Category Archives: Voynich Manuscript

I’ve recently been researching 15th century copies of Johannes Hartlieb’s German translation of Andrea Capellanus’ “De Amore“. My plan is to try to work out if any includes a predecessor of the hand-crossing drawing that appeared in the three 1482/1484 incunabula…

…which, if you recall, is the drawing that Koen Gheuens cleverly suggested might well be linked with a Diebold Lauber workshop drawing and the Voynich Manuscript’s Gemini zodiac roundel…

This is all going OK so far (I now have Alfred Karnein’s magisterial book on the subject, and a copy of his 1985 book should land on my doorstep soon), and as always I’ll post more on this in due course.

However, there’s one other German “De Amore” described by Frank Fürbeth in his more recent book “Johannes Hartlieb: Untersuchungen zu Leben und Werk” that I can’t find anything about. This manuscript, which doesn’t appear in either of Karnein’s books, was (says Fürbeth) Number 3 in American antiquarian bookseller Philip J. Pirages’ 1985 catalogue. But when I emailed the bookseller, Phil Pirages himself kindly replied, saying that he had no record of any such book.

It would seem that something a little odd is going on here. 🙁

Can I therefore please ask any Cipher Mysteries reader who just happens to have easy access to a stupendously good academic library with a copy of Frank Fürbeth’s book “Johannes Hartlieb: Untersuchungen zu Leben und Werk” (currently £80+ on bookfinder.com, somewhat out of my range, sadly) to photograph or scan pages 62 and 63 for me? (This is, according to Google, where Fürbeth discusses the Pirages manuscript.) Thanks very much!

Note: I believe that the 2011 edition of the book is simply a reprint of the 1992 original, but please check to see if these two pages do mention Philip J. Pirages, thanks!

There are now, courtesy of Koen Gheuens and others, numerous web pages exploring possible / probable connections between the Voynich Manuscript’s zodiac roundels and 15th century scribal workshops in German Alsace (most notably that of Diebold Lauber).

In one of my own contributions to this (small) canon, I discussed the McKell calendar, an astrological / medical calendar made in Hagenau between 1430 and 1450 by Lauber’s workshop (it has more recently been dated as c. 1445). Though I only knew of black and white scans online, commenter Helmut Winkler very kindly posted up a link to a webpage from BNU Strasbourg, the institution that recently bought it.

This included colour scans of the pages, including the gold-leaf sun/moon framing device at the top of the medical pages, which surely makes it clear what a top-end manuscript this must have been:

The colours too turn out to be remarkably vibrant, notably the red and blue clothes and gold wings on the calendar’s page for August (Virgo):

The McKell Aries

Though the McKell Aries page scan is now (apparently) missing from the website, I found a copy of it online that Darren Worley had posted in a comment to Stephen Bax’s website. Note that the Aries page’s tree is drawn and coloured very similarly to the Virgo page’s tree (above):

Can you spot anything wrong with this image? Having once lived next door to a goat for a few months (who would happily eat your washing given half a chance), I have to say that this looks to me less like a sheep or a ram than a greedy goat with a goatee doing what goats do best (i.e. climbing up to eat anything it can sink its teeth into).

But to be clear, the three simplest ways to tell a sheep from a goat are:

1. tail direction (goats’ tails go up, while sheep’s tails go down)
2. what they’re doing (sheep usually graze grass, while goats tend to prefer munching everything else)
3. goats stereotypically have goatees (the clue’s in the name)

So even though the McKell Aries illustrator got the tail direction right for sheep, I’m still happily scoring this 2-1 in favour of the goats.

The Voynich Aries

As Prof. Ewa Sniezynska-Stolot wrote to Rafal Prinke in 2001, having examined the Voynich Manuscript: “The genre scenes, eg. Aries eating a bush, suggest that the signs were redrawn from a calendar”. And just as with the McKell calendar’s Aries, it has long been argued that the animal depicted is less like a ram or sheep than, as Albert Howard Carter seems to have first pointed out back in 1946, a goat.

Moreover, there are two of the same animals on consecutive pages, firstly a dark Aries…

…followed by a light Aries…

Incidentally, on the VMS list in 2004 Pamela Richards argued that this must be a goat because “[s]heep don’t have dew claws, those tiny hard horns above the hooves; goats do. And those dew claws are very clearly depicted on each foot.”. However, as Rene Zandbergen pointed out much later, sheep actually have dew claws too (though horses and giraffes don’t, so please be reassured that we can at least rule them out), so this isn’t a great argument.

So… is this actually a goat? By my (albeit simplified) scoring system in the preceding section (tail down, no goatee, but grazing), I ought really to instead score this 2-1 to Sheep United. But it’s a game of two halves, I’m sick as a parrot, the seagulls follow the trawler, etc etc.

Model Books and Calendars

In many ways, though, I think it doesn’t greatly matter if either/both is/are a sheep or a goat, because I think we can tell broadly what happened here.

The McKell Aries was (I believe) most likely copied from a previously made goat drawing exemplar, probably from a model book. And perhaps the artist straightened the tail to make it better resemble a sheep, who knows? I also think that the McKell Virgo tree was copied from the same goat picture (the tree was surely integral to the goat design).

Similarly, I think the Voynich Aries was almost certainly copied from an Aries roundel in an existing calendar. Perhaps the score in that original illustration would have been scored 2-1 to the goats or 1-2 to the sheep after extra time, it’s almost futile to try to say.

