Love it when machine learning re-discovers the field of linguistics every couple months. https://twitter.com/anthrocharya/status/1297288137761345536

Tbf I should also consider this thread; I’ve read the abstract three times (paywall + being on phone) and am still struggling with why this is Nature-/press release-worthy if not just because it’s using computational methods. We know how language gets transmitted, after all... https://twitter.com/thai101/status/1297316349329444864

...and there’s quite a bit of fieldwork in revitalization/documentation about concepts (of course!). Waiting to bestow judgment until I read the paper, but it’s an awfully small dataset and an awfully big fuss for something that (I thought) was pretty clear in ling already?