In Part 1, I looked at lots of different types of evidence that each may (or indeed may not) say something about what happened to Quire 20’s bifolios to leave them in their final nesting order. So, what are the different microtheories these suggest? (What I call “microtheories” are attempts to reconstruct microhistories (e.g. about a specific feature), rather than an overarching macrohistory.)

Microtheory #1: “f116v is Q20’s final page”

It’s easy to conclude from f116v’s strange (and I suspect heavily emended) “michitonese” and doodly shapes that it must always have been Q20’s final page. Certainly, the presence of wormholes on f114/f115/f116 implies that the bifolio nesting/ordering we see now for Q20 was probably still in place back in the days when the Voynich Manuscript had a wooden cover.

However, Wladimir’s wormholes tell a rather more complicated story: because there is a substantial set of wormholes between f114 and f115 (smaller on f114, larger on f115) that doesn’t appear on f116, it strongly seems as though the present state was preceded by a different ordering (though probably also with a wooden cover) where f115 was Q20’s rearmost folio, with f114 nested immediately inside f115.

So it seems that while this particular microtheory was true for the final ordering, it was very probably not true for the ordering that preceded it. However, it doesn’t help us determine whether f116 was originally the final folio. True, it looks like it could have been: but there’s no obvious supporting evidence.

Microtheory #2: “f105r was originally Q20’s first page”

The ornate gallows at the top left of f105r certainly feels as though it ought to have been marking a visual statement of some kind, certainly more so than the rather less whizzy f103r. The fact that f105r has more ‘titles’ seems to back up the idea that this is an important page: and we might reasonably wonder whether the bottom edge of f105 was heavily trimmed (as Anton noted) to remove some kind of ownership mark in the bottom margin.

However, the suggestion that this was excised to remove the ownership mark is (unless someone can dig up any indentations on a different recto page, which – though a little unlikely – might prove possible with some kind of fancy modern scan) just plain speculation. Hence we only really have the ornate gallows and the presence of ‘titles’ to suggest that this was the case. Which is not to say that they’re bad evidence: rather, they’re both just a little weak.

Microtheory #3: “Contact transfers confirm the current ordering”

It’s entirely true that contact transfers (particularly paint contact transfers) seem to support Q20’s current nesting/ordering. I’m thinking here of the contacts between f115v and f116r; between f104v and f105r; and between f113v and f115r (via the vellum gap in the outside edge of f114).

And yet, I suspect that because the paint contact specifically between f115v and f116r is inconsistent with the wormholes highlighted by Wladimir Dulov, this would seem to imply either that much of the painting was done late in the manuscript’s binding state, or that even though the earlier bifolio nesting/ordering state was lost for an intermediate binding, it was magically restored for the final binding.

Similarly, many years ago I went through (pretty much all) the paint contact transfers: only a single instance (on f2v/f3r, see Curse Chapter 4) convinced me that contact between pages had transferred paint from the original binding. Everywhere else, it seemed that paint had transferred when the manuscript was in a later bound state.

So I don’t really believe that the (very few) paint contact transfers we see in Q20 can help us draw conclusions about Q20’s earliest nesting / ordering state, alas.

Microtheory #4: “f58r and f58v are the real start of Q20″

This microtheory gets mentioned only rarely, and yet there’s a great deal to be said for it. If you reverse the f58-f65 bifolio (i.e. turn it upside down, as VViews suggested here), f65 would appear to be the final folio of a Herbal-B section, making f58r likely the original start of the recipe section.

There’s a lot to commend this microtheory. For a start, f58 is the only page outside Q20 that has paragraph stars (though admittedly quite different-looking ones from what we see in Q20 ‘proper’). And Lisa Fagin Davis categorises f58 as “Scribe 3”, the same scribe who wrote ~95% of Q20.

f58r also has a space for an ornate capital at the top left of the page (presumably duplicated from a source page’s layout?), which makes it look as though it ought to be the start of a section / chapter / book (MarcoP mentioned this in 2018, but I’m pretty sure this was noted long before):

Incidentally, even though Lisa Fagin Davis categorises both f58 and f65 as having been done by Scribe 3, there seems to be a visual discontinuity between the two writings, to my eye making the text on f65v look more like Scribe 5’s work on f66 (though perhaps it was using a different quill & ink?):

Furthermore, in Julian Bunn’s folio similarity map, f58r and f58v are clustered closely together (so are probably content-continuous and/or structure-continuous): and they’re not too far away on his map from (say) f105r and f115r. Hence there seems to be no obvious technical reason blocking this general idea.