Incidentally, I have a recollection of Rene Zandbergen once pointing to an Aries calendar roundel where there was a tree in the background that was almost like an optical illusion of something being eaten by the animal in the foreground. But I am unable to dig that up from anywhere, sorry. 🙁

Goats in the Buch der Natur

Finally, Ulrike Spyra’s book might once again be an interesting resource here, because her Synoptic Table of Illustrations lists (on p.384) a number of drawings of goats (“Gaiz / Capra“):

• M590 – 61rb (Munchen, BSB, Cgm 590)
• A497 – 115va (Augsburg, SuStB, 2o Cod. 497)
• GW – 66ra (Gottweig, Stiftsbibl., Cod. 389 rot)
• SG – 69rb (Strasbourg BNU Cod. 2264)
• HD311 – 79v (Heidelberg UB Cpg 311)
• M684 – 84r (Michelstadt, Nic.-Matz-Bibl. Cod D 684)
• WU – 67r Wurzburg, UB M ch f 265

As an example, here’s Strasbourg BNU Cod. 2264’s goat, which I’d say scores a comfortable 2-0 win:

The small point I’m making here is that 15th century artists were clearly more than capable of drawing goats in a semi-realistic way if they so chose.

Given that I couldn’t find any page on the Internet providing links to scans of alchemical herbals, I thought it would be good to try to fill that gap. Not as many as I hoped turned out to be fully accessible, but there are still a good number.

The basic list I used was the one given by Philip Neal (summarizing Vera Segre Rutz, of course), but extended to include the manuscripts discussed by Alexandra Marracini in her “Asphalt and Bitumen, Sodom and Gomorrah: Placing Yale’s Voynich Manuscript on the Herbal Timeline” paper.

Segre Rutz’s & Marraccini’s trees

Before I begin, it’s important to remember that Vera Segre Rutz reconstructed the cladistic tree of alchemical herbal manuscripts: this framework has dominated discussion of alchemical herbals ever since. The root of this tree was a Manuscript X (now lost), which begat Manuscripts Y and Z (both of which are also now lost). All the “direct tradition” manuscripts derived from Manuscript Y or Manuscript Z.

In addition, here’s Alexandra Marraccini’s tree, which laid out where she thought the Voynich Manuscript sat in relation to alchemical herbal manuscripts. Note that the top part of the tree is Segre Rutz’s direct tradition, while the bottom (yellow) part contains Marraccini’s proposed two groups of derived manuscripts:

Manuscripts in Segre Rutz’s Direct Tradition

The four Z-family alchemical herbals are:

• Fermo, Biblioteca Comunale MS 18 (2 pages on YouTube?, 2 more pages here)
• Florence, Biblioteca di Botanica dell’Universita MS 106 – book by Stefania Ragazzini (as Rene Zandbergen has pointed out, this ms has a simple cipher key on fol. 1r)
• Oxford, Bodleian Library, MS Canon. Misc. 408 (166 plants, Latin)
• Pavia, Biblioteca Universitaria MS Aldini 211 (this is the herbal that Segre Rutz wrote “Il Giardino Magico degli alchimisti” about)

The four Y-family alchemical herbals are:

• Paris, Bibliotheque Nationale, MS Lat. 17844
• Paris, Bibliotheque Nationale, MS Lat. 17848
• Vicenza, Biblioteca Bertoliana MS 362 (formerly “MS G.23.2.3”) (Image of the “cleaver and sword” herbal page) (not yet here, sadly)

Manuscripts in Segre Rutz’s Indirect Tradition

• Bologna, Biblioteca Universitaria MS Aldrovandi 151(1)
• Bologna, Biblioteca Universitaria MS Aldrovandi 151(2)
• Bologna, Biblioteca Universitaria MS Aldrovandi 152 (1550-1605)
• Bologna, Biblioteca Universitaria MS Aldrovandi 153 – catalogue entry
• Brescia, Biblioteca Queriniana MS B.V.24 – a page discussed by Marco Ponzi
• Florence, Biblioteca Medicea Laurenziana Ashb. 456 – catalogue description
• Florence, Biblioteca Marucelliana MS 168/C – discussed in the book “I Segreti della medicina verde nell’epoca medicea, da due manoscritti inediti della città di Firenze : (secoli XV e XVI)“
• London, Wellcome Historical Medical Library MS 261 – catalogue entry
• London, Wellcome Historical Medical Library MS 334 – catalogue entry – ‘A contemporary copy of a famous MS herbal preserved at [the] Laurentian Library in Florence. Bought for 700 fr. by Woynich [Voynich] 1912‘.
• London, Wellcome Historical Medical Library MS 337 – catalogue entry
• New York, Pierpont Morgan Library MS 22.222 – I believe this is actually “PML 22222.4” – catalogue entry – “Text derives from an herbal in Pavia, see Bühler, “An anonymous Latin herbal.“
• Oxford, Bodleian Library MS Add. A. 23
• Padua, Biblioteca Universitaria MS 604 – manuscript reference looks incorrect (see here)
• Paris, Bibliotheque Nationale, MS Hebr 1199
• Trent, Museo Provinciale d’Arte MS 1591 – some images are online here
• Venice, Biblioteca Nazionale Marciana, MS It. III.11 (MS 5004)
• Venice, Biblioteca Nazionale Marciana, MS It II.12 (MS 4936)

Tractatus de Herbis Tradition

The herbals forming the “Tractatus de Herbis” tradition deriving from Firenze MS 106 are divided by Marraccini into two groups. Firstly, the group she calls the “Non Flattened Asphaltum” group:

• Vatican, Biblioteca Apostolica Vaticana, Chig F VIII. 188 (not yet digitized)
• London, British Library, Egerton MS 747

There’s also a late (1450) copy of Egerton MS 747 which commenter bi3mw thought should be included here:

• Basel, Universitätsbibliothek, Kräuterbuch “Circa Instans” K II 11

Secondly, the “Flattened Asphaltum” group (which Marraccini believes may well include the Voynich Manuscript):

• London, British Library, MS Sloane 4016 – this is described in the catalogue entry as “An Italian Herbal, classified by Baumann as one of the ‘North Italian group’ and as a copy of Paris, Bibliothèque de l’Ecole des Beaux-Arts, MS Masson 116 (see Baumann, Das Erbario Carrarese, 1974).“
• Vatican, Biblioteca Apostolica Vaticana, Chig F. VII. 158
• Paris, Bibliotheque Nationale, MS Lat. 6823

To this group I expect we should also add the herbal from which MS Sloane 4016 was copied (according to some, though Alain Touwaide vigorously disagrees):

• Paris, ENSBA, Ms.Mas 0116

…and also a closely related herbal manuscript that Rene Zandbergen thinks gets somewhat unfairly overlooked:

• Munich, BSB Clm 28531

Over the years, a small drawing on page 80v of the Voynich Manuscript has triggered what can only sensibly be described as all manner of Unholy Theory Wars.

