A few days ago, Rene Zandbergen very kindly pointed me in the direction of Lat. Borg. 898, a cipher manuscript newly digitized by the Biblioteca Apostolica Vaticana. This has 410 pages of handwritten text, written using an alphabet formed almost entirely of astrological symbols (though with a few words apparently of Arabic right at the start, occasional Latin words fragments near the start, and a page of Italian right at the end).
Now, apart from Giovanni Fontana’s technical notes (early 15th century), the Voynich Manuscript (15th century) and the Rohonc Codex (16th century), you’d be quite hard pressed to find any other book-sized manuscript written in cipher from before 1650 (when shorthand started to become fashionable: Samuel Pepys began his shorthand diary in 1660.) So naturally, Lat. Borg. 898 was something I wanted to know more about.
According to the manuscript’s inside cover, it contains “Itineris de septentrionales fructus”, ‘the fruit of a journey to the North’: which also means it is a copy of Montpellier H.505, a handwritten manuscript written by Johannes van Heeck. “Van Hoo?” you may ask.
Actually, Johannes van Heeck was one of the small group of people who founded the ultra-influential Accademia Lincei in 1603, which ran until 1630. Yet not long after, he was jailed for killing an argumentative apothecary in a vicious brawl: even though his ultra-well-connected fellow founder Cesi managed to extricate him from prison, Van Heeck was still quickly banished from Rome. And so it was that he suddenly found himself with the time to go on a naturalist-themed journey to Poland, Pomerania and all points north to look at all of their flora and fauna.
(Incidentally, Johannes van Heeck wrote a book on the 1604 supernova: moreover, van Heeck and Cesi were both at the famous feast in honour of Galileo Galilei where the ‘telescopium’ was first given its name: Galileo and Della Porta both became Lynxes too.)
Johannes van Heeck (‘Heckius’) wrote his notes in Latin using what David Freedberg calls (though without elaborating) the “Lincean code”, which is presumably the simple substitution cipher described above. It shouldn’t be hard to crack, now, should it?
Computer Says No
It seemed obvious to me that the encipherer had not bothered to try to disguise the lengths of words (what American Cryptogram Association people would call an ‘aristocrat’ cryptogram, as opposed to a ‘patristocat’ cryptogram), so immediately it was highly probable that this was only lightly enciphered.
I therefore briefly looked for obvious Latin language cribs (a word with a highly unusual letter pattern, that would only have one or two possible Latin plaintexts) in the early pages, but noticed only this one:
This could easily be MISSUM or MITTAM: but because I also thought it likely that Latin would have several other words that matched the same pattern, decided not to pursue this further.
However, the cipher shapes had already suggested a likely cryptographic pattern to me: that the encipherer was using the astrological aspect glyphs (conjunction, sextile, square, trine, opposition) for vowels, and the astrological sign and planet glyphs for consonants. As a result, I was confident that I should be able to crack it easily.
But when I transcribed the first half of folio 2r (most of the first folio seemed to have been ripped out) and put it into CryptoCrack… the computer said no. That is, it didn’t find anything remotely like Latin. Or Italian. Or Dutch. Or anything, in fact. And when I tried half a page from folio 52r, that too failed to work (though it suggested a word “DISTILLA”).
That was a bit odd: so, with the briefest of nods to my first headmasterly Latin teacher (Richard Sale), I instead put the first few lines of Caesar’s Gallic Wars (“In tres partes” etc) into CryptoCrack: which decrypted them with ease. So CryptoCrack seemed to be working OK: the problem was apparently with Lat. Borg. 898’s plaintext.
At this point, though, I could see enough of the patterns to have a go at it by hand: and so got almost all of the alphabet.
Why did CryptoCrack fail to work on such a lightly enciphered Latin ciphertext? Having thought about it, it now seems to me that the way Johannes van Heeck and his contemporaries used Latin was essentially quite different from the way Julius Caesar wrote: and so using a single overarching “Latin” corpus / statistics is probably far from optimal.
So if in future I find myself desperately needing Cryptocrack to break tricky Latin texts circa 1600, I might put together a much later Latin corpus and see if I can get Phil Pilcrow to add that in as a “Late Latin” language option.
Through Philip Neal’s eyes
One nice thing about code-breaking is that even though a ciphertext usually has only one genuine plaintext, there are often many paths to that destination: while all will employ cunning of one sort or another, none is exclusively right.
And so it was that Philip Neal was having a good cryptological chortle at my expense: he asked me if I had seen the “glaring crib” in the section headed ‘Contra Tremorem Cordis‘. I thought he was referring to the MISSUM / MITTAM crib-like word above (which actually turned out to be TOLLIT), but it turned out he was actually referring to the two words immediately after it.
Nice crib! I wish I’d seen it, because it would have saved me half an hour of crypto hassle – but both roads led to Rome.
The Eye of the Lynx
While David Freedburg’s book “The Eye of the Lynx” doesn’t mention Lat. Borg. 898, it does mention Montpellier BEM H505 (though note that the description there is hyperlinked not to Johannes van Heeck but to quite the wrong “Johannes Eck” (1484-1543)), the book-sized cipher manuscript from which Lat. Borg. 898 was copied (with a few copying errors, inevitably). H505 is apparently subtitled “Mechanica et Naturalia Ioannis Ecchi Lincei”, and (David Freedburg “The Eye of the Lynx”, p.444 note 10):
“is largely devoted to matters of technical and mechanical interest and contains Heckius’s typically rather awkward drawings of machines, instruments, chemical vessels, and a variety of mechanical devices. Introducing the manuscript are thirty pages of medical notes written in Arabic, Syriac, and a strangely hermetic combination of Arabic and the Lincean code, as well as a passionate and moving invocation to the Virgin to assist him in his exploration of the hidden parts of nature.”
[Update: I found a scanned image from H505 here and inserted it below:]
So it would seem that Borg.Lat.898 is nothing more than the dreary text-only child of H.505: but, almost inevitably, H.505 has not yet been digitized, so as far as Lincean cipher manuscripts go, we currently have what we have (Borg.Lat.898) and no more.
Incidentally, If you want to know more about the Accademia Lincei, then David Freedburg’s (2003) “Eye of the Lynx” is a nice accessible read. For more on Johannes van Heeck, Freedberg mentions Alessandrini’s (1978) “Cimeli lincei a Montpellier” pp.288-293 and pp.68-77, plus Gabrieli’s (1989) “Contributi alla storia della Accademia dei Lincei” pp. 1055-1078): but nothing much else is obviously online. Which is shame, because he was a particularly colourful character.
Transcription of fol 1r
[R] calamenti thimi
pulegt cardui benedic-
ti rosarum menthe cr
ispe anam [l. se.] anisi
feniculi ozimi urthi
ce aneti [an }s vad:] angeli
ce feniculi althee
squille iridis turbit
elle: albi ana [}y. Esula9
propter: }y] asali [}vi] galan
ge cinamomi calami [azo:
an }s] infundantur trita
omnia in aceti fortis
simi [ttx.] tridup in lo-
co calido in uase ui-
treo uel terreo uitre
to deinde bulliant