There are also some aspects of f58r that make it appear a little bit special. For example, f58r and the top half of f58v are made up of chunky paragraphs; f58v has a ‘title’ conspicuously at the end of the first paragraph; while f58r has more words starting with “chol-” than any other page by quite some margin (though what this means I have no idea, I just thought you’d like to know):

It also has more words starting with “chal-” than any other page (a lot of the rest are in Q20):

f58v also has an odd-looking sequence of single-leg gallows on its top line, if you’re into that kind of thing:

So, might f58r have actually originally been Q20’s start page? Personally, I think it’s not a bad microtheory at all, really not bad at all.

Having said all that, I would note that, unlike Q13 and Q20, f58r and f58v contain only a single word beginning with “l-” each. Similarly, while f58r and f58v have more instances of “ed” than is usual for Currier A, they have far fewer “ed” instances than Currier B. Compare this with f66r (in Q8), a Scribe 5 page which is simply stuffed with words that start with “l-” (particularly at the bottom), as well as many more words containing “ed”:

Hence this might well be one of the ‘intermediate’ Currier-language folios that Rene Zandbergen has mentioned at various times, e.g. we might be able to see aspects of Currier-A ‘evolving’ into Currier-B here.

Microtheory #5: “Q20 was originally made up of Q20A and Q20B”

Though I’ve blogged numerous times about this microtheory, it’s still not definitively proven. While it does (from the wormholes) now seem likely that f115v (rather than f116v) was the end folio for much of the Voynich Manuscript’s physical history, we can’t yet prove that f116v was not also the original end folio. And even if the microtheory that f105r was the first page of Q20 now seems less likely than the microtheory that f58r was Q20’s original first page, we can’t use either to prove or disprove this.

Perhaps, though, this will necessarily be more of a culmination / accumulation / combination of other codicological arguments than a supporting argument for them. The jury is still out here.

In Part 1, I collected together a whole load of evidence relating to Q20’s bifolio order that previous analyses had proposed. I thought I’d add a post suggesting other analyses that haven’t yet been done. Who knows, maybe someone will give one or more of these a go?

Vellum thickness

The class of gadget you would use for measuring the thickness of vellum is a non-contact thickness gauge (these are like double-sided laser/LED measuring devices). But why would that be interesting?

The reason is that this offers a low-cost (and non-destructive) way to match up different bifolios’ vellum edges, particularly their long edges. As I understand it, a single sheet of vellum would typically have been large enough to be cut into four bifolios (a very common gathering size): and so there is strong reason to suspect that the vellum sheets that made up the Voynich Manuscript would have been used to form individual quires. (Though very probably not the quires that we see now.)

My hunch is that mapping the Voynich Manuscript’s bifolios’ vellum thicknesses would help give us a little more information about how the original set of gatherings were manufactured and assembled (before their order was later scrambled, probably several times). I believe that this could cast light on Herbal-A pages, Herbal-B pages, Q13 pages and Q20 pages.

Sure, it would take a little bit of planning and a fair bit of subtle reasoning afterwards, but in terms of resources, it would cost relatively little. So why not try?

DNA comparison

This idea – of sampling the DNA of all the different bifolios, and then cross-referencing them – was something I described here back in April 2022. It still strikes me as an attack well worth attempting.

I should note that Michelle Lewis commented (pointing to a voynich.ninja discussion) that she thought that what I was suggesting was a little ahead of the technology – maybe it is, maybe it isn’t. There is certainly a Bovine Genome Database online (the first Hereford cow DNA was sequenced in 2009, says Wikipedia), so there may already be researchers out there who have suitable ’23AndMoo’ skills needed to, ummm, give us a steer. (Did you see what I did there?)

Lisa Fagin Davis too left a comment expressing (what I think came across as) an even more negative opinion of the idea, which was she thought that it would (basically) give no useful information. For what it’s worth, my opinion of her opinion was that there seems to be convincing evidence that the Voynich Manuscript’s support (i.e. its vellum) was prepared (and indeed cut down) in idiosyncratic (or at least unusual) ways, which would seem to point to very much the opposite of the scenarios she invoked to argue against the idea. Still, maybe she’ll come round to the idea in a decade or so.

Revisiting Julian Bunn’s page distance metric

In Part 1, I mentioned Julian Bunn’s Voynich page distance metric webpage: but on reflection, I don’t think I really highlighted how promising this kind of study could be when applied to bifolio reordering.