This drawing has been used as ‘definitive’ proof of the manuscript’s supposed New World origins (i.e. because it is ‘obviously’ an armadillo); or that the manuscript was faked (i.e. according to Rich SantaColoma over at Koen’s blog, “Everybody who was anybody [in the 17th century] had a stuffed armadillo hanging in their kunstkammer“, which may yet turn out to be the Voynich comment of the year); and so forth, endlessly.

Whatever the drawing is actually supposed to represent, all we know for sure is that it appears in a quire / section of the manuscript that seems to be almost entirely related to different aspects of water (there are baths, bathing nymphs, showers, fountains, pipes, and even a rainbow in there). Hence it has long seemed highly likely to me that this will turn out to represent an animal somehow connected to water.

Back in 2009, I suggested as one possible candidate the catoblepas, which nowadays gets far more online attention as a fairly second-rate Neutral Evil Dungeons & Dragons monster (it later became known as a nekrozon, D&D trivia fans), or as a recurring enemy in the Final Fantasy series (it’s called a ‘Shoat’ there, though that actually means ‘piglet’, FF trivia fans) than as an actual mythical beast. Errrm… if you know what I mean.

As an aside, I can’t help but pass on that in Janick and Tucker’s (2018) “Unraveling the Voynich Codex” (whose nutty New World Voynich theory – naturally – relies on this being an armadillo), they mention my 2009 catoblepas page: and on p.360 describe me sweetly as “one of the most expansive and intemperate of bloggers”. Which they can, of course, stick right up their hairy arses, bless them. (Happy now? Good. So let’s move swiftly on.)

But it turned out that Andrew Sweeney had first suggested the catoblepas on the old VMs mailing list back in 2004. Hence I thought it was now time to revisit the entire secret history of the catoblepas, and see what I could find…

Pliny the Elder on the Catoblepas

Our first source is Pliny the Elder (AD 23–79), the well-known Roman writer and naturalist, who is also famous for having died trying to rescue friends and family from Pompeii following the eruption of Vesuvius. He describes the Catoblepas (named from the Greek Κατωβλεψ, ‘looking down’) in his 37-book Naturalis Historia as follows:

In Western Aethiopia [Ethiopia, i.e. West Africa] there is a spring, the Nigris, which most people have supposed to be the source of the Nile… In its neighbourhood there is an animal called the Catoblepas, in other respects of moderate size and inactive with the rest of its limbs, only with a very heavy head which it carries with difficulty — it is always hanging down to the ground; otherwise it is deadly to the human race, as all who see its eyes expire immediately.

Pliny the Elder, Natural History 8. 77 (trans. Rackham)

Aelian on the Catoblepas

When Claudius Aelianus (c. 175 – c. 235 AD) wrote his 17-book encyclopaedia of nature, Pliny the Elder’s book was one of many he shamelessly recycled in his own ‘honey-tongued’ prose. So it should be no great surprise that the account of the catoblepas we find there is little more than an elaborated reworking of Pliny’s account:

Libya [Africa] […] produces the animal called the Katobleps [Catoblepas]. In appearance it is about the size of a bull, but it has a grimmer expression, for its eyebrows are high and shaggy, and the eyes beneath are not large like those of oxen but narrower and bloodshot. And they do not look straight ahead but down on to the ground: that is why it is called ‘down-looking’. And a mane that begins on the crown of its head and resembles horsehair, falls over its forehead covering its face, which makes it more terrifying when one meets it. And it feeds upon poisonous roots. When it glares like a bull it immediately shudders and raises its mane, and when this has risen erect and the lips about its mouth are bared, it emits from its throat pungent and foul-smelling breath, so that the whole air overhead is infected, and any animals that approach and inhale it are grievously afflicted, lose their voice, and are seized with fatal convulsions. This beast is conscious of its power; and other animals know it too and flee from it as far away as they can.

Aelian, On Animals 7. 6 (trans. Scholfield) (Greek natural history 2nd Century A.D.):

Thomas de Kent

From there, we fast forward to the Middle Ages and to Thomas de Kent, the Anglo-Norman author of the 12th century Roman de toute chevalerie, one of several Alexander-themed romances written at the time. There are several manuscript versions of his poem (listed on Arlima), of which Trinity College O. 9. 34. (made circa 1250) is online here. However, even though some monsters are depicted in the Trinity Ms, I don’t believe that any of them is a catoblepas, e.g the sea monster on f24r:

However, the BnF has a digitized copy of the (much more exciting-looking, particularly if you like ornate towers and nicely-coloured horseback battles with swords) 14th century Français 24364, which on fol.68r (in what seems to be an inserted section on mythical animals) has this enchanting image of a catoblepas using its +10 Eyes That Paralyze to kill some poor sap:

Durham’s copy (C. IV. 27 B) I had no luck finding at all, but perhaps others will do better than me.

To be honest, though, this set of manuscripts seems not to have formed the start of any long-running tradition. So this – unless you know better – probably marks the end of this particular line of manuscripts.

Thomas of Cantimpré (c.1200-c.1272)

The mainstream medieval reception of the catoblepas seems to begin with Vincent de Beauvais (1190-1264), who mentions it in Book XVIII of his Biblioteca mundi. However it is with Thomas of Cantimpré’s De naturis rerum that things start to get properly interesting.

