I’ve just watched the National Geographic / Naked Science documentary on the Voynich Manuscript, courtesy of a Stateside friend (thanks!). Regular Cipher Mysteries readers will already know how my review of it is supposed to go – ‘that, despite a few inaccuracies, it was great to see the Voynich Manuscript being brought to a popular audience‘.

But actually, the whole thing made me utterly furious: it was like watching yourself being airbrushed out of a family photograph. Let me get this straight: I researched the history like crazy, reasoned my way to the mid-15th century, stuck my neck out by writing the first properly new book on the Voynich for 30 years, talked with the documentary producers, sent lists of Voynich details for them to look at, got asked to fly out to Austria (though they later withdrew that at the last minute without explanation), kept confidences when asked, etc.

And then, once the film-makers got the radiocarbon dating in their hands, my Milan/Venice Averlino/Filarete theory became the last man standing (Voynich theory-wise). So why did it not get even a passing mention, when just before the end, they thought to edit in a map of Northern Italy with swallowtail-merloned castles and the narrator starts (apropros of nothing) to wonder what will be found in the archives “between Milan and Venice”. Perhaps I’m just being a bit shallow here, but that did feel particularly shabby on their part.

However pleased I am for Edith Sherwood that her Leonardo-made-the-Voynich-so-he-did nonsense merited both screentime and an angelic child actor pretending to be young Leonardo, the fact remains that it was guff before the radiocarbon dating (and arguably double guff afterwards): while much the same goes for all the Dee/Kelly hoax rubbish, which has accreted support more from its longstandingness than anything approaching evidence.

Perhaps the worst thing is that we’re all now supposed to bow down to the radiocarbon dating and start trawling the archives for candidates in the 1404-1438 timeframe. Yet even Rene Zandbergen himself has supplied the evidence for a pretty convincing terminus post quem: MS Vat Gr 1291 was completely unknown in Italy before being bought by Bartolomeo Malipiero as Bishop of Brescia, and so its stylistics could not sensibly have influenced the Voynich before 1457. In fact, 1465 – when the manuscript was carried from Brescia to Rome and became much better known – might even be a more sensible TPQ. And that’s without the cipher alphabet dating (post-1455 or so) and the parallel hatching dating (post-1440 if Florence, post-1450 if elsewhere in Italy).

And I’ll leave you with another thought: a couple of seconds after hearing the Beinecke’s Paula Zyats say “I don’t see any corrections”, the following image got edited in – a part of the f17r marginalia that looks to my eyes precisely like an emendation.

Voynich Manuscript f17r marginalia

Really, what am I supposed to think? *sigh*

The last few days have seen a huge flurry of Voynich-related Internet interest: a University of Arizona press release on radiocarbon testing the Voynich Manuscript’s vellum sparked a Discovery News item and a hundred or more slightly-edited reposts, with even Fox News getting in on the act yesterday. Sadly, though, it’s all the “same old same old”: you’d be forgiven for concluding that – the UofA’s whizzy new dating aside – there has been nothing new under the Voynich research sun since Mary D’Imperio’s (1976) “An Elegant Enigma”.

But actually, we know a huge amount now. For example, I think I can now prove to the satisfaction of 99% of historians that the Voynich Manuscript is indeed a ciphertext. Even though I first described the following in 2006 (The Curse of the Voynich, Chapter 10: “Secret Numbers”), nobody seems to have picked up on it as a basic historical proof.

So here’s my argument why it’s a ciphertext, one rock-solid step at a time.

(1) Medieval page references.

To make a medieval book, you fold a small double-width set of vellum or paper leaves around a single central line: this unbound-gathered-together nested set is called a gathering. Then, you pass off a set of gatherings to a binder, each gathering with its own sequential mark – typically ‘a’, ‘b’, ‘c’, etc – so that you can be sure the binder will bind them together in the right order. Once sewn and bound, the gatherings become known as quires. Within each quire, the individual folios (leaves) are typically numbered ‘i’, ‘ii’, ‘iii’, etc, while the two sides of each folio were referred to as recto (front) and verso (back), usually abbreviated to ‘r’ and ‘v’ respectively.

