Voynich Quire 20 notes…

by ⋅ 11 Comments

Though the Voynich Manuscript has many unusual and interesting sections, arguably the most boring of the lot is Quire 20 (‘Q20’). This comprises a thick-ish set of six text-only bifolios (though with a central bifolio missing), where just about every paragraph has a tiny drawing of a star/flower/comet shape attached to it.

Tally all these up, and you find that Q20 as it now is has between 345 and 347 stars (two are slightly questionable). This is intriguingly close enough to the magic number 365 for some researchers to have speculated that this section might actually be an almanack, i.e. a one-paragraph-per-day-of-the-year guide. It is also close enough to the similarly resonant number 360 for some (and, indeed, often the same) people to speculate whether the section might be a kind of per-degree astrology, i.e. a one-paragraph-per-degree judicial astrological reference. If neither, then the general consensus seems to be that Q20 is some kind of secret recipe / spell collection (there were plenty of these in the Middle Ages).

Yet the problem with constructing an explanation around 365 stars, as Elmar Vogt pointed out in a 2009 blog post, is that Q20 ticks over fairly consistently at about 15 paragraphs/stars per page, and we have four pages missing from the centre: hence we would expect the grand total to be somewhere closer to 345 + (15 x 4) = 405, which is far too many. Of course, the same observation holds true for 360 per-degree paragraphs too.

You could try – as Elmar subsequently did – to start from a count of 365 (or indeed from 360) and hold that as a constraint to generate plausible codicological starting points (i.e. by answering the question “what kind of ordering could be constructed that would be consistent with 360 or 365 starred paragraphs?”). However, you quickly run into the problem that the marginalia-filled f116v looks very much as if it was originally (as it now is) the final page of the quire, which (in my opinion) is also supported by the large, final, unstarred paragraph on f116r, [which, incidentally, I predict will turn out to hold the most informative plaintext of the whole book, i.e. stuff about the author]. And if you accept that f116 was indeed probably originally the end folio, then you run into a bit of a brick wall as far as possible shufflings of Q20 go.

All the same, here’s an interesting new angle to consider. Cipher Mysteries reader Tim Tattrie very kindly left a comment here a few days ago, pointing out that the rare EVA letter ‘x’ appears in 19 folios – “46r, 55r, 57v, 66r, 85r, 86v6, 94r, 104r, 105r, 105v, 106v, 107v [NP: actually 107r], 108r, 108v, 111r, 112r [NP: ?], 113v, 114v, 115v” – which he wondered might suggest a possible thematic link between Q8, Q14 (the nine-rosette fold-out) and Q20. However, what particularly struck me about this list of pages was that (remembering that f109 and f110 form Q20’s missing central bifolio) every folio in Q20 has an ‘x’ except for the first folio (f103) and the last folio (f116), which are of course part of the same bifolio.

Could it be that the f103-f116 bifolio was originally separate from the rest of the Q20 bifolios? Even though f116v looks as though it really ought to be the back page of the quire, I have always found f103r a very unconvincing first page for a quire – with no introduction, no title, no ornate scribal scene-setting, to my eyes it seems more like a mid-quire page, which is a bit of a puzzle. In fact, of all the Q20 starred paragraph pages I’d have thought the strongest candidate for quire-initial page would be f105r, courtesy of its overelaborate split (and dotted, and starred) gallows, as per the below image. Just so you know, f105r also contains two of Q20’s four ‘titles’ (an offset piece of Voynichese text), one on line 10 and one on its bottom line, the latter of which crypto crib-fans may well be pleased to hear that I suspect contains the title of the chapter – “otaiir.chedaiin.otair.otaly” (note that the two other Q20 titles are to be found on f108v and f114r).

Interestingly, Tim Tattrie then pointed out that “f103 and 116 are linked as they are the only folios in Q20 that do not have tails on the stars“: which (together with the absence of ‘x’s) suggests that they were originally joined together, i.e. as the central bifolio of a quire. All of which suggests that the original layout of Q20 may have been radically different layout, with the f105/f114 bifolio on the outside of (let’s call it) ‘Q20a’ and the f103/f116 bifolio on the outside of ‘Q20b’, and all the other Q20 bifolios in the middle of Q20a.

Picking up where Elmar left off before, this is where almanack (365-star) and per-degree (360-star) proponents would somewhat enthusiastically ask how many stars this proposed Q20a gathering would have originally had. Because f103 has 35 stars, and f116 has 11 stars, Q20a has (say) 299 stars across five bifolios. Add in the missing f109/f110 bifolio, and the answer is terrifically simple: extrapolating from Q20a’s stars-per-paragraph figure, we’d expect the original quire to have had (299 * 6 / 5) = 359 stars – hence the likelihood is that Q20a contained some kind of per-degree zodiac reference, most probably along the lines of Pietro d’Abano’s per-degree astrology.

Having said that… the figures are extremely tight, and if the missing bifolio instead mimicked the f108/f111 bifolio (which, after all, it ended up physically next to in the final binding) rather than the overall section average (which may also be slightly lowered by the 11-per-page average of f105), it would have closer to 72 stars rather than 60, which would then bring the 365-star almanack hypothesis back into the frame. It’s a tricky old business this, nearly enough to drive a poor codicologist to drink (please excuse my shaking white cotton gloves, etc).

There are various other observations that might help us reconstruct what was going on with Q20a/Q20b:-

  • Helpfully, Elmar documented the various empty/full star patterns on the various Q20 pages, and noted that most of them are no more than empty-full-empty-full-empty-full sequences, with the exception of “f103r, f104r and f108r”, which he interprets as implying that they were probably originally placed together.
  • f105v is very much more faded than all the other pages in Q20, which I have wondered might have been the result of weathering: if f105 had been the outer folio of an unbound Q20a, then it could very well have ended up being folded around the back of Q20b by mistake for a while early in the life of the Q20 pages, which would have put it on the back page of the whole manuscript.
  • I still believe that the gap in the text in the outside edge of f112 was no more than a vellum fault (probably a long vertical rip) in the page in the original (unenciphered) document from which Q20a was copied, but that’s probably not material to any bifolio nesting discussion.
  • The tail-less paragraph stars on f103r seem slightly ad hoc to my eyes: I suspect that they were added in after the event.
  • The notion that Q20 originally contained seven nested quires (as per the folio numbering) seems slightly over-the-top to me: this seems a bit too much for a single mid-15th century binding to comfortably take.