For example, we can see from his folio clusters map that f106v and f106r are not really that similar to each other, which suggests a change of topic / section / chapter somewhere in mid-f106. In fact, if we look at f106’s two sides, we can see what seems like a change of ‘tempo’ in the last two paragraphs of f106r, as well as a mysterious unstarred paragraph on f106v (paragraph 8, “tshod qokchy…”).

But while these high-level statistics suggest content anomalies, I don’t believe we have yet found a way to determine which specific low-level features drive those high-level statistics. In the case of f106r/v, can we identify the exact paragraph where the dictionary contents typical of f106r ‘flip over’ to the dictionary contents typical of f106v? If this difference is something we can measure, it should be something we can audit, right?

Incidentally, Julian effectively defines his metric as (Number of unique words that appear in both Page X and Page Y) / (Number of unique words in (Page X union Page Y)). I like this because it doesn’t “over-reward” high-frequency words such as qokedy (etc). However, perhaps a really good (and visually simple) tool would be something that allowed you to compare two pages by automatically drawing lines between matching unique words (while also excluding high-frequency words). I suspect that this might help you visually see where content blocks start and end.

Regardless, the other follow-on question here is simply: what recto page originally followed f106v? It of course has to be a recto page: and the ‘nearest’ recto page would seem to be f104r (and by quite some margin). This suggests that in the original nesting order, f106 may well have been followed by f104.

Here again, I think it would be helpful to visualise how the ‘dictionary intersections’ between the two pages work in practice.

I thought it would be a good idea to post up as many of the different research threads relating to Quire 20’s (‘Q20’) original bifolio nesting / ordering in a single place as I could. However, be warned that there are… quite a lot of them.

Ornate Gallows

There is only a single ornate (swirly) gallows in Quire 20. Yet oddly, it’s to be found not at the start of the quire (as you might expect), but instead at the top of f105r (i.e. the front of the third folio in):

So, if your basic assumption about Q20 is that its bifolios are in the correct order, then it would seem that this ornate gallows poses you a problem: it ‘feels’ like it’s in the wrong place.

However, as we’ll see, there are plenty of other types of evidence we can look at, and – annoyingly, but probably unsurprisingly – they all yield a slightly different spin on the same basic question of ordering.

Wladimir’s Wormholes

In a very interesting recent post, Wladimir Dulov noted various patterns of wormholes on Q20, particularly on f114, f115 and f116. One cluster of wormholes starts large on f116 (the end page), becomes slightly smaller on f115, and smaller still on f114: this seems to imply that these were made when the bifolios were in their current (final) nesting order.

As an aside, book-worms are actually woodworms, and so don’t – as Rene Zandbergen has pointed out – like eating vellum. Wormholes in vellum typically mean that the book had had a wooden cover which the worms had eaten through first, before carrying on munching into the vellum (then stopping after a couple of pages).

Incidentally, I’ve long suspected (from the curious quire numbering style and the careful vellum repairs) that the Voynich Manuscript may well have spent time in a Swiss monastery library (and, I’ve argued, probably not too far from Lake Constance): and my guess is that this was probably where it had a wooden cover added. (Perhaps even in a chained library [“Kettenbibliothek”] such as Schaffhausen.)

So far, so wormy. But Wladimir’s nice point is that there is also a second set of wormholes on f114 and f115, which doesn’t go through to f116. In fact, f114 and f115 have an abundance of wormholes (not just the ones Wladimir mentions) not visible on f116, or indeed on any other Q20 page.

This would seem to imply that even though f116 has ended up being bound as the final folio, f115 may well have previously been bound (in a previous binding) as the final folio, with f114 nested just inside it. Additionally, I suspect the absence of matching wormholes on any other Q20 bifolio also weakly implies that at the time when the two outermost bifolios were f114 and f115, none of the other Q20 bifolios was then nested just inside f114. This suggests that the bifolio that was then nested inside f114 might possibly have been the missing Q20 bifolio (f109-f110).

However, from the random bifolio shuffling that seems to have gone on early in the Voynich Manuscript’s life, I’m also fairly certain that when it entered that library, the manuscript was in the form of a set of unbound (or perhaps only very lightly-bound) gatherings. So, even though Wladimir’s wormhole reasoning is sound, I suspect it only takes us back to the manuscript’s wooden cover days, i.e. it doesn’t necessarily tell us about the original bifolio nesting order etc.

All the same, his observation does signal fairly loudly that Q20’s final bifolio order (that we see today) is very likely wrong, giving us confidence that trying to reconstruct the original order is a sensible idea.