Even though De naturis rerum was initially just text descriptions, adaptations and illustrations appeared in manuscript copies before very long. And these were then followed by incunabula and (of course) printed books. Hence Thomas of Cantimpré’s book is a lot like the long shadow of the Batcape that falls over Gotham’s seedy streets: by which I mean that just about everywhere we will find the catoblepas depicted from 1300 onwards, it will turn out to be either in a version of De naturis rerum, or in a book adapted from or strongly influenced by it.

In short, De naturis rerum is the source of the catoblepas Niger. And here’s the ‘cathapleba’ in Valenciennes 320, one of the earliest illustrated mss:

And here’s another (slightly later) one, Brugge MS 411 (1451-1500), found by Ger Hungerink:

Der Naturen Bloeme (ca.1350)

One book adapted (and abbreviated) from Thomas of Cantimpré’s De naturis rerum was Der Naturen Bloeme by Jacob van Maerlant (1230/35-ca.1291). There’s a 2011 edition by Herman Thys, and a detailed 2001 book on its reception by Amand Berteloot and Detlev Hellfaier: in Dutch circles, it’s quite a famous medieval manuscript, so there is plenty of academic literature on it out there if you’re interested.

It is in one of the illustrated copies of Der Naturen Bloeme that we find another early visual representation of the catoblepas. The KB KA 16 manuscript copy (ca.1350) contains miniatures of all the standard Mandevillean monstrosities (people with giant feet, people with no heads, people with two faces, etc), plus an odd-looking catoblepas on this page:

Cathaplebas is een dier
zeer vreselijk en onguur
en is op de Nijl, de rivier,
van de vreselijkste manieren.
Traag is het en niet bar groot.
De last heeft het zwaar ter nood
van zijn hoofd dat hem zwaar weegt.
Van deze beesten is het dat men zegt
komt het op je aan onvoorzien
en tussen de ogen ziet het je
dan ben je weg van het lijf.
Dit dier lijkt op een deel der wijven
die het hoofd dragen gehoornd zo zeer
dat het stinkt voor Onze Heer
en schijnt of het hen verwurgde
dan komt er een dwaas die onge
past op haar ziet en wordt zo gevangen
en van hart alzo ontdaan
dat hij ziel en lijf verliest
en de dood daarom kiest.
Van de c dat neemt hier een einde,
nu hoort wat ik van de d vindt.

In the KB KA 16 copy, Der Naturen Bloeme is preceded by a calendar for Utrecht with local saints’ days (but no Dutch Cisioianus). I should mention that KB KA 16 includes an illustrated zodiac (though no crossbows), plus plenty of marginal whimsy, such as this horny rabbit:

…which is nice.

Here’s another Maerlant catoblepas in the British Library, also found by Ger Hungerink:

Conrad von Megenberg’s “Buch Der Natur”

The other well-known book derived from Thomas of Cantimpré’s De naturis rerum is Conrad von Megenberg’s “Buch der Natur”, which I and Koen Gheuens have both discussed on our respective blogs.

The main reference work for this is Ulrika Spyra’s book “Das ‘Buch der Natur’ Konrads von Megenberg”, a magisterial tome sitting next to me which I have already mentioned here a fair few times. The index references “Cathapleba”, but there’s also Spyra’s extraordinarily helpful “5.2.1 Synoptische Tabelle der Illustrationen in den Buch der Natur Handschriften” (p.382). From this (p.385), we learn that illustrations of the “cathaphleba” are to be found in 68rb of GW (Göttweig, Stiftsbibl., Cod. 389 rot), 83v of HD311 (Heidelberg, UB, Cpg 311), and 87v of M684 (Michelstadt, Nic.Matz-Bibl., Cod. D 684).

Firstly, Heidelberg UB Cpg311 (1455-1460), because it’s easy to get to. 🙂 However, the surprise here is that the drawing on 83v (reproduced as Abb 35 in Spyra) actually depicts a cockatrice rather than a catoblepas, so isn’t a lot of use to us:

This is copied faithfully in Nurnberg GNM Hs. 16538, fol. 50r (Spyra’s Abb. 47), which is hence also no use to us. 🙂

If anyone can find a copy of 68rb of Göttweig, Stiftsbibl., Cod. 389 rot, or of 87v of Michelstadt, Nic.Matz-Bibl., Cod. D 684, please let me know.

Spyra also mentions (pp.304-305) Olim Erbach, Graflich Erbach-Erbach und Wartenberg-Rothische Rentkammer, Cod. cart. ohne Signatur /Mscr. Nr 2. This has (she says) a Cathehaba on fol 50r.

Leonardo da Vinci on the Catoblepas

Leonardo briefly mentions the catoblepas in his Notebooks, though editors have noted that this derives from Pliny rather than from Aelianus:

CATOBLEPAS.

It is found in Ethiopia near to the source Nigricapo. It is not a very large animal, is sluggish in all its parts, and its head is so large that it carries it with difficulty, in such wise that it always droops towards the ground; otherwise it would be a great pest to man, for any one on whom it fixes its eyes dies immediately.

Marcilio Ficino on the Catoblepas

I found this quotation on p.84 here:

‘Are you surprised that the body of one man is contaminated by the rational soul of another? But you are not surprised that one soul is harmed by another when we gulp down alien vices from the company we keep. You are not surprised that your body is easily infected with disease by the vapor of another body as is obviously the case with consumption, epidemy, leprosy, the itch, dysentery, pleurisy, and conjunctivitis. Among the western Ethiopians purportedly lived beasts called the catoblepas that would kill men simply by looking at them (…), so effective is the power in the vapors of [their] eyes (…). Such is the power of the imagination and especially when the vapors of the eyes are subject to the emotions of the soul’

Ficino, TP, XIII, 4; Allen, IV, 195-197: Ficino, De vita III, XVI; K&C, 325.