What are we to make of all this fairly raw stream of codicological stuff? Personally, I’m fairly sold on the idea that Q20 was originally formed of two distinct gatherings, with f105r the first page of the first gathering (Q20a) and f116v the last page of the last gathering (Q20b). Placing all the remaining bifolios in Q20a to get close to the magic numbers of 360 and 365 is a hugely seductive idea: however, such numerological arguments often seem massively convincing for a short period of time, before you kind of “sober up” to their limitations, so I think it’s probably safer to note the suggestion down as a neat explanation (but to return to it later when we have amassed more critical codicological information about the Q20 pages).

Perhaps the thing to do now is to look at other features (such as handwriting, rare symbols, letter patterns, contact transfers, etc) for extra dimensions of grouping / nesting information within the Q20 pages, and to look for more ways of testing out this proposed Q20a/Q20b split. Sorry not to be able to come to a definitive conclusion on all this in a single post, but with a bit of luck this should be a good starting point for further discussion on what Q20 originally looked like. Tim, Elmar, John, Glen (and anyone else), what do you think? 🙂

Update: having posted all the above, I went off and had a look at all the ‘x’ instances in Q20 – these seem to have a strong affinity for sitting next to ‘ar’ and ‘or’ pairs, e.g. arxor / salxor / kedarxy / oxorshey / oxar / shoxar / lxorxoiin, etc.

Voynich Cisioianus cipher crib…?

by ⋅ 11 Comments

When is Easter? A simple question, but one with quite a tricky answer: following the decision of the First Council of Nicaea in 325AD, it is the first Sunday after the full moon after the Spring Equinox (which is simplified to be 21st March): hence, Easter can fall anywhere between 22nd March and 25th April.

A moment’s reflection should be ample to reveal what a dog’s dinner of a calculation this entails: and when combined with leap years, calendrical uncertainty, and subsequent calendrical reform, what a practical mess it yielded in the centuries following. Even Carl Gauss got his own Easter-calculating algorithm wrong first time round (and he was no mathematical slouch).

From the early Middle Ages onwards, the awkward task of determining when Easter fell was known as computus – Latin for ‘computation’. In fact, you might (just about) argue that the Nicaean Council’s curious dating mix of pagan festivals, Metonic cycles, astrology and religion provided the original impetus for the modern digital computer – people in the Church had been computing Easter by hand for the previous millennium or so, and were doubtless thoroughly sick of the whole thing.

Given all the above, the obvious historical question to ask is: how on earth did anyone ever manage to calculate Easter? The answer lies in a motley bunch of tables, diagrams, and mnemonics devised, copied and adapted throughout the Middle Ages and early Renaissance that attempt to make the task do-able. For the most part, these are built upon the 19-year cycle of the moon (the Metonic cycle): this means that any time you find yourself looking at an unusual-looking table or diagram in a medieval manuscript that ‘just happens’ to be divided into nineteen columns or segments, there’s a fairly good chance it will turn out to be some kind of computus-based trickery.

The literature on computus is fairly spotty, because (I think) it tends to fall between two stools: basically, it’s too religious to be of interest to many historians of science, but also too scientific for many historians of religion. However, one decent starting point is a 1954 article in Speculum by Lynn Thorndike (one of my favourite historians, as long-time Cipher Mysteries readers will no doubt recall) called simply “Computus” (here’s the JSTOR page for it).

Thorndike had previously written a 1947 paper “Blasius the Franciscan and his Works on Computus” (again, here’s its JSTOR page), in which he discussed Blasius’ “circio” computus mnemonics and their reception in other manuscripts: for example, “CIRCIO” decomposes into CIR = “January 1st, circumcisio domini, the Feast of the Circumcision”, CI = “C, the third letter of the alphabet, which (I think) signifies the third section of the nineteen-year cycle”, O = “O, the 14th letter of the Latin alphabet, hence Easter falls on the 14th April”. Which is to say, in the thirteenth century (probably), Blasius constructed a tricksy Latin-sounding mnemonic that (it seems) replaced one of the computus tables (though note that I haven’t yet read either Thorndike article, so this is just a guess).

But this was not the only similar mnemonic from this time: what became far better known was the “Cisioianus” / “Cisiojanus” mnemonic. Because this spread mainly through 14th and 15th century German woodblock calendars, there’s a fair bit of German-language literature on this, and (for a nice change) the German Wikipedia page on Cisiojanus is actually quite helpful.

Basically, a Cisioianus mnemonic consists of 12 Latin-sounding (but nonsensical) couplets, padded out so that you step through the number of syllables to remember the saint’s days and feasts in that month. Here’s the couplet for January, from where you can see that the mnemonic got its name from the first two ‘words’:-

císio jánus epí ¦ sibi véndicat óc feli már an
prísca fab ág vincén ¦ ti páu po nóbile lúmen

(In case any passing pub quiz pop trivia fans are wondering, Carol Decker’s band “T’Pau” was named after a Vulcan priestess in Star Trek, not after the “ti pau” in the second line here. Just so you know.)

So: because January has 31 days, the couplet for it has 31 syllables, with the feast days highlighted:-

  1. cí → circumcisio domini, the Feast of the Circumcision
  2. si → (continuation)
  3. o → (continuation)
  4. ján → (a reminder that this is the couplet for January)
  5. us → (a reminder that this is the couplet for January)
  6. ep → epiphanias, Epiphany
  7. í → (continuation)
  8. si → (null)
  9. bi → (null)
  10. vén → (null)
  11. dic → (null)
  12. at → (null)
  13. óc → octava epiphaniae, the eighth day of the Epiphany
  14. fe → Felicis presbyteris
  15. li → (continuation)
  16. már → Marcelli papae
  17. an → Antoni abbatis
  18. prís → Priscae virginis martiris
  19. ca → (continuation)
  20. fab → Fabiani et Sebastiani
  21. ág → Agnetis virginis
  22. vin → Vincentii martiris
  23. cén → (continuation)
  24. ti → Timotei martiris und Titi martiris
  25. páu → conversio Pauli
  26. po → Polycarpi episcopi martiris
  27. nó → (null)
  28. bi → (null)
  29. le → (null)
  30. lú → lumen
  31. men → (continuation)

So, now you know a couplet to remember all the important medieval feast days in January. All you have to do is remember the other eleven couplets and you’ve got the whole year covered, right?

