Or so says Emmanuel Oliveiro; Ruth Schuster reports for Haaretz (archived):
Decades after a number of unknown alphabets were discovered among the Dead Sea Scrolls, and against all odds, Emmanuel Oliveiro of the University of Groningen in the Netherlands believes he has cracked the “impossible” one known as Cryptic B. The code had been considered to be impossible to decipher, mainly because of the sheer paucity of Cryptic B material. All we have are isolated fragments from two scrolls called 4Q362 and 4Q363, and a few spots in other scrolls where scribes briefly introduced Cryptic B in the middle of a Hebrew text, Oliveiro explains, in the journal Dead Sea Discoveries in December.
Oliveiro’s process was based on analysis and intuition, similar to the methodology the scholar Józef Milik used when deciphering Cryptic A in 1955. Both began with assuming that they were dealing with a mono-alphabetic substitution system– where each of the 22 letters of Hebrew or Aramaic is consistently replaced with a specific cryptic sign (as in – say A is always be replaced by $). […]
But the key breakthrough was suddenly realizing that a sequence of five letters in a Cryptic B fragment might represent the five-letter Hebrew word Yisrael, spelled yod, sin, resh, aleph, lamed. It is true that the resh did not survive the eons intact. But looking at the high-resolution image of the age-darkened fragment – the word ישראל (Yisrael) leaps out. “Once you’ve seen it, you can’t unsee it,” as Oliveiro tells Haaretz by Zoom.
Many more details and images, along with history of the scrolls and their other scripts, at the link; I agree with David Weman, who sent me the link, that the quote at the end is delightful:
So, without certainty, Oliveiro cracked the impenetrable. “I told my friends and wife that I am going to try this and they’re like, you could be stuck here for 40 years and never crack the code,” he says. “And what do you hope to find anyway, a secret felafel recipe? But once I saw it – I think it was quite fast.” How fast? About two months to cross the desert of Cryptic B and see Yisrael.
Thanks, David!
Reason to click through, the photo of the key word Yisrael looks like a cubist portrait of Grouch Marx.
I take it there’s not a code so much as very stylized Hebrew signs, or it wouldn’t make sense to talk about the resh not surviving.
The paper is here (Open Access).
Although it is so fragmentary, I find it completely convincing.
Shit, really? That’s not a new-style WOW-style headline? EDIT: Cryptic B has not “definitely” been “cracked”.
@Y:
Yes, it looks sensible. I like its conclusion that the documents in question seem (as far as the extremely fragmentary nature of the evidence goes) to be fairly run-of-the mill (for Qumran.)
Interesting point that the purpose of the code (which is not a terribly obscure or “secure” one if this paper is right) may have been more related to prestige than effective concealment.
I also liked the expression “the מצאתי moment” after a second, when the penny dropped. (So to speak.)
The fragments are so small that it is not instantly clear to me why these texts were chosen to be encrypted.
On the other hand, among the sectarian texts, the Qumran pesharim and some others were presumably available to be read only by initiated Essenes or Ossenes, which name variants I consider to be, despite fifty-some other published proposed etymologies, a clipped version of their self designation, ‘osey hatorah, observers of torah–a name not accepted by, among others, Sadducees and Pharisees.
@David Eddyshaw and Y
The established Israeli Hebrew equivalent of “Eureka!” having for decades been ! מצאתי , its appearance here is not strange.
I don’t suppose for a moment that it is. It was just new to me.
I don’t know Israeli Hebrew, so I had to mentally translate it from Biblical Hebrew to English to Greek. And then English again.
Incidentally, I seem to recall that the psilosis is because A. was actually exclaiming in Doric, but I may have remembered that wrong. Wiktionary just references the Attic εὕρηκα.
Literally, εὕρηκα = ‘I found’ = מָצָאתִי. In my usage, it is not a fixed expression. I would only use it if I had been looking for a physical object, and would more likely add the object pronoun, as in English. For a more general use, יֵשׁ yesh would be more fitting.
Let me guess, LLM was involved?
Y : EDIT3: yeah, it can read the abstract now, thanks.
…hasn’t made it to German.
(Admittedly I don’t think I’ve heard anyone actually say it. But it’s always spelled heureka.)
@V:
I wouldn’t have thought so. I interpreted it as a kind of joke (and quite an amusing one, in context), though if M is right, it may not have been meant as such, but simply as the local equivalent of “eureka.”
Y’s Sprachgefühl seems to suggest otherwise, but they did recently say they’d been out of regular contact with Israeli Hebrew for a while IIRC.
I can’t access the article without a CAPTCHA that does not appear without my ad-blocking
Yes, that happened to me, too, even when I experimented with coming from several different countries. But I did manage to read the paper itself (it’s quite short.)
DE, True, but I have kept up with the language to some degree through family, the internet, etc.
David Eddyshaw: you have to understand that, as someone who actually worked on LLMs twenty years ago and also did AI for computer games, currently calling LLMs “AI” gives me a lot of “what are you talking about” vibe. What the fuck are are talking about — they have nothing in common.
And as someone who has read Marvin Minsky’s ideas about Generalized AI. Those are three completely different things.
Why do fuck do they call LLMs AI?
Absolutely. The marketing of LLMs as “AI” is the central fraud of a thoroughly fraudulent project. (That‘s the reason why they call them “AI.”)
David Eddyshaw: We have nothing to do except celebrate Christmas, saint Jordan’s day, (‘6th of January) and all of of the others, until it all blows away, and we see what what we can to with the remains of the financial system.
what V (and DE) said! i suspect the bubble bursting may gonna make 2008 look like 1928.
Any timeline on these predictions? I am still waiting for bitcoin to disappear in a puff of hot air.
In the end, it’s just Markov chains, as applied to human societies. It’s quite simple. https://clarkesworldmagazine.com/kritzer_01_15/
Language Hat pondered the question of (H)EUREKA in 2010. Nobody knows the answer; speculations include (1) the spelling was influenced by other Greek words starting with eu-, or (2) the story about Archimedes was transmitted only through Latin texts, in which the Greek word was written without breathing marks.