John Jonston (1614)

Skipping past the 16th century (for now), once we get to the 17th century interest in the catoblepas somewhat wanes. Of the two famous 17th century drawings, the first was from John Jonston, which depicts something much closer to the gnu or wildebeest, which was (almost certainly) the source of the original description many centuries previously:

This was from John Johnston’s Historia naturalis de quadrupedibus, Amsterdam 1614.

Edward Topsell (1607)

Finally, Edward Topsell’s description of the Catoblepas in his (1607) Historie of foure-footed beastes (which was basically an English translation of Conrad Gessner’s epic 1551-1558 “Historia animalium“) was reproduced in John Swan’s 1643 “Speculum Mundi”, p.649:

The Gorgon or Catoblepas is for the most part bred in Lybia and Hesperia. It is a fearfull and terrible beast to look upon, it hath eye-lids thick and high, eyes not very great, but fiery and as it were of a bloudie colour. He never useth to look directly forward, nor upward, but always down to the earth; and from his crown to his nose he hath a long hanging mane, by reason whereof his body all over as if it were full of scales. As for his meat, it is deadly and poison full herbs; and if at any time this strange beast shall see a Bull or other creature whereof he is afraid, he presently causeth his mane to stand upright, and gaping, wide he sendeth forth a horrible filthy breath, which infecteth and poysoneth the aire over his head and about him, insomuch that such creatures as draw in the breath of that aire, are grievously afflicted, and losing both voice and fight, they fall into deadly convulsions.

Topsell’s drawing / engraving looks like this:

So… What To Make Of All This?

It is, alas, a complicated picture. If there is a common thread to be had, it is that nobody prior to John Johnston seems to have had the faintest idea of what it was they were drawing. Catoblepas get rendered as cockatrices, catty things, doggy things, odd blue things, whatever.

The one detail that got Ger Hungerink most excited was the apparent visual parallel between the Voynich Manuscript’s scaly ‘armadillo’ and Topsell’s scaly catoblepas. But at the same time, I should immediately caution that commentators on Topsell usually conclude that Topsell got confused in his translation, and so merged Gessner’s catoblepas with Gessner’s gorgon.

If you want to read Gessner’s chapter on the catoblepas, it is online here (pp.137-139), though there is no drawing or artwork illustrating it (and the chapter swiftly moves on to discuss the Gorgon). But really, unless someone can dig up a sixteenth century catoblepas print that Topsell could well have referred to, I’m currently not at all sure that we can, on the visual evidence we have so far, trace any kind of viable copying path from any of the Cantimpre versions all the way through to Topsell’s scaly catoblepas.

However, there are still many missing mss above, and there are also two sets of entirely different sources which I still need to go through properly, which I’ll have to cover in a separate post (because this one, I think it’s fair to say, has ended up somewhat out of control). So there’s a little way to go yet…

Back in 2006 when I wrote The Curse of the Voynich, I included in the book a whole lot of notes relating to the internal structure of ‘Voynichese’ (i.e. the language, dialect, or manner of writing/encipherment found in the Voynich Manuscript, whichever you happen to feel easiest which).

To be clear, I didn’t claim to have deciphered so much as a single letter: rather, I wanted to communicate the high-level view of Voynichese I had built up (not too far from that of Brigadier John Tiltman) as a collection of smaller ciphers, all artfully arranged into an elegant overall system.

The mystery of EVA d and EVA y

For example, I believed (and in fact still do believe, and for a whole constellation of reasons) that EVA -d- (word-middle) and EVA -y (word-final) are probably kinds of scribal abbreviations (e.g. contraction and truncation respectively): and that to successfully read Voynichese, we will ultimately need to reconstruct how its words are abbreviated.

At the same time, I believe that EVA d- and EVA -y (both word-initial) work differently again, i.e. that the same two letter-shapes are doing ‘double duty’, that they mean different things when placed in different parts of a word.

In Latin, the shorthand shape ‘9’ (the same as EVA y) behaves very similarly to this, insofar as it stands in for com-/con- when it appears word-initially, and for -us when it appears word-finally. This was still in (admittedly light) use in the mid-fifteenth century, so the idea that something could mean different things in different positions within words was still ‘in the air’, so to speak.

Really, what I was trying to do was understand how the Voynichese ‘engine’ worked: to not only identify the individual cogs and pinions (i.e. Tiltman’s smaller component ciphers) but to also move towards identifying how these meshed together to form not just a collection of adjacent tricks, but a coherent (if subtly overlapping) system.

The overall metaphor that seemed most productive to me was that of architecture: that the components that made up Voynichese were laid out not haphazardly, but had a kind of consistent conceptual organization to them, yielding what appeared to be rigid use-structures and language-like rules.

Yet at the same time, attempts to produce formal Voynichese grammars to capture these have proved unfruitful: even though thousands of statistical experiments seem to back up the overwhelming intuition that there’s something there if we could only see it, we remain blind to exactly what is going on.

Yes, It’s Unpigeonholeable

Some Voynich linguists try to argue against my view by claiming that I’m describing it purely as a cipher, which (in their view) ‘of course’ it simply isn’t. But the problem is that that’s really not my position at all.

Rather, one of my overall beliefs about Voynichese is that the person who constructed it would have been able to almost entirely (though perhaps not necessarily 100% completely) read it back off the page. And so a lot of what I’m talking about isn’t so much cryptography as steganography, “hiding in plain sight”: and that in turn isn’t so very far from being a linguistic problem.

So if (as I suspect) Voynichese turns out to be equal parts cryptography, steganography, shorthand and language, decoding it will require a significant collaborative effort: but it will also require people to stop trying to pigeonhole it into a single category. Is there any real likelihood it is pure language, or pure shorthand, or pure steganography? For me, the answer is no.