Wladimir’s Pre-binding

Wladimir has another interesting observation about Q20. He points out a puncture mark near the bottom of each of the bifolios f103, f104, f105, f106, and f107 (he is unsure from the scans whether or not there is a matching puncture mark on f108, but thinks that there probably is). He describes this as a “pre-binding mark”, where a binder has run a small piece of twine through the bifolios to line them up ready to bind properly.

This would seem to have happened just before the final binding we see today, so it’s quite late on in the overall codicological timeline. However, because there seems to be no evidence of pre-binding elsewhere in the Voynich Manuscript, I can’t help but wonder whether this weakly implies that Q20 was bound separately to the rest of the manuscript.

Vellum Tricks

Back in 2016, I posted about the vellum colour of each of the Q20 bifolios. Here, contrast enhanced, are the vellum colours of f103-f116, f104-f115, f105-f114, f106-f113, f107-f112, and f108-f111:

My observation back then was that f103-f116 and f106-f113 seemed different from the others two bifolios: hence it seemed likely to me that the four were cut from a single (large) piece of vellum. It therefore further seemed likely to me that these other four originally sat next to each other.

(Of course, we could do better than my eye by physically sampling DNA from these bifolios and comparing their sequences, but there currently seems to be no appetite for doing this.)

To my mind, this implies that there were probably two quires / gatherings:

• Q20A – f105-f114, f104-f115, f107-f112, and f108-f111
• Q20B – f103-f116 (almost certainly on the outside), f106-f113, and perhaps the middle bifolio

Anton and f105

Voynich researcher Anton [Alipov] (whose name you may recognise from voynich.ninja) commented on Wladimir’s brief summary comment of the above, noting that:

Folio 105 is also somewhat excessively trimmed from the bottom which looks strange.

It’s a neat, clean upward cut across the bottom that continues across to the other half of the bifolio. Without close physical examination (Lisa, have you looked at this?), it’s hard to be sure whether this was in the vellum right from Day One, or whether it was cut off at a later date.

As far as I know, the two main things that typically get added at the bottom of pages are quire numbers and ownership marks. So that suggests to me that there may well have been some ownership mark added to the bottom right of f105r (or the bottom left of f105v) which a later owner wanted to remove. (Not all ownership transfers are transactions later owners want to advertise, as the heavily erased Sinapius signature on f1r seems to attest.)

Similarly, the last folio of Quire 19 (i.e. facing f103r) seems to have had a large chunk taken out of it, which is consistent with an ownership mark there also being excised: this possibly suggests that that may have been the last page of a book / section.

(Incidentally, I pointed out in 2010 that I thought that f105v shows more sign of weathering than just about every other page in Q20, which would seem to imply that it spent a good period of time on the back of the quire. However, looking at f105v again now, I’m not really sure what I was seeing back then.)

This might make the bottom part of f105v the likely location for a quire mark (in one binding), and also the bottom part of f105r the likely location for an ownership mark (in the reversed binding). What a codicological mess!

Unusual Glyph Patterns

Despite its ‘language’ similarities to other Currier B pages, Quire 20 also has a number of glyph pattern idiosyncrasies. In a 2010 comment here, Tim Tattrie pointed out:

“lo” as a separate word is only found in f104r, 106r and 108v.

“rl” as a separate word, or word beginning is only found in f104r, 108v and 113r.

“llo” as a series of letters is only found in f104r, 108v,111v, 113v,116r

Looking at these in voynichese.com, there seem to be just as many free-standing “lo” words in Q13 as in Q20, so I’m not quite taken: and the number of “llo” glyph sequences is extremely low (5 matches).

I’d add that most of the places we see “lr” are in Q20: and similarly for “dl”. There are also some instances of “dr” clustered on f105v, and similarly for “dd”. Q20 is also where the Voynichese glyph “x” appears most often. Sean B. Palmer also thought that the only “genuine” occurrence of “aa” in the Voynich Manuscript was on the third line of f115r (which ends “cholor daar oraro”).

Generally, I do think that the glyph content of Q20 words seems a bit more ‘variable’ than Q13 words, but the overall pattern doesn’t seem wildly different. Unless Rene has some stats on this I don’t know?

Repairs to f116

An interesting 2015 blog post by David Jackson takes a look at the holes, waterstains and repairs on Q20’s final folio (f116). He speculates (for several different reasons) that f116 was trimmed down to the bare minimum in response to damage to that page.

