Very unusually, this is for pleasure! I’m off to Copenhagen to see Radiohead 😀

No worries! The pedagogical habit is too ingrained for me to just provide a transcription, but I will add that ‘hu.’ looks like a non-standard abbreviation of ‘huius’, and that the next word agreed with it. (This brought to you from a coach to Heathrow, so should come with a slight caveat.)

It does indeed!

Of course! I didn’t mean the whole article.

I can believe it – people have obviously been scrambling to publish on LLMs, so a bit of duplication is no surprise. As long as they’re making good points, the more the merrier!

Mind you, Wyclif got there first on LLMs being indifferent to the truth (“non curat sive vera sive falsa dixerit”) and their outputs being bullshit (“qui nutriebantur in croceis amplexati sunt stercora”) despite the occasional truth (“quandoque vera balbuciat per accidens”). bsky.app/profile/brun...

#doomscrolling

I’d never thought to try Biblissima instead of Gallica! After a quick experiment, I can see how it would be quicker. Less fussy download process for a single page, too.

Complaint shamelessly lifted from the third comment here (archive link because Roger's site is currently down, lack of shame because the comment was mine in the first place): web.archive.org/web/20250419...

Wait, had they recently replaced the one they’d had for years, whereby a search for ‘8 E XV’ would find Royal MS 8 E XV but a search for ‘8 E.XV’ wouldn’t, and (unbelievably) a search for ‘Royal 8 E’ wouldn’t find anything whereas a search for ‘Royal MS 8 E’ would? Because that one seriously sucked.

🤣

“What they do not offer is replacing the need to learn the basic skillset of the discipline first”. In the case of palaeography, crappy HTR is now useful in much the same way as crappy OCR of early printed books has long been useful: it can help competent researchers find needles in haystacks. (2/2)

Yes, this! As Georgy Kantor has put it in a piece about AI: “what unites successful applications is that they allow already expert human beings, capable of evaluating and directing the work, to deal with the amounts of data which would have been unmanageable otherwise…” (1/2) x.com/GeorgyKantor...

Good sleuthing! There may be some material for comparison in DMLBS s.vv. ‘pugnata’ and ‘pugillus’ 1. In a similar vein, there’s also ‘palmata‘ (s.v. ‘palmare‘ 6) and ‘palma‘ 2b. Agreed that the definition of Occitan ‘ponhierat‘ sounds dubious – quots or it didn’t happen! logeion.uchicago.edu/pugnata

Thanks! Working from home as usual, but hopping the Channel (or going under it) a few times a year. A massive relief, obviously. Fingers permacrossed for something wicked to come your way too...

Thanks! It sounds great, but I’m only 3 months into a 3-year postdoc (😌) so I’ll give it a miss. (Incidentally, there’s a bugbear of mine in the FPs: knowledge of Latin won’t be assessed via a test. I do hope the interview makes up for it, because qualifications are obviously not a reliable guide.)

I’m afraid not! The only piece of secondary literature that comes to mind (apart from the one reviewed at the link below) is actually about Greek, though it isn’t mentioned in Tanner’s article: Peter Maxwell-Stuart, Studies in Greek Colour Terminology (2 vols 1981). bmcr.brynmawr.edu/2016/2016.03...

I had to read up on χλωρός for the DMLBS entry on ‘viridis’. Thirteen years later, all I can really remember is that it’s hard to determine where the balance lies between green and yellow. Anyway, the relevant part of the entry (among others, perhaps!) is sense 6: logeion.uchicago.edu/viridis

Also got me wondering what ain’t no country I’ve ever heard of. Checking the reds, I found only the Comoros. Confirms my suspicion that my knowledge of African countries comes from my dad’s stamp collection (Kenya c1960) where it’ll have been disguised as Madagascar: en.wikipedia.org/wiki/Postage...