What many of us moderns forget is that the Renaissance (and particularly the fifteenth century) was a time long before the borders between intellectual specializations had started to be so anxiously patrolled. Back then, there was no hard line between language and cipher, between fact and fiction, between Arts and Sciences, even between past and present: thinking was far muddier, and far less clearly defined. Or, if you want to be charitable, much more fluid and creative. 🙂

And so I think we really shouldn’t be surprised if the creator of the Voynich manuscript trampled gleefully over the flower beds of what we now think of as convention: it would be several hundred years before intellectual “Keep Off The Grass” signs would start to appear.

Vowels, Consonants, Numbers, And, The

Regardless of all the above, I think that anyone trying to make sense of Voynichese really has to start with the most basic questions. Surely the biggest ones (and these have bugged me for nearly twenty years) are the classic questions of both cryptologists and linguists alike:

• Where are the vowels?
• Where are the consonants?
• Where are the numbers?
• Where are the ‘and‘ words?
• Where are the ‘the‘ words?

Unfortunately, many people who go hunting for vowels in Voynichese take its letter shapes completely at face value: and by that token, EVA a / i / o would ‘surely’ be standing in for (plaintext) A / I / O. Even though this at first seems to move you forward, what immediately happens next is that you find yourself utterly, ineffably stuck: that even though “vowel = VOWEL” may (briefly) feel like a plausible starting point, Voynichese doesn’t actually work like that at all.

And so the more well-organized vowel hunters move on to applying linguistic algorithms (such as Sukhotin’s) to determine which letters are vowels, and which are consonants. This normally (e.g. depending on which transcription you are using, how you parse EVA letters into glyphs, etc) will yield much the same kind of result: which also gets you basically nowhere.

This also doesn’t even begin to attempt to answer the question of where the numbers are (for in a manuscript that size, there must surely be numbers aplenty in there, right?); where the ‘and‘ words are hiding; and just as much where all the ‘the‘ definite articles are to be found.

Honestly, how is it that researchers can collectively invest so much time staring at Voynichese and yet they almost all never try to formulate answers (however hypothetical or speculative) to such basic questions?

Shape Families

Despite our continuing inability to read Voynichese, I think we can identify – purely from their shapes and the similar ways they appear – a number of distinct groups of letters:

• EVA e, ee, eee, ch, sh  (the ‘c-family’)
• EVA t, k, f, p (the ‘gallows family’)
• EVA or, ar, ol, al
• EVA an, ain, aiin, aiiin
• EVA air, air, am, aim
• EVA d, y
• EVA qo
• EVA s

Oddly, many of the shapes inside each of these groups can often be substituted for one another (e.g. gallows can normally be substituted one for the other to form similar words): and this alone forms a kind of skeletal “shape-grammar” for Voynichese. (Though quite why this should be the case remains a mystery.)

One of the things I have long wondered about these shape families (which, once again, wasn’t not far at all from what Brigadier Tiltman had suggested) was whether each of them might have previously expressed some kind of individual cipher-like trick: for example, I wondered whether the ololol-like repeats of the or/ar/ol/al group might have originally been specifically used to disguise Roman numbers.

In which case Voynichese wasn’t itself a work of invention so much as one of careful assembly, its creator stitching (and adapting) a set of pre-existing tricks together to form the illusion of a coherent whole.

In which case, the intriguing question then arises as to whether we might be able to reconstruct what each of these families is trying to conceal. Might we be able to work out the secret history of each of these sub-tricks?

On The Vowel Trail

All the same, the question of the day comes down to this: which of these distinct families might be hiding the vowels?

Back when I was writing Curse, I speculated whether the series of ‘c’-like shapes in Voynichese (EVA e, ee, eee, ch, sh) might somehow be standing in for vowels. After all, the members of this set do seem to share some kind of visual ‘family connection’ as far as their shapes go (i.e. they’re all formed of right-facing semicircles, and there are (superficially, at least) as many of them as the number of vowels you might typically expect to find in a typical European text (i.e. five).

A famous medieval monastic cipher also replaced vowels with clusters of dots (e.g. one dot for a, two dots for e, etc), so the idea that a cipher and/or alphabet might ‘thematically obfuscate’ a connected group of letters in the same way is visually (and indeed historically) quite appealing.

At the same time, I think that while this may well prove to be true (or even largely true) for Currier A pages, at the same time something odd is going on with Voynichese Currier B pages that this isn’t capturing. So Voynichese as a whole remains subtler and more awkward than this is able to completely account for.

Strike-Through Gallows

What I also find hugely intriguing is not that there are families of shapes, but that there are also mysterious areas of overlap between those families.

These are the places where I think the creator of Voynichese used his cunning to ‘hybridize’ them, i.e. to adapt the area between a pair of families, to turn the overall set of families into a complete system.

Nowhere is this kind of overlapping clearer than with the strike-through gallows. These are instances where shapes in the gallows family (EVA t, k, f, p) are kind of ‘struck-through’ by a ‘ch’ shape. The difficulty of rendering these struck-through gallows as text led to a lot of debate between people proposing various Voynich transcription alphabets.

In the end, the EVA transcription rendered the ‘ch’ shape as two half-letters so that struck-through gallows could be rendered with a ‘c’ and an ‘h’ either side of it, e.g. EVA k -> ckh, t -> cth, f -> cfh, p -> cph. But remember that this is no more a handy transcription convention, and really shouldn’t be interpreted as endorsing any particular view of what is actually going on ‘under the hood’.

Because that’s another big question researchers have been all too content to avoid ever since EVA arrived: in short, what on earth is going on with these struck-through gallows?

