Linear A Linked to Linear B.

Patricia Claus at Greek Reporter writes Minoan Language Linear A Linked to Linear B in Groundbreaking New Research:

The Minoan language known as “Linear A” may finally be deciphered with the help of the internet, which can be used to uncover previously-hidden links to the much-better understood Linear B language, which developed later in the prehistoric period. […] Linear A, which was used by the Minoans during the Bronze Age, exists on at least 1,400 known inscriptions made on clay tablets. The language has baffled the world’s top archaeologists and linguistic experts for many years.

But thanks to the pioneering work of Dr. Ester Salgarella, a Junior Research Fellow in Classics at St John’s College, Cambridge, understanding the Minoans’ most ancient language, used during the period of their civilization’s highest flourishing, may be at hand. Her research breakthrough, which has been hailed as “an extraordinary piece of detective work,” could provide the key for linguists to unlock the secrets of the Minoan language – and learn more about its society and culture. […]

Taking an interdisciplinary approach using evidence from linguistics, inscriptions, archaeology and paleography (the study of the handwriting of ancient writing), Dr. Salgarella examined the two scripts. To compare them more easily, she created an online resource of individual signs and inscriptions called “SigLA – The Signs of Linear A: A paleographic Database.” She explains: “At the moment there is a lot of confusion about Linear A. We don’t really know how many signs are to be taken as core signs; there’s even been a partial misclassification of signs in the past. “This database tries to clear up the situation and give scholars a basis for advancement. We don’t have a Rosetta Stone to crack the code of Linear A, and more linguistic analysis is required, but this structural analysis is a foundation stone.” […]

“What we see with Linear A, Linear B, and the two different languages notated (Minoan and Greek) is that the graphic system (script) was continued without considerable change (but of course there are differences), but the languages differ. Upon the transmission process of Linear A to Linear B there was therefore ‘linguistic shift’. However, the graphic system was not deeply affected,” she says.

Now, I haven’t paid a lot of attention to the Linears, but I thought it was common knowledge that A and B were related scripts, and I’m not clear on what the purported breakthrough here is. I got it from the Facebook feed of Slavomír Čéplö (bulbul), and he didn’t mock it, which is some sort of praise considering how readily he mocks things. So what’s the story here? Anybody have more information?


  1. Not sure that there is actually a breakthrough, but her work is aimed at figuring out a Linear A sign-set using computers.

    Surprising thing is, why wasnt this tried earlier?

  2. @zyxt: Yeah, for some reason, I thought this had already been done, but I guess I had just been thinking of the Cretan hieroglyphs, for which the number of witness inscriptions is several nepers smaller.

    One thing a comprehensive online database (with concordances) could be quite useful for would be determining computationally whether the Linear A inscriptions are actually all in a single language or whether there are multiple languages represented.

  3. As far as I can tell, the groundbreaking work here is compiling paleographic variants of the signs in both scripts, and thus establishing their equivalences. I’m not sure. It looks like the work was published in a recent book. It’s nice to see a lot of careful work giving modest but trustworthy results, rather than vice versa vice versa vice versa vice versa, which the bulk of the usual crackpot literature is.

  4. Very true!

  5. Here’s a presentation of a talk on “Diversity and Variation in the writing systems of Bronze Age Greece”, given recently by Anna Judson (who did her dissertation on “The Undeciphered Signs of Linear B”), and by Salgarella (whose dissertation became this recent book). I can figure out a little of the gist of the talk. It looks really interesting.

  6. If I got it right, they seem to say that if we have Linear B syllabic sign which as we know is pronounced “di” and there is a corresponding Linear A sign which looks almost the same then it is likely that that sign was also pronounced “di”.

    If we accumulate enough of such similar signs, we will be able to say how some Minoan words were pronounced (though we still won’t know what they meant).

    That’s progress of course, but still very far from “Linear A may finally be deciphered”.

  7. hat,
    and [bulbul] didn’t mock it, which is some sort of praise
    Third time I’ve heard this assessment of me this week. Also, very true.