Broadly speaking, I’m not completely convinced by his argument: the main waterstain (near the top of the folios) is visible all the way through Q20, and vellum – animal skin – is basically waterproof. I think David highlights some interesting features, but they don’t quite fit together for me the way they do for him.

Contact Transfers

There are a few places in Q20 where we can see colour transfers between adjacent pages:

• Paint from the top two red paragraph stars on f116r has transferred to the facing page (f115v)
• Paint from the sixth red star on f113v has transferred past the vellum edge flaw on f114 to f115r
• There’s a stray red paint spot that appears near the top of f104v and f105r
• There’s a stray faint paint spot between f114v and f115r (look between stars #10 and #11)
• Green paint (presumably from Q19?) has ended up on the outer edge of f104r, partly because f103 seems to be about 1cm narrower than f104. (And because f103-f116 is a bifolio, this 1cm different also suggests that the outside edge of f116 was not trimmed down, i.e. that was how that bifolio was originally cut.)

Contact transfers from the paragraph stars highlight the issues of (a) whether the paragraph stars were drawn in the original construction phase, and (b) whether the paragraph stars were painted in the original construction phase – because if they were, then f115v originally faced f116r and f113v faced f115r (with f114 optionally in the middle).

Lisa Fagin Davis’ Scribes

As far as Lisa Fagin Davis’ well-known (and much-cited) scribal analysis of the Voynich Manuscript goes, Quire 20’s palaeography might seem to be one of its less interesting features. She writes (p.17):

The entire Quire is written by Scribe 3 with the exception of folio 115r, where the first twelve lines were written by Scribe 2.

These twelve lines span four short paragraphs, and look like this:

For our present challenge of reconstruct the bifolio nesting order(s) of Q20, all this really does is suggest that f115r might possibly have been the original start of a quire or a section. But you’d almost certainly need to combine this with more information to form the outlines of a proper argument.

Q20 Titles

In Voynich researcher terminology, a “title” is a short sequence of Voynichese text that is positioned on a page in a slightly anomalous (and non-paragraphy) way. For example, everyone knows that f1r (the very first page of the Voynich Manuscript) has four of these ‘titles’ (on lines 6, 10, 21, and 28), a fact that has given rise to the broadly-held speculation that the paragraphs on f1r with titles might be using them to hold section / chapter / book names (in some way). Page f8r similarly has 3 titles (on lines 8, 13, and 21)

Q20 also has some of these Voynichese titles, according to John Grove’s somewhat ancient list:

• f105r, line 9
• f105r, line 36
• f108v, line 52
• f114r, line 34

As with the scribal information, it’s hard to be sure what exactly to make of this: but it certainly makes f105r seem like a page of structural interest, title-wise.

Dictionary Distance Metric

Back in 2010, Julian Bunn posted up an interesting page computing distance metrics between folios in terms of how similar their unique word lists were. I emailed him some comments (which he then incorporated into the web-page) on what this had to say specifically about Quire 20. Here’s what I sent:

Having played with Julian’s results a bit […], it appears that while some pages’ recto and verso sides are very similar, others are wildly different. For example, just in the recipe section:-
103    good
104    very bad
105    very good
106    bad
107    excellent
108    excellent
109    (missing)
110    (missing)
111    excellent
112    good
113    excellent
114    excellent
115    very bad
116    n/a

Looking at pages within recipe bifolios, however, yields different results again: for example, even though both f104 and f115 are both “bad” above (and are on the same bifolio), f104v is extremely similar to f115r, while f104r is extremely similar to f115v (which is a bit odd). Furthermore, the closeness between f111v and f108r suggests that these originally formed the central bifolio (but reversed), i.e. that the correct page order across the centre was f111r, f111v, f108r, f108v. However, f105 / f114 seem quite unconnected, as do f106 / f113 and f107 / f112.

Re-reading my comments 12 years later, it’s clear that there’s quite a lot of suggestive information here. For example, even though the two bifolio pairs f105r/f105v and f114r/f114v are close, the disparity between f105 and f114 weakly suggests that their bifolio sat towards the outside of a quire or gathering.

Similarly, the closeness between the two halves of the f104-f115 bifolio suggests that this may possibly have been a central bifolio (though what precisely is going on remains something of a mystery).

Finally, the fact that f106r/f106v aren’t close but their opposite folio half f113r/f113v are very close suggests that there may have been a change of topic or structure somewhere on f106.

What Have I Missed?