Back when I wrote Curse, I pointed to a 1455 Milanese cipher where, very unusually, ‘subscriptio’ was rendered in a very similar strike-through way: and so proposed that this might well be what we are looking at with strike-through gallows. While this made good hypothetical sense at the time, I have to say it also didn’t really sit well with the idea that EVA ‘ch’ might be in some way part of a vowel family. And so I was left not seeing how these two families and their overlap might have been meshed together

But a couple of years ago, I had an idea as to how all these different pieces could have been reconciled into a single system…

Cicco Simonetta and Q

Philip Neal’s exemplary translation of Cicco Simonetta’s 1474 Regule (‘rules’) for codebreaking includes his translation of Simonetta’s notes on the weakness of the letter ‘Q’:

Consider if in the published writing there be any cipher which always and everywhere is followed on by one and the same cipher, for such a cipher is representative of q, and the other following is representative of u, for always after q follows u, and the cipher which follows on the cipher representative of u is a vowel always, for always after q follows u and another vowel follows after u.

What, then, are codemakers to do to avoid people using QU as a giveaway? Apart from adding in nulls, Simonetta suggests possibly “putting one sole letter in place of q and u”.

Now, what I found interesting about this is that in 1474 (actually, I strongly suspect that Simonetta was copying out a document that had been compiled some twenty years previously, so perhaps in 1454 or so), Milanese codebreakers were aware that leaving ‘q’ and ‘u’ adjacent was a crypto ‘tell’, that could be used to break their ciphers.

And yet in the Voynich Manuscript, there was apparently no sign of any mechanism or shape family being used to obfuscate a ‘qu’ pair. Or… was there?

Revisiting EVA ch

And so I finish this with the thought that struck me a couple of years ago. What if the strike-through gallows were simply formed by a ‘Q’ shape being struck through by a ‘U’ shape?

For if that were the case, we could probably conclude that not only is EVA ‘ch’ a vowel, but the letter it is standing in for is U/V.

Ah, some might say, but there are 18 instances of EVA ‘chch’ in the Voynich Manuscript. However, I would point out that many/all of these could very easily have been copying errors for the (almost microscopically different) EVA ‘chee’ (e.g. ‘dchchy’ could instead have been ‘dcheey’, etc).

Similarly, even though there are 755 instances of EVA ‘chee’ in the VMs, there are only 33 instances of EVA ‘eech’. Perhaps this is representative of words beginning ‘V’+vowel, or of specific diphthongs, I don’t know. There are 4989 ‘che’ instances, but only 180 ‘ech’ instances: maybe this is something that can be mined for more information and insight.

Of course, I don’t know that I’ve got this right: but the suggestion that EVA ‘ch’ is ‘U/V’ is a hypothesis that’s based on good observation and good crypto history, and offers plenty of space to explore and to work with.

For example, it would suggest that ckh is actually the same as (k)(ch), which may help normalize a lot of the text (and please don’t try to argue back to me that k ‘can only’ maps to a single plaintext letter, Voynichese is much too subtle for that, or else we wouldn’t get qokedy qokedy etc).

Lots to think about, anyway.

Diane O’Donovan recently stumbled across a reference to a relatively little-known Italian-Jewish engineer / cryptographer / magician called Abraham Colorni (Abramo Colorni) who was for a short while at Rudolf II’s court: and wondered aloud (in some comments to Cipher Mysteries) what we might learn from his 1593 book on cryptography.

“Scotographia etc etc”

As you might expect, Colorni’s book title is badly afflicted by the prolixity so typical of the age: “Scotographia, ouero, Scienza di scriuere oscuro, facilissima, & sicurissima, per qual si voglia lingua : le cui diuerse inuentioni diuisi in tre libri, seruiranno in più modi, & per cifra, & per contra cifra : le quali, se ben saranno communi a tutti, potranno nondimeno usarsi da ogn’uno, senza pericolo d’essere inteso da altri, che dal proprio corrispondente”. That is, “Scotography, or the science of concealed writing, most easily and most securely, etc etc etc“.

Various physical copies exist: MIT Library, in the Cryptology Collection of UPenn’s Van Pelt Library (I always wondered what happened to Lucy), Harvard Library, BnF, the British Library, the Hungarian Academy of Sciences’ Library, Bayerische Staatsbibliothek, and the Klau Library at Hebrew Union College in Cincinatti to name but eight. There are also microfilm copies at Herzog August Bibliothek and the British Library, if squinting into dusty old back-lit magnifying boxes floats your boat.

Obviously, what you’d actually like to know is what online versions exist. The BNCF website includes only a ragged copy of the first couple of folios of MAGL.3.8.24, which is not that impressive:

The Museo Galileo’s website has a complete set of scans of the BNCF MS, though (perhaps because the whole book has an unusual aspect ratio, i.e. it’s much wider than it is tall), all the Museo’s scans have come out vertically stretched by a factor of three in their reader (the “Reader” icon at the top of the page). This is also true of the PDF download option, e.g. how it is (left) and how it ought to be (right):

Alternatively, you can read the same pages from the index webpage, though only one at a time, and the (unstretched) image goes off the right hand edge of the web page unless you really widen the size of the browser window, which is annoying in a quite different way.

Having said that, none of this is fin du monde etc.

The Book’s Contents

As normal, the book starts with a seven-page laudatory preamble praising Colorni’s most magnificent patron, Holy Roman Emperor Rudolf II, and explaining the symbolic meaning behind the four specific zodiac signs chosen for the frontispiece (Scorpio, Libra, Virgo, Leo):

The book continues with three main chapters (though the middle chapter is tiny), and then finishes up with an enormous enciphering table (more than half the remainder of the book). It also includes some interesting cryptographic figures which I don’t recall seeing elsewhere.

From what I have read, it seems to me (and feel free to correct this impression) that Colorni was not a theoretical codemaker or codebreaker. Though his cipher history account starts with the normal SCYTALE (long thin message wrapped around a stick) cipher yarn, his writing doesn’t seem informed by the work of contemporary crypto theoreticians such as Bellaso.