Incidentally, January was named after the two-headed gate-keeper Janus, god of doors and gates (though personally I would prefer it if we had stuck with the Anglo-Saxon “Wulfmonath”, the perishingly cold month when hungry wolves try to enter villages, the original ‘wolf from the door’). And also… Macrobius relates that Roman boys would play with a coin called the “as” (which had Janus on one side and a ship’s prow on the other), calling “capita aut navia?” – (‘heads or ships?’), which presumably morphed into the modern “heads or tails”… but I perhaps have digressed a tad too far here!

Of course, human nature being what it is, people then went on to construct rude and/or ridiculous versions of this basic cisioianus mnemonic that were easier to remember, but that’s a story for another day. 🙂

“Fascinating, Nick… but how on earth is this all linked to the Voynich Manuscript?“, I hear you (very reasonably) ask. Well… this all started with an intriguing email from Steve Herbelin, who got the online Voynich / historical research bug a while ago. He had been particularly intrigued by the circular picture on f67r2, which seems to be built around some kind of rational, 12-way division, presumably depicting something calendrical… but what?

f67r2-400x500-enhanced

Specifically, Steve wondered if this (or something similar) might reappear in other medieval manuscripts. After some protracted searching, he found this online image from a manuscript from Auxerre from circa 1400 which has plenty of circular computus diagrams (hence all the discussion of computus above), and the following 12-way circular diagram on fol. 9v:-

AuxerreMS240-fol9v-centre

Decoding this: the outer ring (#1) is a reminder of which cisioianus couplet to use, ring #2 is the month name (“januari9” is at about 8 o’clock), #3 is the kalends, #4 is the nones, #5 is the ides, and the innermost ring (#6) says whether the month belongs to the third (lunar regulars) or fifth (new moon calculation) cycle.

Basically, Steve wonders whether these two images might somehow be part of the same (cladistic / stemmatic) family-tree of manuscripts: that is, whether the text in f67r2’s twelve segments might encipher the same kind of information on the Auxerre MS’s fol. 9v.

Having thought about this for a few days, though the precise details probably don’t quite mesh as well as they at first appear, I really don’t think you can dismiss this comparison out of hand. Mnemonics were useful and not widely known (and so might well fall into the category of “secret practical knowledge“): and it has long been noted that the “medallions” in the centre of the Voynich Manuscript’s zodiac pages do seem to hark back to the kind of illustration you’d find in early German woodblock-printed calendars, so there may well be some kind of reasonably direct influence there.

My own take on f67r2 (The Curse of the Voynich, pp.59-60) has long been that it seems to link a 12-way division around the outside with an 8-way division in the centre, and so (as astrology historian David Juste suggested to me several years ago) could very easily depict or signify some kind of calendrical conversion between a 12-way (lunar) zodiac/administrative calendar and an 8-way (solar) pagan/agricultural calendar. All of which is very neat: but fails to explain the 12 coloured moons or the structure of the text.

Of course, if we could only find the way in which any one ring of the f67r2 diagram enciphers the same information as a ring on the Auxerre MS fol. 9v, then we’d have an almost unbeatably good crib to crack the VMs’ cunning cryptography. However, nothing to do with the Voynich has ever proved to be that straightforward…

For a start, there don’t seem to be 30-31 syllables in each of the 12 segments (however you try to count them), so we can probably rule out a full cisioianus plaintext: so matching this in some abbreviated way would require a bit of thought. Also, I don’t (yet) know the details of Blasius’ “circius” mnemonic, but that might possibly be a better match (as long as it is a 12-part mnemonic rather than a 19-part mnemonic). Furthermore, I can’t see an obvious match with month names (which others have tried to do here for decades), and we don’t even know where the sequence of twelve segments start (or indeed end).

Interestingly, there’s a marginal mark at the top left of f69r2 which came out artificially sharply in the enhanced image above. At full resolution it looks rather messier, but might possibly include a left-to-right-flipped “J” at the bottom:-

f67r2-top-left-detail

Might this be indicating where to start on the diagram; or might it instead signify the start of the quire or chapter? (This was formerly the frontmost page of Quire 9, before it was rebound along the wrong fold, pace John Grove).

At this point, I have to call a halt on this (already far too long) post: once again, I don’t have all the answers, but perhaps I have managed to ask one or two reasonably good questions. All credit to Steve Herbelin!

Voynich Summer Camp, transcript of session #1…

by ⋅ 2 Comments

For the recent Hungarian Voynich summer camp, I offered to do a couple of IM sessions over Skype, both of which seemed to go down very well. I thought many Cipher Mysteries readers might enjoy going over the transcript, so here it is (lightly edited for house style, as usual, and with after-the-event section dividers to make it not quite so unwieldy). I’ll be posting the Sagittarius spreadsheet mentioned here to the the blog very shortly, have no fear! 🙂

Note: Skype IM session on 27/08/2009
vc = “voynich camp”
NP = “Nick Pelling”

o. Introduction

[19:01:42] vc: Thank you very much for the possibility.
[19:01:57] NP: No problem – may I ask who is there this evening?
[19:02:41] vc: We are currently five pople but we are six and tomorrow we will be seven.
[19:02:50] NP: five is fine 🙂
[19:03:11] NP: Do you have names, or are you Devo?
[19:03:19] vc: Two linguists, and 3 geekz.
[19:03:46] NP: I’m not sure who would win the fight. Best to work together.
[19:03:57] vc: Zsolt, Ancsi, Norbi, Miklos, Anna, Andris.
[19:04:13] NP: OK, got it. Where shall I begin?