I tried to follow Wladimir’s discussion of Quire 20 “gaskets” (as Google Translate put it), but wasn’t really able to (I believe it’s to do with things attached to the spines of individual quires that were used to bind them to a spine but without gluing them). However, I don’t believe this affects the issue of bifolio ordering I’m trying to tackle here.

I’m unaware of any multi-spectral scans of Q20 that we might be able to refer to so that we can compare the vellum used on different bifolios.

Also, I’m unaware if there are any studies that have specifically used page-to-page word list difference metrics (of the kind Julian Bunn did) to exhaustively evaluate all the different nesting permutations, i.e. to suggest what the original nesting order was.

Regardless, if you know of any other analyses of Q20 that might have some impact on the bifolio nesting order, please mention it in the comment section below, thanks very much!

After all the factuality of Part 1 and the opinionicity of Part 2, it’s time to actually start attacking Quire 20’s block. The issue of how exactly to attack remains tricky, but let’s give it a go and see how far we get.

Ditto?

In part 2, I argued for the presence of some quite specific parallels between Q20’s recipe-like blocks and the recipes in Lat 6741. One particularly interesting comparison was between Philip Neal’s “horizontal Neal keys” (pairs of single leg gallows, typically found 2/3rds of the way along the top line of paragraphs) and the second coloured initial in Lat 6741 that typically separates a recipe’s title (e.g. “ad faciendum…” etc) from the main body text of that same recipe.

However, not all recipe titles are the same kind of length. Notably, there are some short ones that use “Ad idem” to mean “ditto“, i.e. ‘use the same title as the preceding recipe‘. You can see these in the following Lat 6741 recipes:

#6…

#30…

#37…

#44…

#62…

#80…

#84…

#88…

Can we see any hints of this “ditto”-like behaviour in the top line of the Voynich Manuscript’s Quire 20? We can certainly see many Q20 paragraphs that resemble this kind of thing:

f103v:

f104r:

f105r:

f106v:

f107r (note the pair-heavy EVA “aralorar” in the first word of the line):

f107v:

f112r:

f113r:

On balance, though, the honest answer is… not really. In fact, even though there are plenty of lines that kind of resemble the ad idem recipe pattern, the sheer variety of word types in, between, and around these single-leg gallows is really quite… impressive. And confounding. And infuriating. But here we are.

A Duplicated Recipe?

Villela-Petit (in “Copies, Reworkings, and Renewals”) points out (p.170) that Alcherio accidentally copied the same recipe from Brother Dionysus twice (“Si vis in colore viridi tingere ossa, ligna, tabulas, scutellas ligni, manubria cutellorum, filum et pannum lini“, recipes 40 and 81), though with the second word (“vis”) misplaced and a few minor spelling differences:

To be fair, I expect that Brother Dionysus himself had copied the two by mistake, and that Alcherio and le Begue retained the duplicate in their scrupulously copied copies.

Might there be a similar duplicated recipe anywhere in Q20? Having access to the same source text that has been manipulated in two different ways within the same overall system may well be extremely revealing, cryptologically speaking, like a kind of “internal Rosetta Stone”.

We’d be looking for a single-paragraph recipe where the title is about 25% of the length of the paragraph (2 lines out of eight). #40 is the third of three slightly longer paragraphs in the middle of medium-sized paragraphs (6.5 6.5 6 // 10 8 8.5 // 6 5.5 11, while #81 is the first of three slightly longer paragraphs after three short paragraphs (3.5 5.5 4 // 8 8 9.5 // 3 4 7).

Generally, I’d expect to see a Voynichese paragraph about five or six lines long, like this one (on f103v):

However, I must confess that I struggled to find even one paragraph (let alone two) that met all (or even most) of the criteria I had in mind. I’ll revisit this in a few days’ time (when I have a clearer head).

Thoughts, Nick?

I think it’s fair to say that the way I’ve previously described block paradigm attacks has been fairly abstract and non-specific: but it is actually something a bit tricksier than just a ‘known plaintext attack’. Rather, it’s more like a kind of scribal jigsaw puzzle, where you’re hoping to glimpse some sign of a scribal structure or ‘meta-pattern’ peeking through the edges of whatever confuddlery has been applied to its plaintext. This post is hopefully only the beginning, I suspect it will take a while to really get under Q20’s skin.

As always, our attacks have all the less ‘bite’ for not having managed to reconstruct more of the original state of the manuscript. Villela-Petit was even able to reconstruct various aspects of the original state of Brother Dionysus’ two gatherings (p.170), but frankly we have a long way to go to catch up with that level of reconstruction.