So, when medieval writers wanted to refer to individual pages, the ‘address’ of that page was its quire index (a/b/c/…), its folio number (i/ii/iii/…) and its side (r/v). For convenience, these three elements were usually fused together into a short string, i.e. the first few pages of the first (i.e. the a-th) quire would be referred to as air, aiv, aiir, aiiv, aiiir, aiiiv, etc. Everyone who worked with manuscripts knew these were medieval page references: this was the cultural norm.

(2) Voynichese appears to be full of medieval page references.

For someone circa 1450 looking at the Voynich Manuscript, one single feature of its mysterious text would have almost jumped off the page – the medieval page references, all apparently to the first (a-th) quire. For example, here you can see an “aiiv” group with an “air” group immediately below it:-

Whether you like it or not, there it is: the Voynich is utterly stuffed full of these medieval page references. However…

(3) These don’t function like medieval page references.

If you count these up, you discover that aiv/aiiv/aiiiv (1675/3742/106) occur more than eight times as often as air/aiir/aiiir (564/112/1). Simply put, the stats are wrong. Also, there appear to be no references to other quires, just to the a-th quire. 

(4) These are not medieval page references

So, if these things so closely resemble medieval page references (but aren’t), what are they? The famous WWII codebreaker Brigadier John Tiltman wondered whether these might actually some kind of obfuscated Roman numeral scheme, but the stats were wrong for that idea too. Hence they don’t match any obvious numerical or referential scheme.

(5) These appear to be integrated within the text

Even if we cannot directly read them, we can see that they recur throughout the text, and – as in the example shown above – that they are often integrated with other Voynichese letters into words. But… if they appear as consistent groups, are integrated within words, and resemble something that they clearly aren’t (i.e. medieval page references), then the only rational historical inference is that…

(6) These are cipher shapes.

QED. But… if even one element of the writing scheme is written in cipher, then the conclusion has to be that…

(7) The Voynich Manuscript is a ciphertext.

Again, QED. However, even though medieval page references were known circa 1450, their heyday had passed: they were far more part of the medieval world of the monastery scriptorium than the emerging Renaissance world. Hence the presence of fake medieval page references in its ciphertext inexorably leads (I believe) to a further curious conclusion:

(8) Voynichese was constructed to resemble a medieval document in a fake archaic language

I contend that this is the only logical conclusion consistent with the radiocarbon dating: an early Renaissance mind looking back at the Middle Ages, cunningly appropriating medieval textual tropes for his/her ciphertext’s alphabet.

Now, please tell me – which part of this argument do you disagree with?

A few days ago, I posted here about my dissatisfaction with the current Wikipedia page on the Voynich manuscript, and about the kind of changes I proposed should be done to make it a better ‘shopfront’ for Voynich research. Basically, my idea was to move all the speculative theory-filled discussion across from the main Wikipedia page over to two or more new pages, so that the main page could instead focus on what the mysterious manuscript actually is (rather than on what it possibly might be). Separating the subjective from the objective in this way would, I think, be a much better way of presenting Voynich-related information, which tends to suffer from conceptual sprawl and lack of focus.

Since then, I’ve been working on a new ‘Voynich theories’ page as a container for a good number of the various historical theories surrounding the Voynich manuscript. What quickly became apparent to me here was that this similarly requires a structured way of presenting this information to visually separate each individual theory from the critique of that theory. Thinking about it all, it may be even better to separate the discussion of each theory into three distinct parts:

  • the key supporting evidence (that which is true)
  • the speculative theory itself (that which is proposed or inferred)
  • the main critiques of the theory (the key problems with the theory)

So, here’s a link to my first draft ‘Voynich theories’ Wikipedia page – what do you think of it? I haven’t even tried to complete the citations and references yet, I’d really like to get some initial reactions before I develop it further or even (*gasp*) add it into Wikipedia. Please leave your comments here, thanks!