Rather, I suspect what happened was that Colorni collected together a series of cryptographic tricks (such as nulls, verbose cipher, etc) and then adapted and extended them into something cunning and ingenious which he believed to be practically impregnable. So I think his book (to answer one of Diane’s questions) documents various cunning “peasant ciphers” rather than being part of a theoretical crypto mainstream.

Incidentally, just about the only ciphertext given in Colorni’s book (there are no challenge ciphers) is:

GWGHPCXKGBEDMMYWOPWQPWO
HMAAHXNAYLPKOOBPXKFFLTGWYIXG

Feel free to try to crack it if you wish. 🙂

Colorni and the Voynich Manuscript?

But, Diane continues, might it have been Abraham Colorni who brought the Voynich Manuscript to Rudolf II’s Golden Court in Prague? Superficially, Colorni would certainly seem to tick many of the boxes, and there’s unlikely to be evidence out there that explicitly proves that he didn’t bring it. (After all, what are the chances a letter now turns up saying “It wasn’t me, Abraham Colorni, who sold that scandalous naked women cipher book to the Emperor, it was that blasted John Dee“?)

All the same, I don’t believe that Colorni’s book’s introductory dedication to Rudolf II (written in 1593) mentions anything sounding at all like the Voynich Manuscript (as always, please feel free to correct me if I’m wrong). It does namecheck Oedipus, but presumably for broadly the same reasons that Georg Baresch also (independently) namechecked Oedipus several decades later.

Perhaps a more productive route to take would be to look at Colorni’s correspondence, and see if that casts any light on the subject. And, very helpfully, there are (at least) two freely downloadable 19th century articles by Professor Giuseppe Jarè that might assist us in this regard:

• Abramo Colorni : ingegnere Mantovano del secolo XVI. ; con documenti ined.
• Abramo Colorni : ingegnere i Alfonso II. d’Este ; memoria letta nell’ adunanza 30 marzo 1890

Both articles include transcriptions of a number of letters (in both Latin and Italian) culled from numerous archives. In fact, the second article contains so many that I suspect that Jarè must have had Colorni as an ongoing research interest for some twenty years or more.

Though some of these definitely mention Colorni’s Scotographia, I didn’t notice anything related to the Voynich Manuscript in there. However, others more observant and diligent than me may have more luck: and wouldn’t that be nice? 😉

Secondary Literature on Abraham Colorni

Though I’ve tried to limit my discussion here of Abraham Colorni to primary evidence, there is also a pretty good modern literature on him if you’re interested:

• The age of secrecy : Jews, Christians, and the economy of secrets, 1400-1800 – Daniel Jütte
• Or, in German: Das Zeitalter des Geheimnisses : Juden, Christen und die Ökonomie des Geheimen, (1400-1800) – Daniel Jütte
• Trading in secrets : Jews and early modern quest for clandestine knowledge (Isis, Vol. 103 (2012), p. 668-68)
• Il prestigiatore di Dio : avventure e miracoli di un alchimista ebreo nelle corti del Rinascimento – Ari’el To’af – Milano : Rizzoli, 2010
• Rene Zandbergen also points out there is a chapter on Colorni by Vladimir Karpenko in:Alchemy and Rudolf II, Exploring the Secrets of Nature in Central Europe in the 16th and 17th centuries, edited by Ivo Purš and Vladimír Karpenko, Artefactum, Prague (2016), though probably building on Daniel Jütte’s book.

A Needle In A Haystack?

For twenty-plus years, Rene Zandbergen and a whole host of others have invested a lot of time into trying to dig up references / historical evidence relating to the Voynich Manuscript’s (probable) time at Rudolf II’s court: but have so far found nothing.

From what I know, I don’t currently believe that Abraham Colorni will turn out to be the missing link, the “Herald” (in Joseph Campbell / Hero’s Journey terms): rather I think that if it did make its way to Rudolf II’s court, it was very much towards the end of his rule (notionally at Rudolf II’s death in 1612, but he was under a kind of house arrest by his brother Matthias for the last few years – families, eh, who’d have ’em?). And with Colorni dying in 1599, the two therefore probably didn’t overlap in Prague.

All the same, I find Professor Giuseppe Jarè’s articles hugely heartening, because he was able to collect together from a whole list of archives all manner of correspondence relating to Colorni: and that gives us access to a evidential slice cutting through Colorni’s life.

So perhaps the right thing to do, Voynich-wise, is to stop looking for a needle in a haystack – i.e. a single perfect piece of evidence – and to instead start looking for a sewing box in a haystack. By this I mean collections of diligently-collected letters and documents not unlike Jarè’s collection of Colorni’s correspondence, but for technical-minded court insiders who were at Rudolf II’s court nearer the end.

The best attempt at doing this so far has been by looking at the correspondence between Duditius and Tadeáš Hájek z Hájku (1525-1600), who was Rudolf II’s Imperial Astronomer, as studied by Josef Smolka (with help from Rene Zandbergen). I previously discussed their lack of (Voynich-related) success here, and concluded that the 1600-1612 period might be more fruitful.

But do we have a list of people who we might even consider as candidates for this kind of search? One would have thought that the 15 volumes of Tyco Brahe’s correspondence (in Tychonis Brahe Dani Opera omnia) to 1601 would have been thoroughly mined by Voynich researchers by now. Christoph Rothmann (of Kassel) similarly died in 1600, while Caspar Peucer died in 1602. Even so, I suspect we are likely to have no luck with any of them.

Has anyone trawled through Kepler’s correspondence looking for partial or indirect references to the Voynich Manuscript? I’m thinking that perhaps the best way forward would be to look at the network of correspondents linked to Kepler in the 1600-1615 period. The letters between Kepler and Galileo are well-known, but they surely can’t make up even 25% of Kepler’s correspondence, right?

Perhaps one of these letter writers will have heard mention of the Voynich Manuscript: and perhaps this is how the first big piece in the Voynich jigsaw will be found, who can say? 🙂