    It looks like the work was published in a recent book
    Indeed, and in fact, this book – “Aegean Linear Script(s): Rethinking the Relationship Between Linear A and Linear B” by Ester Salgarella, CUP 2020 – is also mentioned in the article.

    I did not comment on the article because a) Greek is far from my specialty and b) I have not read the book. The research does sound legit and very promising: it basically examines language shift WITHOUT, um, script shift. A not uncommon phenomenon (Sumerian > Akkadian > Hittite), but I think the only instance where the first language is unknown.

    Surprising thing is, why wasnt [figuring out Linear A using computers] tried earlier?
    Because no one cared enough to put in the hard work of collecting, transcribing, annotating and digitizing the data. I know this because that’s the answer to 99% of these questions and also the article mentions this database by name.

  8. @bulbul
    🙂 agree

    Has the horse already bolted? Linear A is already encoded in Unicode. How will they handle changes to their sign-set?

    Or this might be a case where encoding an unknown script in Unicode helps in its decipherment.

  9. @SFReader: A handful of Minoan words have already been picked out that way, but with everything digitized, the process can be expanded and automated. That’s part of how, as I alluded to above, this can be used to distinguish whether Linear A always encodes the same language, by looking to see whether different groups of inscriptions use consistently different vocabularies and phonotactics.

  10. The Linear A corpus isn’t very large, and I’m guessing that whatever can currently be done in terms of matching up signs with Linear B to “read” it phonetically has already been done; that doesn’t seem to be the new part of this. As for LA encoding multiple languages, that seems unlikely a priori given that the large majority of inscriptions are from Crete and a few adjacent islands and more or less from the same period; anyway it would be hard to compare vocabulary without knowing the language(s), and syllabaries are notoriously bad at showing phonotactics.

    Thanks for the link to the presentation, Y. From what I can make out (I don’t know nearly as much about this stuff as I’d like), part of the argument is that sign variants in LB sometimes continue variation that already existed in LA, which if true would have interesting implications for the transmission scenario.

  11. We know that the underlying Linear A script was adapted to write other languages more than once. Besides the Linear B used by the Myceneans, there were the Cypriot syllabaries that, over time, were used to write both Greek and Eteocypriot (another pre-Indo-European insular language that may have been related to Eetocretan/Minoan). So it’s not impossible that even the earlier Linear A may have also been used to write neighboring languages, but it’s by no means certain either.

  12. I suspect what Salgarella believes to be a novel discovery is that a set of signs that had been assumed to be different in A vs. B is instead contained within the variation she has found, so that there is a more, um, linear relationship between the two scripts. Thus, from the Cambridge Blog post about her book:

    >The combination of these two models strongly suggests that LA and LB are more closely intertwined than had hitherto been assumed and unlikely to represent self-contained systems, as generally thought. As a result, these models emphasise strong continuity within the writing practice as a whole, on structural, palaeographic and phonetic (in terms of continuation of sound values) levels.

    In fact, large parts of the book seem to be available for free, for now, here:

    I’d note this interesting line:
    >Before analysing LA and LB as outlined above. it is worth exploring the context in which they ‘were (re-)born’ to us, ie the origin of their nomenclature and classification as separate writing systems. The research question addressed here is why are LA and LB considered two different scripts.

    In other words, while I haven’t seen her use the terminology (yet?), I think the trend of her argument is that they are not two different scripts, and that she has been able to add new findings of previously overlooked consistency that give her confidence rechristening them as a single script – “Linear 1.0 and 1.1”, maybe?

    Anyway, several of the people here… oh, heck, probably almost everyone here knows a lot more about this than I do, so I hope one or two look at the book excerpt and report back.

    EDIT – And indeed, a bit further on, she distinguishes between the “Linear signary” which she argues is constant, and the “writing system” — “the overall repertory of signs (syllabograms, logograms, klasmatograms, numerals) which make the two systems stand apart.”

  13. The distinction might have been clearer if I’d given her definition of ‘Linear signary,” which she calls the “logo-syllabic set.”

Speak Your Mind