A few days ago, Andy Egan noted on the 617 blog that what happens in Matrix 3 will determine, in part, the aesthetic qualities of Matrix 1. (Permalinks bloggered, but right now it’s the second post from the top.) In this case the connection was because what happens in Matrix 3 (partially) determines what is true in Matrix 1, and the relevant fictional facts are relevant to the aesthetic value of the whole.
[Warning: What follows contains spoilers about Matrix 2, the first season of 24, Ulysses and possibly the Iliad, depending on what I feel like writing about.]
When can a later story determine what is true in an earlier story? This turned out to be harder than I expected to figure out. Let’s start with one case where it does make a difference, and another case where it does not make a difference. Both are fictional, though they could be close to real cases.
Martian Leopold
Joyce survived his illnesses in 1941, but this wasn’t obviously for the greater good of literary creation. Deciding that the future was in sci-fi, he decided to write the story of June 17, 1904, where it was revealed that Bloom was really a Martian, and that the Circe episode was not hallucinatory, as everyone had previously suspected, but a literal representation of what happened in Nighttown.24 by Committee
As in the real-world 24, it is revealed towards the end of the first series that Nina is a traitor, and has been throughout the show. This changes what we think about the earlier episodes, including I think their aesthetic qualities. As in the real-world show, it was not decided until the early episodes had been completed, and even screened, that Nina would be made to have been a traitor all along. Unlike, I think, the real-world version, 24 was written by a very fluid committee. Although there was some continuity from week to week, the committee of writers who made Nina treasonous had no members in common with the committee that wrote the early episodes.
I hope you agree that even if Ulysses Part 2 is written, in Ulysses Leopold is human not Martian, and Circe is a hallucination (or perhaps several hallucinations). And I hope you agree that changing the writing structure of 24 in this way does not affect the truth value of claims about the early episodes. To really make trouble, I need a slightly harder case.
Intended Martian Leopold
Joyce survived his illnesses in 1941, which gave him the chance to write the sequel to Ulysses he’d always planned: the story of June 17, 1904, where it is revealed that Bloom was really a Martian, and that the Circe episode was not hallucinatory, as everyone had previously suspected, but a literal representation of what happened in Nighttown.
I think that even here, it is not true in Ulysses that Bloom is a Martian. I guess this is a contentious intuition, but there’s a way to back it up, sort of. Imagine that Joyce intended to write this sequel, but really did die before he had a chance to write it. In this case I think the complete absence of textual clues means that Bloom really isn’t Martian, authorial intention be damned. I know this is an unfashionably fashionable view for an analytic philosophy blog, but I think it’s correct. Call me crazy, call me Derridian, call me a sellout. (Just don’t call me late for breakfast.) I don’t think that in that case Joyce’s intentions matter, and I don’t think his writing out those intentions in a later work matters either.
What makes the difference between these cases? At first glance one might think that a later work matters to what’s true in an earlier work if (a) the later work is written by the same author as the earlier work, and (b) it carries out the intentions the author had in describing what was happening in the first work. But these are neither necessary nor sufficient conditions, as our cases above illustrate. Or at least, so I think they illustrate. One might argue that the later episodes of 24 do have the same author as the earlier episodes, since the author in each case is a committee, and the same committee writes each episode, even if it has different members at the different times. (Mereological essentialism is not true of writing committees, you see.) But I don’t think that will do. Make the earlier episodes written by a single person, who gradually adds co-authors, then gradually drifts out of the process. The later episodes still matter. (Or do they?) And it’s hard to say that we have the same author. (Or is it? My intuitions have quit for the day.)
So what does make the difference? Beats me. It’s easy to say that the later work matters iff it is part of the same story, or narrative, as the first work. This seems true, but the problem is now analysing the concept of being part of the same story. And I have no idea how this will be done.
For anyone who likes more and more absurd examples, it’s fun to play with variations of the 24 case where one of two co-authors stays on, and writes out her intentions for the back story behind what they co-wrote. Or, even better, completely changes her mind about the back story, and instead writes a story that meshes perfectly with what her co-author thought the back-story was, even though she was ignorant of her co-author’s wild views.
Looks like I didn’t get around to spoiling the Iliad. Maybe tomorrow.