1. Page Reference Discussion

[19:04:49] vc: can you speak about your own theory about the seem to be page links?
[19:05:29] NP: OK – the apparent page references
[19:08:00] NP: Large medieval documents were bound into quires, and the quires were numbered
[19:08:16] NP: But for centuries, people used letters rather than numbers
[19:08:21] NP: a b c d e etc
[19:08:46] NP: folio numbers were then given in roman numerals for that quire
[19:08:50] NP: i ii iii iv v etc
[19:09:11] NP: The side of the page was also given, with r for recto, v for verso
[19:09:57] NP: So, for centuries people referred to page numbers pimarily in tiny little letter patterns of the form: air, aiv, aiir, aiiv, aiiir, aiiiv
[19:10:38] NP: Few people nowadays (apart from codicologists) bear this in mind, but this was the primary way in which books were structured circa 1400
[19:11:22] NP: So, if we were to look at the VMs through the eyes of someone living 500 years ago, the aiiv / aiir patterns would immediately jump out a t them as being page references.
[19:12:05] NP: Specifically, page references to the first bound quire of a book (for there is no bir biv biir biiv pattern in the VMs)
[19:12:30] NP: So, you might say, what if these are indeed page references to the first quire?
[19:13:09] NP: Well… if they are page references, they’re quite strange ones: references to recto (r) pages would then be vastly outnumbered by references to verso (v) pages
[19:13:51] NP: Ultimately, it comes down to a paradox of vision – we’re presented with something that *looks* like page numbers, but which does not have the statistical profile of page numbers. How can that be?
[19:14:26] vc: We are listening, go on
[19:14:36] NP: Once you start saying “well, these are only designed to resemble page numbers”, then you’re into the realms of cryptography and steganography
[19:15:36] NP: I think that the presence of these non-page-number page numbers is a persuasive reason to look for cryptographic solutions in preference to linguistic solutions
[19:15:43] NP: But then…
[19:16:13] NP: You have to ask the question, what are these non-page-number page numbers actually doing? What is their function?
[19:16:43] NP: When I went to the Beinecke in May 2006, I spent an hour looking at a single page of the VMs trying to work it out.
[19:16:58] NP: f38v
[19:17:32] NP: I had borrowed a magnifying glass from the front desk and went over every aiiv-style pattern on the page, over and over again, looking for a clue
[19:17:46] NP: And then I noticed it, one tiny piece at a time
[19:18:15] NP: And in the two years since, I’ve fleshed it out into a complete story of how they came about, and why they have the shape they do.
[19:19:20] NP: In the first phase, the author wrote a simple “v” shape and added a carefully placed dot above the row of i’s. This indicated one of the Arabic numerals 1-9 (I think)
[19:19:55] NP: However, after that early Herbal A phase, the author revisited it and thought it was too obvious, and so added scribal flourishes.
[19:20:20] NP: These loops start from the dot, go up and around, and rejoin the unflourished “v” shape at the top right
[19:20:54] NP: In the Currier B phase, the author used a completely different way of enciphering numbers (but using the same aiiiv family pattern)
[19:21:13] vc: Yes, we can see it in Psp
[19:21:26] NP: So, you can see the layers of development of the cipher over time. Voynichese wasn’t a static system, it was an evolving system
[19:22:13] NP: The problem with using statistical tools on Voynichese is that the system moves over time, which is why it is important to apply tests to each phase in turn
[19:22:21] vc: That’s why you are so eager about to reconstruct the original binding order…….
[19:22:32] NP: That’s one of the reasons, yes
[19:23:02] NP: Mainly it comes back to the evolution of the cipher, though. By tracking the evolution, we can perhaps find a way into it
[19:23:08] vc: because the question arises – why one can observe this on 38 v first?
[19:23:37] NP: I chose the page because it was clearly written and had plenty of aiiv patterns on to work with.
[19:23:54] NP: Some herbal pages are clear but don’t have much text
[19:24:50] NP: I suppose one of the cutting edge research things I’d like you all to come away with from this is the idea that the Voynich grew and developed
[19:25:13] vc: Is it observable on other pages? for example, on a page written by another Currier hand?
[19:25:14] NP: It was not a “fait accompli”, a huge static thing.
[19:25:24] NP: Only on Currier A pages
[19:25:31] NP: Hand 1 pages
[19:25:40] vc: Hmmm…
[19:25:58] NP: On Hand 2 pages, the “v” flourish is written in one go, but takes a number of shapes
[19:26:11] NP: I think that the Arabic numeral is enciphered using those shapes
[19:26:17] NP: Look for yourself
[19:26:41] NP: I should say: “the v and the scribal flourish are written together in one go”

2. The vertical column on f49v

[19:27:15] vc: Speakin’ bout arabic numerals, have you checked with your own eyes on the real thing F49v?
[19:27:20] NP: yes
[19:27:27] NP: not original
[19:27:31] vc: There are arabic numbers on the left margin.
[19:27:40] NP: Yes, 16th century Arabic numerals
[19:27:52] NP: Yet the quire numbers have 15th century Arabic numerals
[19:28:09] vc: It seems on the reproductions that they are about the same color and style as the Voynichese letters nearby
[19:28:10] NP: Doesn’t add up, unless the f49v letters were added later
[19:28:19] NP: They’re pretty close, but they’re not the same
[19:28:28] vc: 🙁
[19:28:44] NP: It fooled Brumbaugh – he erected a huge theory on that alone
[19:28:54] vc: Next Q from us:
[19:29:22] vc: Then how would you interpret these single character lists on the left of some pages?
[19:29:56] NP: (1) Page titles
[19:30:00] NP: (2) cipher keys
[19:30:02] NP: (3) both
[19:30:05] NP: (4) neither
[19:30:07] NP: 🙂
[19:30:10] vc: XD
[19:30:46] vc: …So you were able to turn the original pages with your bare hands?
[19:30:49] NP: Oh yes
[19:30:52] vc: Jeez…
[19:31:00] NP: They were clean
[19:31:01] NP: 🙂
[19:31:20] NP: Seriously, the Beinecke loves people to look at its books
[19:31:36] NP: But they do insist you take your tinfoil hat off at the door
[19:32:26] NP: Actually, I looked at the VMs for three whole days, which was extremely generous of them, and that was much appreciated

3. Steganography or cryptography?

[19:32:20] vc: Right. So: (we haven’t read your book…yet) Steganography or cryptography?
[19:32:30] NP: Both
[19:32:55] NP: It is a cipher that has been disguised as a language
[19:34:05] NP: One of the back-end ciphers (the verbose ciphers I blogged about earlier today) turns a dense, information-rich series of tokens and turns it into an information-sparse language-like entity
[19:34:21] NP: XYZ –> BABEBI
[19:34:33] NP: is a trivial example
[19:34:57] NP: where X –> BA, Y –> BE, Z –> BI
[19:35:38] vc: Right -we have just thinking about the same conception.
[19:35:44] NP: This gives the apparent CVCVCV behaviour that makes the VMs seem so language-like
[19:35:56] vc: Just to reveal some of our works —
[19:36:03] NP: ok…
[19:36:14] NP: http://ciphermysteries.com/2009/08/27/voynich-cipher-structure
[19:37:36] NP: And here’s a test I did in 2003, so I’m pretty consistent: http://voynichcentral.com/users/nickpelling/pairs.gif
[19:37:45] vc: We will definitely read your article later…
[19:38:41] NP: And here’s the article I couldn’t find earlier: http://ciphermysteries.com/2009/06/06/the-voynich-cipher-for-code-breakers
[19:41:27] vc: Ok, we have just made a quick look on them for some of us have’nt seen these before.
[19:41:49] NP: There’s lots of good stuff on my blog. Lots!
[19:42:06] vc: The author knows 🙂
[19:42:25] NP: I wouldn’t post it if it wasn’t good. 😉
[19:42:42] NP: You should see the stuff I don’t post! ;-o
[19:42:57] vc: We downloaded your whole blog and copied it to identical CD’s