In Part 1, I tried to foreground everything you’d need to know about Quire 20 and BnF Lat 6741 in order to make it possible to try to systematically match their two respective recipe structures. I also tried to put my own thoughts and assessments to one side for this Part 2 post. So let’s get to work…

Visual Comparisons

As we saw before, the recipe section in Lat 6741 starts with a marginal number and a scribal flourish.

The start of (what I call) Q20B (on f105r, near the start of Q20) isn’t wildly different, visually speaking:

It has an extravagant-looking gallows initial character, and pretty much an identical number of lines. The “tailed star” is what I guess means “ytem”, a kind of medieval marginal ‘bullet point’, normally found within a list of items.

Moreover, it may just be a coincidence but the pair of single-leg gallows in the top line aligns strikingly well with the start of the red “Experimenta de coloribus” part of the Lat 6741 line.

It’s certainly easy to see why other people have previously suggested that what we are looking at in Q20 might well be a set of recipes, or possibly a set of 360 astrological predictions / prescriptions (as Tiltman suggested in 1975, mentioned in D’Imperio).

Horizontal Neal Keys?

Furthermore, this does make me wonder whether the top-line behaviour observed by Philip Neal (where we see pairs of single-leg gallows often appear two thirds of the way across the top line of a Q20 paragraph) might in fact be specifically mirroring structural behaviour we see often in Lat 6741. For example, its recipe #3 looks like this:

What we’re seeing here is a short recipe intro section (“ad faciendu[m] azuriu[m] finu[m]”), followed by an inserted coloured capital that precedes the body of the recipe. This basic pattern appears throughout Lat 6741’s recipes:

• coloured capital (red or blue)
• title
• coloured capital (blue or red)
• body

Compare this with the third paragraph of the Voynich Manuscript’s f105r:

Here we can see (if you ignore the “title” line inserted at the right hand side) a single-leg gallows at the start of the line, with more single-leg gallows (in fact, there are three of them) just about halfway through the paragraph’s top line.

Note, at this point I’m not claiming that what we’re seeing in the above is a translation or obfuscation or encryption of Lat 6741. However, I can’t help but suspect that what we are seeing here might well be structural parallels between the two documents (or, at least, between two classes of document).

Numbered recipes

One other aspect that strikingly emerged from the comparison was the numbering: the Arabic numerals Jehan le Begue’s used in his 1431 copy were a few decades ahead of the general adoption curve – and yet, because of the gaps in the numbering, it seems highly likely to me that these were gaps in the numbering present in Giovanni Alcherio’s original copy. All of which is why I suspect Alcherio’s original recipes were numbered (which I’m also guessing was why they appeared at the start of Jehan le Begue’s copy, so that le Begue could extend the numbered series into his own collection of recipes).

However, what I suspect we’re seeing in Quire 20 is (though I of course can’t prove it) more like an unnumbered list of “ytems”. Does that means we are looking at a completely different document? Or perhaps at a copied document where the marginal numbers were removed or lost in the copying? It’s hard to reason based on only a single text, however generous le Begue’s internal annotations were.

Materiality

My nextinferential jump is a little hard to explain, but it’s ultimately about materiality. Giovanni Alcherio was an illuminator, and asked lots of artists for their colour-making tricks, i.e. how to get best-in-class performance out of raw writing materials. Similarly, the way that the Voynich Manuscript’s vellum was prepared was (I think) very much trying to squeeze the best ‘performance’ out of not-quite-top-league support material, in a way that – to me – speaks of someone who also understood the nature of scribal material, and who considered it an important facet of production.

Putting these two together, I’m broadly quite liking the idea that Quire 20 may have contained scribal secret recipes of the sort that Alcherio and le Begue collected: that is, someone who cared about the quality of the support material must surely have also cared about the quality of the writing materials.

As an aside, a number of manuscript studies have applied modern spectroscopic analytical techniques to inks and paints to try to narrow down the precise recipes used (for example, both Nancy Turner and Ines Villela-Petit have done this). According to the Yale photographic facsimile edition of the Voynich Manuscript’s chapter on “Physical Findings”, the only blue pigment found was coarsely ground azurite with traces of red iron oxide (hematite, Fe2O3) and/or (as per McCrone) red cuprite (Cu2O), as well as (unusually) barium. Zyats et al (who wrote the chapter) suggest these “may be related to the geological origin of the azurite”, though the painstaking work of identifying where in the world this actually was was, umm, generously left as an area of “future research” for the reader. (pp. 34-36, plus footnotes)

(A quick Google search revealed that a Cornell group [Smieska et al] has been working on identifying the location of different azurites using the presence of barium and barites, but I don’t know how their work has advanced since 2016/2017.)

