The Voynich Manuscript

The infuriating Voynich Manuscript (A.K.A. “Beinecke MS 408″, or “the VMs”) contains about 240 pages of curious drawings, incomprehensible diagrams and undecipherable handwriting from five centuries ago. Whether a work of cipher genius or loopy madness, it is hard to deny it is one of those rare cases where the truth is many times stranger than fiction.

Its last four hundred years of history can be squeezed into eight bullet points (though there’s much more detail here if you’re interested):-

• Circa 1600-1610, it was (very probably) owned by Holy Roman Emperor Rudolf II
• Circa 1610-1620, it was (very probably) owned by Rudolf II’s “Imperial Distiller” Jacobus z Tepenecz
• Circa 1630-1645, it was owned by (otherwise unknown) German Bohemian alchemist Georg Baresch
• Circa 1645-1665, it was owned by Johannes Marcus Marci of Cronland, who gave it to Athanasius Kircher
• For the next few centuries, it was (almost certainly) owned by Jesuits & moved around Europe
• In 1912, it was bought (probably for peanuts) by dodgy antiquarian book dealer Wilfrid Voynich
• He bequeathed it to his wife Ethel, who bequeathed it to Anne Nill, who sold it to H. P. Kraus in 1961
• In 1969, Kraus donated it to Yale University’s Beinecke Rare Book & Manuscript Library

However, before 1600 things quickly get murky, to the point that the list of “very probably true” things we can say about the Voynich Manuscript’s early art history is embarrassingly short:-

• Radiocarbon tests carried out in 2009 date itsvellum to between 1404 and 1438 with 95% certainty, though as yet there is no cast-iron proof that the text and drawings were added straight away
• The clear, upright handwriting is most often described as being reminiscent of either Carolingian minuscule (800-1200) or its Italian Quattrocento revival form, the “humanist hand” (circa 1400-1500) – the radiocarbon dating points to the latter
• Several of its drawings have parallel hatching (similar to Leonardo da Vinci’s); so it was probably made after 1410 if from Germany, after 1440 if from Florence, or after 1450 if from elsewhere
• Two owners have added writing in [what appear to be] fifteenth century hands; so it was probably made before 1500
• Some marginalia (in the zodiac section) appear to be in Occitan, where the spelling most resembles that known to be from Toulon; so it is probable that the manuscript spent some time in South West France
• There is strong codicological evidence that the current page order and binding order differ from the original i.e. that both the folio (leaf) numbers and quire (group) numbers were added at a later date
• A small number of the manuscript’s plant drawings do seem to depict actual plants (f2v has a water lily, for example), though most do not

It should be pretty clear that we have two quite separate types of historical data here – pre-1500 (codicological) and post-1600 (archival) – with no obvious way of crossing the roughly century-long gap between them.

My opinion (which you can take or leave) is that if we put more palaeographic effort into reading the VMs’ marginalia, we would very probably improve on this unsatisfactory situation. For example, I believe that the top line of f116v says [something like] “por le bon simon s(int)…“, and that this was possibly even written by the original author. Furthermore, I suspect taht some of the ‘chicken scratch’ marginalia may be ink blots saying “Simon”, and that these were added in the middle of the 15th century, near the start of the VMs’ life. But who was this ”Simon”?

Putting all the wobbly factuality to one side, this VMs account would be woefully incomplete if it failed to mention the sheer intellectual romance of such a mystery-filled mega-object, the tragi-comedy of all the mad theories surrounding it, let alone the blood-spattered trail of ruined reputations and wasted lives dripping behind this inscrutable “Sphinx”. For centuries, it has acted as a blank screen for numerous people to project their (often somewhat demented) historical / cryptological / novelistic fantasies onto, or if not that then an academic cliff to throw their hard-earned reputation over: yet recently there are signs that a few people are (at long last) starting to look at the VMs with (relatively) clear eyes. (Better late than never, I suppose!)

Arguably the biggest question to face up to is this: when people try to understand the VMs, why does it all go so wrong? I suspect that the confusion arises from the central paradox of the Voynich Manuscript – the way that its text resembles some unknown (perhaps lost, secret, or private) simple language while simultaneously exhibiting many of the properties you might expect to see of a complex ciphertext (i.e. an enciphered text). Any proposed explanation should therefore not only bridge the century-long historical gap, but also demonstrate why the VMs appears both ‘language-y’ and ‘cipher-y’ at the same time.

To illustrate this, here are some practical examples of the way Voynichese letters ‘dance’ to a tricky set of structural rules. Individual letter-shapes frequently occur…

• …as the first letter of a page (e.g. the ornate “gallows” letters, EVA “t”, “k”, “p”, “f”)
• …as the first letter of a paragraph (e.g. EVA “t”, “k”, “p”, “f”)
• …as the first letter of a line (e.g. EVA “s”)
• …as the last letter of a line (e.g. EVA “m” or “am”)
• …as the first letter of a word (e.g. EVA “qo”)
• …as the last letter of a word (e.g. EVA “y” or “dy”)
• …as separated pairs on the top line of a page (e.g. EVA “p” or “f”)
• …as a paired letter (e.g. EVA “ol”, “or”, “al”, “ar”)
• …unrepeated, except in EVA “ee” / “eee” / “ii” / “iii” sets.

…and so on. From a code-breaker’s point of view, this basically rules out Renaissance polyalphabetic ciphers, because they use multiple alphabets (or offsets into alphabets) to destroy the outward signs of internal structure – and what we see here has even more signs of internal structure than normal languages. Yet just to be confusing, some of the letter-shapes resemble shorthand both in their shape and their apparent positioning within words.

So… is ‘Voynichese’ a language, a shorthand, a cipher, or perhaps some carefully-orchestrated jumble of all three? Right now, nobody can say – but perhaps it is this ‘hard-to-pin-down-ness’ that has managed to keep the Voynich’s mystery alive for all this time. Once you can appreciate that Voynichese is almost the opposite of chaotic – that its absence of randomness is possibly its most remarkable aspect – but yet none of the many visible patterns seem to help us decrypt it, you’ll perhaps begin your own journey into its mystery. Enjoy!