4. Occitan sources?

[19:43:17] vc: Ok. Next q:
[19:43:36] vc: About the Occitan language.
[19:44:06] vc: Where does the info on the month names come from?
[19:44:30] NP: Jorge Stolfi asked a load of Occitan researchers back in 1997
[19:44:46] NP: There’s a page on it, it’s certainly in the 1997 vms-list archive
[19:45:29] NP: Here’s this week’s post on it: http://ciphermysteries.com/2009/08/22/jaume-deydiers-livre-de-raison
[19:45:36] vc: hmm…
[19:45:53] NP: Basically, the closest match was with Toulon Occitan
[19:46:05] NP: But with a bit of German thrown in (for “augst”)
[19:46:26] vc: (also: “may”)
[19:46:36] NP: But is it maij or may?
[19:47:41] NP: However, it’s very likely that the Toulon Occitan was merely one person (Jaume Deydier) so it’s a bit of a small sample to be doing inferences from
[19:48:17] NP: Unless anyone knows better, there aren’t that many written examples of pre-1600 Occitan to be found anywhere
[19:48:34] NP: And spelling was a bit of a fluid thing back then, too
[19:48:58] vc: yes, our lingusist here told us 🙂
[19:49:03] NP: I’ve just read a book on artisan autobiographies, and many artisans deliberately spelt things how they sounded
[19:49:22] NP: a kind of anti-academic independent spirit

5. Voynich Marginalia questions

[19:50:27] vc: Do you think that the marginalia and other features written with latin alphabet originate from more than one person?
[19:50:40] NP: Yes, but perhaps not in the way that you might think
[19:50:51] vc: ?
[19:51:03] NP: The problem with the marginalia is that they have Voynichese embedded in them
[19:51:18] vc: This is not a problem. 🙂
[19:51:36] NP: Well.. you’ll see why it might be…
[19:51:48] NP: If you look at f17r under an ultraviolet black lamp, there’s a piece of Voynichese tacked on the end
[19:52:04] NP: it reads: oteeeolair
[19:52:28] NP: Rene Zandbergen independently observed it earlier this year, so I didn’t just imagine it
[19:52:30] NP: 🙂
[19:52:48] vc: Do one usually have a portable ultraviolet black lamp when you go to the Beinecke? 🙂
[19:53:00] NP: Luckily they have one behind the desk you can borrow
[19:53:06] NP: I left mine in the car
[19:53:07] NP: 🙂
[19:53:36] NP: Similarly, there’s Voynichese embedded in the f116v marginalia
[19:53:48] vc: But that can be seen.
[19:54:02] vc: It would be really nice if they could be used as cribs
[19:54:23] NP: That’s what Newbold thought…
[19:54:37] NP: But why can’t we read the marginalia on f116v?
[19:54:47] vc: Yes… Why?
[19:55:10] NP: I think the answer is that successive owners have emended (altered incorrectly) the text, trying to rescue fading words
[19:55:52] NP: And so f17r, f66r, and f116v have all been rescued but worsened
[19:56:44] NP: Unfortunately, the Beinecke’s scans aren’t good enough for us to be able to do forensic uber-restoration on these
[19:56:56] NP: I’ve tried, though, believe me I’ve tried
[19:57:25] vc: We have consulted with a hungarian medieval scholer who suggested that text on the 116v is actually a prayer or something like that in the realm of magic
[19:57:31] vc: a rune
[19:57:38] NP: Well, a prayer is close
[19:57:46] NP: ahia maria
[19:57:51] NP: + + +
[19:57:57] vc: the characters like + as separators justifiy this theory.
[19:58:00] NP: yes
[19:58:35] NP: However, if you look at the “rescued” letters at the beginning of the same line, you can see the contrast with the faded “ahia maria” letters
[19:58:57] vc: He said that this is usual at medieval spells. At these places one might draw a cross in the air. how is it said in english?
[19:59:20] NP: the same, probably “trace a cross shape in the air”
[19:59:33] NP: “trace out a cross shape in the air” probably better
[19:59:49] vc: As Catholics still do
[20:00:16] NP: It could be a prayer, or a curse, or a spell (protection?) or anything
[20:00:52] NP: Remember that the boundaires between magic and liturgy in the 15th century was painfully thin – both were usually carried out by broadly the same people.
[20:01:17] NP: See Richard Kieckhefer’s book “Forbidden Rites”, recommended
[20:01:19] vc: (What is lighter than a witch? A stone perhaps? Or a duck?)
[20:02:00] NP: I’m getting that sinking feeling
[20:02:47] NP: necromancy, charm, spell, prayer – pretty fluid lines between them all
[20:02:58] vc: Whatever the main language of the spell, the last words seem to be in german. What is your opinion about this?
[20:03:31] NP: I think that this was probably added by the first obsessive Voynichologist (who just happened to be German)
[20:03:34] vc: (The duck comes from Monthy Python and the holy Grail…)
[20:03:39] NP: Georg Baresch
[20:03:44] NP: 🙂
[20:04:17] NP: I think it was probably Baresch who is responsible for most of the attempted rescuing of the marginalia
[20:04:32] vc: So you are saying that this last page is from at least 2 or 3 hands?
[20:04:39] NP: But it was Baresch who ensured that it got passed down to us, so we should be grateful to him
[20:05:03] NP: Yes, but on top of each other rather than side by side
[20:05:20] NP: Blow the page up and look again at the varying density of the ink and quills
[20:05:33] vc: We will certainly do that.
[20:06:34] NP: Also, put the various marginalia side by side to scale (all blown up), it’s an interesting viewpoint
[20:06:53] NP: Print them out at A1 scale – huge!

