stories as species

February 23, 2022 9:40 AM

Forgotten books: The application of unseen species models to the survival of culture "According to a new paper published in the journal Science, {paywalled] an international team of researchers has adapted an ecological "unseen species" model to estimate how many medieval European stories in the chivalric romance or heroic tradition survived and how much has been lost. "
posted by dhruva (6 comments total)

The team behind the study made a website too.

I'm of too minds about this. On the one hand, it's reassuring, as Matthew Driscoll says, that this study reaches the same conclusions philologists had reached a century or so ago. Partly because I've noticed some revisionism lately discounting old scholarship on this topic, which makes the study valuable.

On the other, it feels weird that so much attention has been given to a flashy way of confirming what scholars knew already a hundred years ago, when genuinely fresh scholarship gets little or no attention.

That said, it never ceases to strike me as remarkable how much literature has survived from the medieval era. In the Icelandic case, 77 percent of literary works, and 38 percent of Old English and Middle English. I'm glad to get some secondary confirmation.

Also, the phrase "some were no doubt lost in library fires" from the article made me laugh. Indeed, no doubt. Just the only manuscript containing the most famous Old English text of them all, Beowulf, was badly charred in a fire. And the 1728 fire in Copenhagen did a lot of damage too, including to Icelandic manuscripts which had been shipped to Denmark (though most was saved by a handful of Icelandic students in the city who carted off what they could take).
posted by Kattullus at 10:32 AM on February 23, 2022 [6 favorites]

I love to see scientists in different fields trading tools. This sounds pretty reasonable. I have a colleague who used a statistical technique that analyses cattle breeding data to predict the host range of fungi. It doesn't totally work, but it works well enough to get people to look at the problem in fresh ways.
posted by acrasis at 2:27 PM on February 23, 2022 [1 favorite]

(To explore these themes on board game form, I recommend Veritas, a game in which each player controls the spread of a particular text through a network of monasteries which occasionally burn down.)
posted by kaibutsu at 2:29 PM on February 23, 2022 [4 favorites]

So somewhere there is Chaucer's equivalent to Terry Pratchett?
posted by The otter lady at 8:45 PM on February 23, 2022

The spread of manuscripts with occasional typos is a metaphor for the evolution of DNA. Bioinf techniques have been mobilized to determine the ms-zero for The Canterbury Tales [Nature 1998] and The Phylogeny of Little Red Riding Hood [PLOS One 2013].
The tragedy of Sophocles is that, of the 123 plays he is known to have written, 116 have disappeared. I haven't seen a convincing case that only the best have come down to us and we're better off not having to trudge through his 100+ workaday pot-boilers.
posted by BobTheScientist at 1:18 AM on February 24, 2022 [1 favorite]

I was surprised to learn that the "Chao1 method" is so simple, even a math-challenged person like me can understand it:

S_chao = S_obs + ( f₁² / 2f₂ )

where S_obs is the total number of distinct surviving works, f₁ is the number of works with 1 surviving copy, and f₂ is the number of works with 2 surviving copies. In other words you're estimating the richness of the total population based purely on the diversity of your sample and the relative number of species that you only found once or twice. It's kind of amazing to me that this simple calculation produces such good results.
posted by Gerald Bostock at 7:29 AM on February 24, 2022 [4 favorites]