests we should be cautions about the truths we learn through exposing ourselves to fiction, or through exercising our own imaginations.

Luckily I can't believe any of these possible worlds are real, as the possibly still living and thus-believing Lewises do (or as our this-worldly Lewis was said to).  But it's possible that they are.  And then all you could do is hide.

(The whole thing is Lemian or Borgesian. Lewis would be committed to think, I am sure happily, that the whole Library of Babel really existed, extensionally. For Quine, the whole library may be found condensed, amply, in the alphabet.)

Wednesday, October 19, 2011

Clock-watching, or Truth, sixty times an hour

In the Critique of Pure Reason, Kant offers some "axioms about time in general":
Time has only one dimension; different times are not simultaneous but sequential (just as different spaces are not sequential but simultaneous).... Different times are only parts of one and the same time.... the proposition that different times cannot be simultaneous could not be derived from a universal concept.... Hence it is contained directly in the intuition and presentation of time.  --Pluhar's translation, A32 = B48
Sequence is how time is related to counting, which is to say that time can be measured by a numbering clock always counting upwards.  Einstein's thought experiments about time really consist of showing the kinds of things that can act as clocks: viz., any physical system that follows the then recently articulated fundamental axioms of arithmetic: the idea of a defined interval and the idea of a succession of such intervals.  (Thus two mirrors bouncing light back and forth to each other are a clock.) Since pretty much all things can be clocks (movies pre-eminently, ticking at 24fps), the universe of things is a universe subject to time. (The fact that Einstein proved Kant wrong is not really germane here: what matters is the domain that space and time map out.)

At any rate, I thought about some of this when I spent a couple of hours at Christian Marclay's The Clock, the real-time exhibition to end all real-time exhibitions.

As everyone knows, Marclay spliced together bits of film totaling twenty-four hours, each bit containing somewhere a (legible) clock, whether a digital sign like the one above, a wrist-watch, a sun-dial (I'm guessing: I don't think I saw one), or someone telling someone else the time.  Some movies, parts of some scenes, appeared more than once -- if a character worried about the time keeps looking at his watch in some movie, his watch may appear more than once in that movie, as time hurries by or seems to drag on forever.  Since most movies aren't in real time even when they're showing continuous time, the time difference between those clips is different (and more accurate) in The Clock from what it was in the original.

In His Girl Friday about twenty minutes passes on the clock in the twelve real minutes that Hildy and Walter are talking; in 24 the commercials allow for elisions but also compress the time a little bit so that particular deadlines can be met during the show and not during the ads.  As far as I know High Noon is in genuine real time, as is Robert Wise's great boxing movie "The Set-Up" (1949).

This is from The Set-Up, at the end, and,
for all I know, from The Clock

You can certainly set your watch by The Clock, and one of the interesting parts of the experience is knowing what time it is as you watch; you don't have to worry about checking.  Watching is checking.  And yet you kind of forget that also: the fragments are all so gripping that you're hooked -- 60 times an hour.  We regret leaving every scene for the next, or would regret it if the next scene weren't so immediately compelling.  The movies!
Part of being gripped, though, is looking for the watch or clock or time-piece in each scene.  (It's a pleasure to see clocks you recognize from the real world.  My watch was there! (Someone trying to pawn it for drugs, at just about 1:17 p.m.  Kind of early in the day, come to think of it; must have been a heroin addict.)  My grandmother's clunky alarm clock!  So we watch (!) somewhat differently from the way we would the original movie.  It might therefore seem surprising that we're hooked, scene after scene.

And yet in another way it's not surprising.  The experience is decontextualized, in the best possible sense, the way movie trailers decontextualize the scenes they show.  We get little fragments of intensity, all the more intense for being unexplained, unassimilated to some reasonable desire or goal or way of making oneself secure in the world, the goal of all characters in sequential narration.

Sequence again, then.  The Clock quilts together scenes from a century or so of movies.  We're not looking at the sequential unfolding of time.  We're looking at different spaces, maybe 2,880 of them (the clips are not the same length; they probably average thirty seconds apiece, but who knows?) that are not simultaneous, and different times that are not sequential, but all brought to the spatial simultaneity, the spacial equivalent of simultaneity, on the video screen.  Time and space are made mosaics of each other.  We recognizes bits and pieces -- here a scene, there some sort of time-keeper.  They all come together.  It's beautiful.