6. About the Hungarians…

[20:06:59] vc: We still have a lot of questions… do you still have some time?
[20:07:14] NP: I have about half an hour now
[20:07:50] vc: Great.
[20:07:51] NP: but I’d like to rest my fingers for five minutes – so, please, tell me what you each think of the VMs (one sentence each, please!)
[20:09:22] vc: (Miklos): Hi. I personally think that this is a kind of constructed language or maybe a written glossolalia.
[20:10:04] NP: it’s still 100x more rational than believing the moon landings were faked 🙂
[20:10:26] NP: even if it is wrong. 🙂
[20:12:57] vc: We are by the way started our searches on a few different pathes. One group is investigating the mythological context of the drawings with girls (if there’s any). The geekz are rather trying to draw some consequences using the digram and trigram entropies, and comparing our results with various 17th century documents, on different languages (and even as exotiic stuffs as Tamil and Hebrew)
[20:13:29] vc: On the other hand
[20:14:23] vc: we’d like to produce some “grammar” here, as well, so, we decided (similarly as Stolfi did it once) to look at the distributions of different words.
[20:14:26] NP: http://ciphermysteries.com/2009/06/18/q13-and-voynich-balneology-sources
[20:14:58] vc: If we concider the first words in the herbal the names of the plants, than maybe that could bring us somewhere.
[20:15:22] NP: There have been many grammars proposed over the years – though I would strongly caution you to treat Currier A and Currier B separately, as the language changed along the way
[20:15:56] vc: Also, we are comparing the Hungarian flora with the drawings, and meanwhile with other old herbals (e. g. ashmole codex). We also about to address a hungarian herbelist in this topic.
[20:16:22] vc: Sure, we’d like to treat the two couriers differently.
[20:16:29] NP: Have you seen the “De Aqua” Voynich theory that was posted to YouTube last month? Its German author has invested a lot of time trying to identify the plants depicted. http://ciphermysteries.com/2009/08/24/de-aqua-voynich-theory-on-youtube
[20:16:48] NP: Erm, don’t shoot the Currier 😉
[20:16:49] vc: Not yet, but we are most curious about it
[20:17:20] NP: If you can read German quickly, there’s plenty of interesting stuff (but a lot of filler too)

7. The f1r paragraph “titles”…

[20:18:13] vc: You can see on f1r that the text is separated to 3 paragraphs.
[20:18:54] vc: Each paragraph ends with some words flushed right.
[20:19:06] NP: Yes… and let’s not forget what John Grove called “titles”, short pieces of text appended to the bottom right of the paragraphs
[20:19:14] vc: is it possible that these are quotations?
[20:19:20] NP: could be anything, who knows?
[20:19:43] NP: As an aside, I should say that the word “possible” sets my teeth on edge
[20:19:50] vc: quotations and the name of the persons who said them
[20:19:58] NP: Probably not
[20:20:28] vc: Not matches with the age?
[20:20:28] NP: My prediction is that they are section titles
[20:21:09] NP: book 1 – agriculture
book 2 – astronomy
book 3 – water
[20:21:10] NP: etc
[20:21:24] vc: hmmmmm
[20:21:43] NP: I would also predict that they relate to the first phase of the VMs’ production
[20:22:20] NP: because the final object has many more sections
[20:22:27] NP: they’re also in A, not B
[20:22:44] NP: but you don’t have to believe me
[20:23:13] NP: I don’t have a gun to your heads
[20:23:28] NP: 🙂

8. Plaintezt language?

[20:23:29] vc: Personally what do you think abozt the plaintext language and the plaintext alphabet length?
[20:23:48] NP: Plaintext language: Italian.
[20:24:05] NP: Maybe German, but if pressed I’ll stick to Italian
[20:24:24] NP: Plaintext alphabet length: just normal, nothing fancy
[20:24:41] NP: I do, however, think that there is something quite special about the enciphering
[20:24:54] vc: namely?
[20:25:21] NP: I think that it was enciphered on wax tablets using a combination of shorthand (to compress it) and verbose cipher (to expand it), such that the overall size of the text matched the original

9. Copy or original?

[20:26:11] NP: I think that there is evidence that the original layout and manuscript features were duplicated, even though the text was enciphered – and so I think that each line of ciphertext corresponds to a line of plaintext
[20:26:23] NP: key evidence for this:-
[20:27:13] NP: (1) the fake vellum hole on f34r – this is a hole that was made in the vellum to copy a a hole that was present in the original copy
[20:27:58] vc: (VBI index of this very feature of your theory?)
[20:28:01] NP: (finding a page number for you…)
[20:28:09] vc: Sorry, could’n miss that one
[20:28:51] NP: (2) f112 has a space on the right hand side. I think that this was a vellum flaw in the original, which was faithfully copied as part of the enciphering
[20:29:35] NP: 10000 VBI points! Ding ding ding!
[20:30:12] NP: Look up close at the hole on f34 – it was rubbed through the vellum, but why?
[20:30:53] NP: Sergio Toresella suggested that the author rubbed through the vellum in some kind of sexual frenzy, but I think he may just have got that wrong
[20:30:55] NP: :O
[20:31:31] vc: VBI index of THIS one is not small.
[20:31:38] NP: Plenty of people have proposed that the VMs is some kind of copy – I just added in some other evidence to say what kind of a copy
[20:31:57] vc: Yes, We have the same opinion.
[20:32:15] vc: It can’t be done in one iteration.
[20:32:28] NP: Essentially, that it’s an enciphered copy but one retaining many aspects of the original layout
[20:33:12] NP: There are also a number of places (particularly in Q13 and the pharma section) where you can see two layers…
[20:33:36] vc: Checkin’…
[20:34:38] NP: f77v (the house at the top)
[20:34:57] NP: f79v (the woman in the pool)
[20:36:39] NP: f88v or the page next to it (curious two-layered spherical jar)
[20:37:13] NP: Sagittarius page – the top left nymph and her odd barrel
[20:37:22] vc: So how these two layers point to the original layout?
[20:37:33] NP: I also think f57v was written in two or more passes
[20:37:50] NP: It’s hard to tell what to think
[20:38:26] NP: What I *suspect* is that there is an odd game of expressing and hiding going on here
[20:39:06] NP: I don’t honestly think that most of the water nymphs have any function apart from distracting your eye from what is really going on on the page
[20:39:22] vc: Mhm