OK, so it’s not the full radiocarbon paper we’re all (still) waiting for. But here’s a new webpage with a little bit more Voynich-related information from Greg Hodgins at the University of Arizona. Hodgins notes that dating the inks would be nice too (but sadly that’s not really possible): and because he scalpeled vellum slivers from the folio edges, there was the strong possibility that “a lot of finger oils adsorbed over time” into those slivers, which could easily throw the dating process off if they weren’t cleaned thoroughly.

Yes, the manuscript itself might once again have somehow managed to give even the most assiduous of researchers the finger. Now wouldn’t that be a surprise?

Last week  (3rd February 2011) saw the US premiere of “The Book That Can’t Be Read”, the long-awaited National Geographic channel airing of the recent ORF documentary on the Voynich Manuscript. Though it prominently features the benign beardiness of everyone’s favourite Voynich expert Rene Zandbergen, for a pleasant change the star of the show is undoubtedly the manuscript itself, with the Beinecke Rare Book & Manuscript Library’s acquiescence to radiocarbon dating of the vellum the shining jewel in the Austrian documentary makers’ crown. If you missed it, it’s showing again shortly (10th February 2011, 4PM): it’s a fairly up-to-the-minute introduction to the VMs, so you should definitely fetch a mid-sized bag of toffee popcorn and settle down on your sofa for this one.

Interestingly, I don’t know if they significantly re-edited the programme for an American audience, but I was pleased – no: delighted, actually – to see some scans of the manuscript the film researchers had taken dotted among the set of low resolution Voynich promo photos on the NatGeo webpage (all © ORF). For example, slide #11 has the infamous erased signature on f1r (the frontmost page of the manuscript), which – with a bit of low-impact Gimp-fu – looks like this:-

Voynich Manuscript f1r, "Jacobj a Tepenece" signature, uv, enhanced

Voynich Manuscript f1r, erased signature

Having long ago slaved to produce not-quite-as-good versions of this from the RGB scans, it’s a pleasure to finally see this in its non-visible ultra-violet glory: to my eyes, it reads “Jacobj à Tepenece / Prag”, but I’ll happily defer to palaeographers working from higher resolution scans.

Slide #12 contains another UV scan, this time of my personal favourite piece of Voynich marginalia – the tiny letters at the top of f17r. Despite its ridiculously low resolution, what should be clear from the image (again, slightly Gimp-enhanced) is that the Voynich letters at the end (“oteeeol aim”, as per The Curse of the Voynich pp.24-25, 30) are an integral part of the writing, just as I claimed when I first saw them in 2006. The point being that if you accept that, then it becomes very likely that this and (by implication) the “michiton” marginalia on the end page were added not by a later owner, but by the encipherer of the VMs himself/herself. All fascinating stuff that, in my opinion, cuts deep to the heart of the VMs’ historical nature, but I’d be a little surprised if the documentary has been edited to cover it.

Voynich Manuscript, f17r marginalia, uv, enhanced

The "meilhor aller" marginalia from f17r of the Voynich Manuscript

Finally, the last photo of immediate interest to Voynich researchers is slide #13, which shows a close-up of the exposed quire bindings (i.e. with the manuscript’s cover partially removed). This kind of view offers a lot of information that you can’t normally see, because the bifolios are so firmly bound together that you can’t get at all close to the sewing holes in the spine of each quire – which is good for conservation, but bad for codicology.

Voynich Manuscript binding, close-up

View of the Voynich Manuscript's binding

Here, the features that particularly intrigue me are the faint writing on the inside cover (bottom left arrow); the non-continuous line of marks across the quire spines (mid-right arrow); and the many redundant sewing stations (needle holes from earlier bindings, indicated by short red underlines). These inexorably point to the manuscript’s complex reordering and rebinding history, i.e. where its quires and bifolios have danced a complicated quadrille over time to end up in their final order. What I don’t really understand is why codicologists don’t have entire conferences devoted to the Voynich Manuscript, because to my eyes it is surely the Everest of codicology – a complex, multi-layered artefact whose secret inner history can only practically be revealed through prolonged, collaborative, non-textual forensic analysis. And yet it’s only me who seems to have published anything substantial on it!

