I was sitting on a train trying to reconcile the Voynich Manuscript’s vellum dating (1404-1438) with its art history dating (1450-1470), while also pondering the various layered aspects of its codicology (such as the pictures apparently behind some the water nymphs), when an unexpected thought popped into my head.
Might the VMs have its (erased) plaintext as a palimpsest beneath its ciphertext?
That is, might the Voynich Manuscript be an enciphered copy of the document that was originally written in broadly the same place on the same pages? The overall timeline could work like this:-
- 1420 or so: the author compiles his/her books of secrets onto brand new vellum.
- 1460 or so: the author copies each line or paragraph onto a wax tablet, erases the line or paragraph from the vellum, and replaces it with an enciphered version.
If this is correct (and though I admit that the odds are against it, I do think it’s worth considering), the issue would then become whether the original plaintext might somehow be made legible again – that is, whether the plaintext’s ink made a sufficiently permanent contact with the vellum that some kind of imaging might now reveal it, even if it was heavily erased at the time.
Over on the MapHist mailing list in recent days, there has been a lot of ink-related discussion about the Vinland Map (which has its own curious cryptographic angle that Jim Enterline has been doggedly pursuing for years): this interesting analytical paper on vellums and inks was cited, as well as “Ink Corrosion” by Gerhard Banik. This latter paper discusses Haerting’s “conclusion that only inks containing iron(II) salts can cause ink degradation damage. The other components of the ink […] do not cause noticeable damage to the support medium.”
At first glance, you’d have to say that the VMs really doesn’t appear to be a classical palimpsest: but this is because palimpsests were typically washed clean and the second layer of writing put down at right angles to the original direction of the text.
But perhaps some of the original ink layer wasn’t erased, in forms other than that of almost imperceptible damage to the support material: I’m thinking in particular of the “hidden house” on f77v that I discussed here back in June 2009.
I’m pretty sure that this originally depicted some kind of house structure (drawn in faint, straight lines), but that the “water nymph” and the “wolkenbanden” were subsequently added in a later pass (in a completely different ink) to mislead the eye.
Codicologically, what is most interesting here is that the original ink (i.e. from the “house”) might be identified and separated out into its own layer… and that understanding its particular composition (and the way that it interacts with the support material) might point to a way of imaging the manuscript’s original (1420?) ink layers from beneath the subsequent (1460?) ink layers (if it exists, of course).
Basicaly, why even try to break the cipher, if it might be possible just to read it? Now wouldn’t that be a surprise for everyone!