10. The Sagittarius spreadsheet…

[20:39:30] NP: The zodiac nymphs’ poses probably do encode some kind of information, though
[20:40:06] NP: I’m sure a page full of naked women was probably even more distracting to the eye 500 years ago
[20:40:23] NP: talking of which…
[20:40:31] NP: I have a spreadsheet you might like to see
[20:40:41] vc: Sure
[20:40:50] NP: It’s taken from the Sagittarius page of a 14th century manuscript
[20:41:15] NP: and it might just contain basically the same data that is enciphered on the VMs’ sagittarius page
[20:41:42] NP: would you all like to have a look?
[20:41:53] vc: YES
[20:42:06] vc: Take a Y! an E! an S!
[20:42:25] NP: I heard you the first time. 😉
[20:43:27] NP: I should mention that what you have here is an annotated version of a scan of a photocopy of an old b&w photograph that is copyrighted by the Warburg Institute
[20:43:57] NP: I added in the red lines to try to reconstruct the table.
[20:44:24] NP: Which means “please don’t post this on the internet or I’ll get busted”
[20:44:21] vc: Ok, we are seeing the reference. Now what?
[20:44:33] NP: Now, here’s the clever bit
[20:45:00] NP: What the table is encoding is an astrological table per degree.
[20:45:00] vc: ok… won’t post
[20:45:45] NP: That is, each of the 30 degrees has a specific planet, fortune, gender, unlucky day assigned to it
[20:45:58] vc: aham
[20:46:32] vc: so if i transform it into circles, I basically get something very similar semantically to the VMS astro circles?
[20:46:48] NP: Theoretically – the tricky bit is working out how that mapping works
[20:47:32] NP: What it needs is a bunch of clever people looking at it and throwing around ideas for a couple of days
[20:47:49] NP: Which I why I thought of you lot 🙂
[20:48:25] vc: I think we are getting your suggestion. 🙂
[20:48:58] vc: Psp: rectangular to polar
[20:49:20] vc: my very first idea
[20:49:21] NP: The other tables didn’t align half as well as Sagittarius, so it would introduce too much uncertainty into the range of maps to choose from
[20:49:24] vc: of course it’s just me
[20:49:33] vc: Ok, ok
[20:49:42] NP: I think… it might not be the answer
[20:49:52] NP: …but it’s a start 🙂
[20:50:20] NP: Sagittarius is interesting because all the nymphs are facing right
[20:51:02] NP: If there is stuff being encoded in poses / clothes / accessories… look for crossed legs, outstretched arms, arm behind hip
[20:51:15] NP: head-dresses, stars
[20:52:21] NP: Incidentally, I think most of the breasts on the page were added in the second phase.
[20:53:11] NP: However… it may be that the breasts from the first phase (I can barely believe that I’m typing this as a sentence) encoded some kind of information. So you might usefully look for breasts that weren’t added later. :O
[20:53:33] vc: (VBI)
[20:53:50] NP: Bearing in mind that the author seems to have a predilection for adding dots in the first phase and hiding them later
[20:53:55] NP: VBI 100000!
[20:53:59] NP: off the scale
[20:54:12] NP: But… look at it for yourself, and make up your own mind
[20:54:29] vc: ROTF
[20:54:52] NP: At least you have something to compare with the page that nobody else (apart from me) has seen in the last 550 years
[20:55:04] NP: Real data (even if does have drawn on breasts)

11. Finishing up…

[20:55:41] NP: OK, team Voynich Budapest
[20:55:50] NP: I have to go now
[20:56:03] NP: I hope you’ve all broadened your view of the VMs
[20:56:12] NP: And not been exposed to too much VBI
[20:56:22] NP: 🙂
[20:56:23] vc: Thank You very much indeed for the most fruitful conversation.
[20:56:36] vc: Not enough i must say
[20:56:46] NP: Plan A is to return for a second session on Sunday
[20:56:55] NP: Is that correct?
[20:57:11] vc: Yes
[21:01:06] NP: Right – good luck, everyone & talk with you on Sunday!
[21:02:02] vc: Ok, all the best until then
[21:02:06] NP: byeeeeeee
[21:02:13] vc: bye 🙂

Voynich Quire 8… what happened?

by ⋅ 14 Comments

I’ve just had a nice email from my old friend GC, asking what I think happened with Quire 8 (“Q8”). You see, the problem is that Q8 contains a whole heap of codicological oddities, all of which fail to join together in a satisfactory way:-

  • f57v has a bottom-right piece of marginalia that (I think) looks rather like “ij” with a bar above it – yet it’s not one of the quire numbers, and doesn’t appear on the back of a quire.
  • f66r has some bottom-left marginalia (the “mus del” nymph): yet unlike most similar Voynich Ms marginalia doesn’t appear on the front or back of a quire.
  • The first (f57 + f66) bifolio contains both circular diagrams and plants
  • The second (f58 + f65) bifolio contains two text-only pages and two herbal pages
  • f58r and f58v have stars linked to most of the paragraphs: but these have no tails, and hence are more like the “starfish” and “stars” found in Quire 9 (Q9) than the paragraph stars used in the recipe section at the end.
  • The page numbers on f65, f66, and f67 all appear to have been emended by a later owner (you can still see the old faint 67 to the right of the new 67)
  • And don’t even get me started about the circular diagram on f57v (with the repeated sequence on one of the rings). Put simply, I think it’s not a magic circle, but rather something else completely masquerading as a magic circle.
  • But sure: at the very least, f57v’s circular diagram would seem to have much more in common with the circular diagrams in Q9 than with herbal quires 1-7.

Generally speaking, though, Q8 seems to be broadly in the right kind of place within the manuscript as a whole. Because its bifolios contain both herbal and diagrammatic stuff, it seems to “belong” between the herbal section and the astronomical section. However, the bifolios’ contents (as we now see them) appear to be rather back-to-front – the circular diagram and the stars are at the front (next to the herbal section), while the herbal drawings are at the back (next to the astronomical section).

This does suggest that the pages are out of order. And if you also look for continuity in the handwriting between originally consecutive pages, I think that only one original page order makes proper sense: f65-f66-f57-f58. When you try this out, the content becomes:

herbal, herbal, text, herbal, herbal, // circle, stars + text, star + text

Where I’ve put the two slashes is where I think the first (herbal) book stops and the second (astronomical) book begins: and I believe the “ij-bar” piece of marginalia on the circle page is one owner’s note that this is the start of “book ij” (book #2).

So, I strongly suspect that what happened to Q8 was a sequence very much like this:-
1. The original page order was f65-f66-x-x-x-x-x-x-f57-f58
2. The bottom right piece of marginalia was added to f57v (start of book “ij”, I believe)
3. The pages were mis-/re-bound to f66-f65-x-x-x-x-x-x-f58-f57 -OR- (more likely) the front folio (f65) simply got folded over to the back of the quire, leaving f66r at the front: f66-x-x-x-x-x-x-f57-f58-f65
4. The nymph & text marginalia were added to f66r.
5. The pages were mis-/re-bound to f57-f58-x-x-x-x-x-x-f65-f66.
6. The quire numbers were added to f66v.
7. The page numbers were added to all the pages.
8. The central three bifolios were removed / lost.

