Linguistic Calvinball

Translation is not a cross between free association and the Game of Telephone

One of the most difficult challenges in linguistics is translating written text when you (1) don’t know what the character-to-sound correspondence is (including whether the script is alphabetic, syllabic, or ideographic), (2) don’t know what language the script represents, and (3) don’t know whether it’s read left-to-right, right-to-left, or alternating every other line (boustrophedonic script). This was what Arthur Evans, Alice Kober, and Michael Ventris were up against with the Linear B script of Crete. That they succeeded is a testimony not only to their skill as linguists and to their sheer dogged persistence, but to the fact that they had absolutely astonishing pattern-recognition ability. Despite my MA in linguistics and decent background in a handful of languages, I can’t imagine taking on such a task, much less succeeding at it.

The problem becomes even thornier when you consider that what appears to be a script might be asemic—something that looks like a real written language but is actually meaningless. (Just a couple of weeks ago, I wrote here about an asemic text called A Book From the Sky that the creator himself said was nonsense, but that hasn’t stopped people from trying to translate it anyhow.)

Which brings us to the Rohonc Codex.

The first certain mention of the Rohonc Codex is in the nineteenth century, although a 1743 catalog of the Rohonc (now the city of Rechnitz, Austria) Library might refer to it—it says, “Magyar imádságok, volumen I in 12“ (”Hungarian prayers in one volume, size duodecimo”).

As you’ll see, that the text represents prayers, or is even in Hungarian, very much remains to be seen. The size matches; duodecimo means “twelve sheets, approximately 127 millimeters by 187 millimeters in size,” and given that some of the earliest guesses about the book’s contents were that it was a prayerbook in archaic Hungarian, it’s possible that the catalog entry refers to the Codex. The paper it’s printed on appears to be sixteenth-century Venetian in origin, but of course this doesn’t mean that’s when the book was written—only that it’s unlikely to be any older than that.

A facsimile of the Rohonc Codex [Image is in the Public Domain]

The drawings are rather crude, and the lettering doesn’t resemble any known script, although various linguists have compared it to Hungarian runes, Dacian, a dialect of early Romanian, and some variant of Hindi. Others think it’s simply a forgery—asemic, in other words—with a sizable number attributing it to the antiquarian Sámuel Nemes, who was known to have forged other documents.

There’s no sure connection between Nemes and the Rohonc Codex, however. He’s not known ever to have handled the document, and certainly never mentioned it. So this seems as tentative as all the other explanations.

Attempts to use the statistical distribution of clusters of symbols, invoking such patterns as Zipf’s Law—the tendency across languages for the word rank to be inversely proportional to word frequency—have also failed.

Like with A Book From the Sky, this hasn’t stopped hopeful scholars from claiming success. Some of them have been eye-rollingly bad, like the solution proposed in 1996 by one Attila Nyíri of Hungary. Nyíri combined some Sumerian symbols with chance resemblances to the Latin alphabet, and used such expedients as rearranging letters and letting the same symbol correspond to more than one sound, and still came up with gibberish like, Eljött az Istened. Száll az Úr. Ó. Vannak a szent angyalok. Azok. Ó. (”Your God has come. The Lord flies. Oh. There are the holy angels. Them. Oh.”

I’m perhaps to be excused for being reminded of the Dick and Jane readers. “Oh, Jane, see Spot. See Spot run. Oh, Spot, don’t roll in that dead squirrel. Oh.”

Another attempt, this one only marginally more plausible, was made by Romanian linguist Viorica Enăchiuc, and hypothesized that the document (1) is read right-to-left and bottom-to-top, and (2) was written in a Dacian dialect of Latin. Enăchiuc came up with lines like Solrgco zicjra naprzi olto co sesvil cas (“O Sun of the live let write what span the time”), which still isn’t exactly what I’d call lucid writing.

Then there’s the Indian linguist Mahesh Kumar Singh, who said the Codex is written left-to-right and top-to-bottom in Hindi, using an obscure variant of the ancient Brahmi script. Singh translated one passage as, He bhagwan log bahoot garib yahan bimar aur bhookhe hai / inko itni sakti aur himmat do taki ye apne karmo ko pura kar sake (”Oh, my God! Here the people is very poor, ill and starving, therefore give them sufficient potency and power that they may satisfy their needs.”) His “translation,” though, was immediately excoriated by other linguists, who said that he was playing fast-and-loose with the script interpretation, and had come up with symbol-to-sound correspondences that were convenient to how he wanted the translation to come out, not what was supported in other texts.

So the whole enterprise has turned into the linguistic version of Calvinball (from Bill Watterson’s brilliant Calvin and Hobbes). If you make up the rules as you go, and never play by the same rules twice, anything can happen.

The upshot of it all is that the Rohonc Codex is still undeciphered, if there’s even anything there to decipher. Like the more famous Voynich Manuscript, it retains its aura of attractive mystery, because most of us can’t resist a puzzle, even if a lot of the best linguists think the script is nonsense. Because how do you prove decisively that something isn’t sensible language?

After all, there are still people who think that Donald Trump’s speeches make sense, even when he says shit like, “I saw engines about three, four years ago. These things were coming—cylinders, no wings, no nothing—and they’re coming down very slowly, landing on a raft in the middle of the ocean someplace, with a circle, boom! Reminded me of the Biden circles that he used to have, right? He’d have eight circles, and he couldn’t fill ’em up. But then I heard he beat us with the popular vote. He couldn’t fill up the eight circles. I always loved those circles, they were so beautiful, so beautiful to look at.”

So maybe “Oh. There are the holy angels. Them. Oh,” isn’t so bad.

The bottom line is that linguistics is not some kind of weird hybrid of free association and the Game of Telephone. When translating a previously undecipherable script succeeds, it always hinges on finding the actual rules and patterns governing the language in question, not making shit up as you go along.

In any case, I’m sure there’ll be further attempts to solve the Rohonc Codex. Which falls into the “no harm if it amuses you” department. And who knows? Maybe there’s a team made up of this century’s Evans/Kober/Ventris triumvirate who will actually succeed.

All I know is that attempting it is way above my pay grade.