Anyway, set your PVRs to stun record and let me know what you think of the Naked Science documentary. Hopefully the documentary makers will now celebrate the occasion by releasing more information,data and photos on the Voynich Manuscript that they took during their research (hint: high quality versions of the above three images would be a very good start)!

A few days ago, I was emailed by Gerry Scott from Cornell who recently, with the help of a friend, started putting together his own Wiki (a set of webpages editable by anybody) to try to give structure to the seething mass / wobbly jelly that is Voynich Manuscript research. Here’s a direct link to what they’ve done so far.

One of the nice things about this is that Gerry has tried to take my (many) criticisms of the Wikipedia page on board, and so has consciously…

tried to segregate facts and speculation. The wiki includes separate sections for textual, linguistic, provenience [sic], and art-historical research, and uses distinct “theory” and “fact” subsections within each section.

He’s aiming high, which is admirable: but it has to be pointed out that the challenge involved – basically, building an online ‘Encyclopaedia Voynichiana’ – is nothing short of gargantuan. It’s at least a decade’s work, and with Wilfrid Voynich’s 2012 centenary looming, we only really have a year before the next tsunami of documentaries hits our virtual shores.

Personally, I think there’s a better way: fix the Wikipedia entry. It’s the #1 resource served up by just about every Voynich-related web-search, as well as the #1 link given by just about every inane half-troll writing up their own gee-whiz account of the VMs: whether we like it or not, it’s going to remain the public face of Voynichology for quite some time yet.

The problem is that it’s, well, pants – it’s overlong, overcondensed, underreadable, and a reader coming to the topic fresh doesn’t really leave the page any the wiser. Structurally, the page’s core problem is that it has no clear distinction between facts, evidence, observations, hypotheses and suppositions: at the same time, over time its text has expanded to about 55K, which is just about the right point to start splitting it up into smaller, more useful pages. But how should it be split?

Personally, I think the content has been squeezed out by a barrage of meta-content – most of the text now seems to be taken up with theories about theories. Honestly: the moment any Wikipedia page fixates so heavily on theories that the thing itself gets lost, something has gone badly wrong.

But what to replace it with? I think there should be a guiding strategic principle in play: no theories on the root page, just facts and evidence. Furthermore, I’d split it up so that Voynich theories (Bacon, Filarete, Leonardo, Ascham, Dee/Kelley and, errrm, Bacon again, etc), Voynich meta-theories (hoax, glossolalia, exotic language, artificial language, hybrid language, shorthand, ciphers, etc), and Voynich history/provenance each have their own page. Which is not to say that those topics are not interesting in their own right: but rather that they’re secondary topics, and not essential to building up a primary understanding of the object itself.

At this point, some might say… “but take away all that stuff, and what would be left?” Actually, I think a surprisingly large amount would remain, pretty much all of which is what people new to the VMs primarily want to know about.

The Wikipedia page is the shopfront to our community and our research, and it’s not serving us well… so it’s time to fix it. If you would like to have a say in what happens next, please join in the debate on the Voynich Wikipedia talk page, or just leave a comment here.

I really don’t know how I managed not to pick up on it, but last year a group of German artists put on a VMs-themed installation at the Grauerhof in Aschersleben entitled “DAS VOYNICH MANUSKRIPT: eine künstlersicht auf ein rätsel” (an artist’s view of a mystery), featuring pieces by Rüdiger Giebler, Moritz Götze, Olaf Holzapfel, Alicja Kwade, Daniel Lergon / Gregory Carlock, Via Lewandowsky, Johannes Nagel, Jorinde Voigt, and Ralf Ziervogel. If you go to the site, clicking on any of the pictures launches a pop-up 32-slide slideshow tour of the exhibition, which is rather nice.