But what happened to pages 59 to 64, which apparently got lost along the way?

Currently, my best guess is that these were never actually there to be lost: there is practically no difference in quill or handwriting between f58v and f65r, which suggests to me that they originally sat adjacent to each other… that is to say, that Q8 probably only ever contained two bifolios. And so, the proper page numbers added (at 7 above) were probably 57-58-59-60, which would make perfect sense.

So… why were they later emended to 57-58-65-66?

My suspicion is that, temporarily bound between Q8 and Q9, there was an extra tricky set of pages, which the page-numberer skipped past before continuing with 67 (in the astronomical section). But what tricky block might that be?

Could it have been the nine-rosette fold-out section? Might the page-numberer have skipped past that, before subsequently noticing that an earlier owner had given it a higher quire number, and so moving it forward to its proper place? It’s a bit of a tricky one to argue for, but I do strongly suspect that something in someone’s system broke down right around here, causing more confusion than we can easily sort out.

However, I’ll leave the nine-rosette section for later: that’s quite enough codicology for one day! 🙂

Navigating Voynich Manuscript f1r…

by ⋅ 3 Comments

Seeing the Voynich Manuscript for the first time is quite an intimidating experience: you’re looking at something which is so uncertain in so many different ways – how should you try to “read” it?

In general, when you look at a page of text, you do two different types of reading: (1) you work out how everything is laid out (you navigate the page) and (2) you read what is contained within it (you read the text). In computer science terms, you could describe the layout conventions and text conventions as having two quite separate ‘grammars’.

For instance, if you picked up a Hungarian newspaper, I would predict that you would stand a good chance of being able to work out its structure, even though you may not be able to understand a single word. It’s perfectly reasonable, then, to be able to navigate a page without being able to read it.

What’s not widely known about the Voynich Manuscript is that researchers have identified many of the navigational elements that structure the text (even though they cannot actually read them). I thought it might be helpful to post about these (oh, and I’m getting emails mildly berated me for posting too much about the wrong ‘v’, i.e. that it’s not “Vampire News“).

As a practical example, let’s look at the very first page of the manuscript proper: this has the name “f1r” (which means “the recto [front] side of folio [double-sided page] #1″). You may also see this referred to as “f001r” (some people use this naming style so that their image files get sorted nicely), or even as “1006076.sid” (this is the Beinecke Rare Book & Manuscript Library’s internal database reference for the high-resolution scan of f1r, which they store as a kind of highly compressed image). This is what f1r looks like:-

Note that the green splodges aren’t actually part of the page itself – they’re green leaves painted onto the reverse side of the folio (that is, on f1v, “folio #1 verso [back]“) that happen to be visible through the vellum. I’ll leave the issue of whether this is because the paint is too thick or the vellum is too thin to another day…

If we use a tricky colour filter written by Jon Grove (more on it here), we can make a passable attempt at removing the green splodges: and if we then bump up the contrast to make everything a little clearer, we can get a revised image of f1r:-


Red areas: these form the first four paragraphs of the text. These often start with one of four large vertical characters (known as “gallows characters”), and appear to have been written from top-left down to bottom-right, as you would English, French, Latin etc.

Blue areas: these are known as “titles”, and are typically right-aligned words or short phrases added to the end of paragraphs. It has been proposed that the text contained in these might actually be section titles (which seems fairly reasonable). There’s a brief discussion on this by (a differently spelled!) John Grove here, who first suggested the term.

Yellow area: this is a cipher key arranged vertically down the right hand side of the page that someone has written in (and only partially filled before giving up) in a 16th century hand. Though a bit indistinct, you can still make out “a b c d e” at the top left and a few other letters besides.

Bright green areas: these odd shapes appear nowhere else, and are generally referred to as “weirdoes” (for want of a better name). Interestingly, these are picked out in bright red: f67r2 is the only other place with red text that I can think of (the page that was originally on the front of what is now Quire 9).

Dull green area: this is where the earliest proven owner wrote his signature (something like “Jacobus de Tepenecz, Prag”, though it is very hard to make out), which a subsequent owner appears to have (quite literally) scrubbed off the page (if you look carefully, you can see what appears to be two or more watermarks at the edges of the area). The question of why someone would want to do this is a matter for another day…

Pink area: hidden in the top right corner next to some wormholes and the folio number (“1”, in a sixteenth century hand) is a very faint picture, possibly of a bird. Surprisingly, this subtle piece of marginalia doesn’t appear in GC’s otherwise-very-good gallery of Voynich marginalia: so here’s an enhanced picture of it so you can see what I’m talking about:-.

So, even if we can’t yet read f1r’s text, can we navigate its layout? I believe we can! From the presence of red text, I’m fairly certain it was the first page of a quire: and from the signature and weathering, I don’t see any reason to think this was ever bound anywhere apart from at the front of the manuscript. This leads me to predict that the set of four paragraphs forms an index to the manuscript as a whole, and so very probably describe four separate “books” or “works”, where the “title” (appended to the end of the paragraph) is indeed the title of that book.

If you were looking for cribs to crack the titles 🙂 , my best guess is that the first book (section) is a herbal, the second book is on the stars (astronomy and astrology), the third book is on water, while the fourth book comprises recipes and secrets. I also suspect that this index page was composed about three-quarters of the way through the project, and that the (really quite strange) Herbal-B pages were added in a subsequent phase. But, once again, that’s another story entirely…

Hidden writing on f1v…?

by ⋅ Leave a Comment

I’ve spent a long time (though “far too long” probably covers it better) hunting down obscured fragments of text in the Voynich Manuscript: so my Spidey-sense tingled almost uncontrollably when I saw a claim for hidden text on f1v in the “Marginal Writing” picture gallery on Glen Claston’s Voynich Central.

I’d never heard of this before: just to be sure, I checked Reuben Ogburn’s 2004 page on “Writing in and around plant illustrations” in case it had slipped in there, but no sign. If you run this through Jon Grove’s colour separator filter, you can see that the brown ink used for the drawing and the brown paint used to fill it in are very slightly different: in the image below, the white area is where the filter thinks the overpainting happened.

But is there writing beneath? If you squint at the topmost image here long enough, you can start to make out something that might almost be writing. But if you filter it slightly differently, I think the answer emerges: the “signal” (below) appears to be not writing, but only compression artefacts from the MrSID wavelet encoding. Sorry, guys: false alarm! (Though next time I’m at the Beinecke, I’ll have another quick look, just to be completely sure…)