How would you resolve if a parallel effort not affiliated with the challenge recovered a passage meeting the length and completeness criteria? eg this was not affiliated with the challenge https://www.theguardian.com/books/2024/apr/29/herculaneum-scroll-plato-final-hours-burial-site

@LarsDoucet Can this resolve?

@bohaska As i understand it this doesn't resolve Yes yet, because it's not a "known lost work".

It has to fulfill this criterium: "The work in question was known to us to have existed, but we don't currently have more than 10% of it"

@Mqrius Correct. And even if it was confirmed to be a work known to have existed, I would want final confirmation on the word count too. Pretty sure this meets the word count criteria, but AFAIK this is not a work we knew of, although it certainly is new.

Chances are looking really good that we get this by 2030 though.

@LarsDoucet it doesn't even match the word count actually;

And in addition, the submission includes another 11 (!) columns of text — more than 2000 characters total.

2000 characters doesn't make 1000 words in Greek.

@Mqrius Gotcha, and so we wait!

How much text needs to be recovered to say one "work" is recovered? Would a poem, a chapter, an entire scroll, or an entire book be required?

@WesleyJB Let's stipulate that the complete work has to be at least 1000 words long.

As for what constitutes a work -- it's possible that e.g. a collection of all of Sophocles' plays, which contains within itself complete records of several named plays, but the scroll that contains them is itself only 85% recoverable or whatever -- would definitely count.

So let's further stipulate:

• It has to be a known work, that is known "by name" (it doesn't have to have a consistent title -- like the literal words of the title doesn't appear anywhere in the work itself --, but it should be known in the field by a particular name, and when recovered, recognized definitively to be that work)

• Very short works e.g. poems and letters, if they are under 1000 words, don't count

• Whatever the work is, if it's over 1000 words in length, it counts

• If the work in question is e.g. one of Sophocles' plays, but found within a larger document that is itself a collection, and the larger document is incomplete but the work in question is complete, then that feels like it should count. It definitely counts in the case of Sophocles's plays.

• Resolving that last one I will reserve the right to rule on if it's ambiguous. This leaves some subjectivity but I've painted myself in a corner and caused even more ambiguity in the past by trying to nail things down too precisely ahead of time and winding up with a coin flip land on its side anyways. "I'll know it when I see it" and promise to try to be fair.

@LarsDoucet I feel like the 1000 words thing is challenging -- I would expect the modal result to be, "well, we have a bunch of fragments of some Sophocles stuff, but we don't have anywhere near the whole play, but that's still enough to study for a hundred researchers' lifetimes." But it's entirely plausible to me we'd get like 85% of a 1000-word work that is known by name to classicists within that broader jumble...

@DaveK Sure, but that’s beneath my threshold! This particular market is about works that are more than just a few pages in modern typesetting, and getting essentially complete ones.

@LarsDoucet I think my point is more that I have no clue about the relative frequency of by-name and by-limited-quotations letter of ~1500 words in the potential Vesuvius corpus versus longer works.

I do think "we recovered ~85% of words in at least one 1500 word stretch" feels very plausible at this point (fingers so, so, so crossed that they pull it off)

@DaveK True enough. Want to make a separate market for that? I’d bet people would bid on it

@LarsDoucet Sure, why not, here: https://manifold.markets/DaveK/will-the-vesuvius-challenge-recover-1bf59f4046bc?r=RGF2ZUs