I particularly like Lergon and Carlock’s ‘book object’, with its spurious botany and implausible fold-out page arrangement. But perhaps the standout contemporary art piece of the show was by Berlin-based Via Lewandowsky (1963-) called “Okay“, formed of the Voynichese letters spelling ‘okay’ (in EVA) in striking green neon.

If you want to see ‘Okay’ for yourself, it’s currently on display at the Galerie Karin Sachs in Munich until the 3rd March 2011 as part of a show of Lewandowsky’s work called “Archäologie der Ähnlichkeit“.

Here’s a bit of fun for you that’s only running for a few more days: a Voynichese-style challenge cipher courtesy of everyone’s favourite hirsute cipher reclusive Tony Gaffney. Here it is (click on it for a more detailed image):-

Tony Gaffney challenge cipher

He says:-

The above could almost be a missing page from the VM. If anyone cares to have a go at deciphering it, it is the start of a very well known Italian story – the plaintext is Italian and it reads left-to-right and top-to-bottom in the normal fashion.

What kind of cipher is it?

Here’s a basic transcription (into EVA) to get you going, assuming that is indeed a genuine cipher:-

p aiin deey eedy lched otoched r qochedchedy aiin eedy chedeed otoeed ch
qochedy lched otochedy chy cthey dchedy siin chdy daiin otoch dcthey cthey
otochedaiin eedy qo otod aiin lched eedy lched otochedoto qochedaiin etey
qochedee dotochedy otoaiiny daiin otocheds daiin eedy chedeed eeds qochedy
aiin eeotochedy
p l daiiny lched otochedy s eedr r otochedchedeed dy eey chedeed daiin eey
dchched dcthey otochedchedeed dy lched eedoto qochedy oto dotochedy eey
eedaiin eedy otoaiiny daiin otoch dcthey cthey otochedaiin eedy qoched siin
eedaiin s otoched
p chedeed eedaiiny r qochedy s otoched qochedy eey chedeed dch l qochedy
lched otochedy m otol l dy qochedy l daiiny daiin dchedy aiin dcthey
eedy otoched ry oto qochedeey cthey otochedcthey otochedaiin eedy
lched otochedy chedeed dch cthey dy chedeed eedr eedch dcthey eedy 
p chched dch eedy otoched lched otochedeey l qochedcthey cthey otochedy
dy chedeed cthyched otochedy eey dch dy chedeed dcthet cthey otoched
oto eedy qochedy eey dee eedy daiin oto eedaiin eedy chedeed eedaiiny
chedeed cthyched otoched s otochedt lched otoched chedeed qoeeed odaiin ch
doto eed 
            *          *          *          *          *          *
            |          |          |          |          |          |

Enjoy! 🙂

Though technically they’re probably not in cipher (rather, they’re almost certainly three wobbly dictionary codes), they definitely form an historical mystery: and even today, the Beale Papers’ promise of 19th century treasure continues to inspire people to borrow a distant cousin’s mini-diggers and covertly dig implausible holes not too far from where Buford’s Tavern once stood. Which is, of course, both foolish and most likely illegal, so don’t expect me to condone anything like that for a microsecond.

What I’m far happier to praise is Andrew S. Allen’s animation “The Thomas Beale Cipher”, which I’ve already mentioned a few times along the way. Anyway, now that its tour of independent film festivals is (presumably) over, Allen’s very generously placed a copy of his film on the web right here for all to see (but expand it to full screen for best effect). You should be pleased to hear it doesn’t offer a faux solution (how gauche that would be) or even the pretense of a clunky explanation, but just the lightest touch of 1940s G-man cryptological paranoia amidst a glorious barrage of vintage textiles. Oh, and a nice brass-section soundtrack too. Go and have a look: I think you’ll like it a lot! 🙂