Which Metric(s) to Match?

Anyway, moving back to the block paradigm attack, which metrics should we consider collecting and comparing?

Though, as per Part 1, my first instinct was to compare [number of lines] in each recipe, I’m also now considering collecting and comparing [distance to first single-leg gallows] with [distance to first embedded coloured capital].

Incidentally, if you use voynichese.com as a quick way to highlight all words containing EVA f or EVA p on f111r (the page Rene Zandbergen highlighted as having many ‘probably fake’ paragraph stars), I think it makes it starkly clear why the top section there ‘reads’ so unlike the rest:

Here, even though the stars are almost all “tailed stars”, there’s a dearth of single-leg gallows as compared to the rest of Q20. If we broaden our viewpoint to include f108r, f108v, f111r and f111v (all on the same bifolio), I think you can clearly see the change in paragraph ‘tempo’ from the middle of f108v to the lower middle of f111r that I mentioned in Part 1, along with the fake-feeling area at the top of f111v:

As a different (more visual) ‘style’ of structure matching, it might therefore instead make sense to print out thumbnails of Lat 1641 and compare the above with patterns of coloured capitals, to see if the comparison reveals anything with the Q20 p/f word map on voynichese.com (printed out as per-folio thumbnails).

Interestingly, I should note that a strong structural match might imply the preservation of structure between documents even though the underlying text itself has been translated (never mind encrypted). That is, even if everything goes to plan here, we still might have no direct proof that the Voynichese is an encrypted version of the text in the other – the text itself may have been translated (e.g. to Italian or French) before being encrypted or obfuscated.

In that kind of case, though, it’s possible that the best route through would be to look for extremely distinctive words in the text (cribs) that would survive even translation and encryption. But let’s worry about such a theoretical problem only at such time as it becomes an actual problem. 😉

No, I’m not blogging about a Joe Cornish / John Boyega medieval mash-up, but about applying my block paradigm attack approach to the Voynich Manuscript’s Quire 20.

This is a “known plaintext attack“: in non-crypto-bro English, it means ‘working out whether a given text is (somehow) the plaintext of the last section of the Voynich Manuscript’, even though we can’t read a word of the latter.

To do this, we’d need a pretty strong candidate text: in fact, we’d want one with broadly the right kind of structure, broadly the right kind of length, and which (preferably) might well have been considered a “trade secret” circa 1400-1450. But which we now have a copy of.

Since 2014 or so, my #1 candidate has been Jehan le Begue’s 1431 collection of colour-related recipes, which famously appeared in Mary Philadelphia Merrifield (1849). These recipes are in Latin and in French (plus Merrifield’s sons’ English translation), though Jehan le Begue helpfully notes that most were copied from Giovanni Alcherio’s collection of colour-related recipes (though le Begue added more of his own in French), even listing the sources that Alcherio listed. And so I’m arguably slightly more interested in Alcherio’s collection than le Begue’s.

However, this kind of begs the question: how can you be sure of the structure of an enciphered text? Making this even harder is the fact that there is convincing evidence that many of the Voynich Manuscript’s bifolios have ended up shuffled around (and according to no obvious pattern).

Hence this post starts with what we know about Quire 20, before moving on to Giovanni Alcherio’s collection of recipes. As always, wrangling all the individual pieces into one place is extraordinarily time-consuming, so that’s what this first post concentrates on doing – Part 2 will try to bring all these together to do the actual attack. (Which is arguably 10x harder. But you have to start somewhere.)

Notes on Quire 20

From my numerous blog posts on Q20 (starting with 2010), I have floated numerous tentative conclusions about how Q20 ‘works’, none or all of which may be useful in this context:

• Tailed paragraph stars may well be a steganographic ‘y’, short for ‘ytem’ (i.e. a bullet point)
• Tail-less paragraph stars may have been added to make the tailed paragraph stars less obvious
• Q20 was probably originally two separate quires/gatherings, that were later shuffled together
• f105r (with the ornate gallows at top left) was probably the start of Q20A
  • I would be unsurprised if the Voynich ‘title’ at the bottom of f105r was Q20A’s book title
• f103r was probably the start of Q20B (i.e. with f116v as its last page)
  • The stars on f103 and f116 are almost all tail-less
• The right margin gap on f112 is probably a copy of a vellum tear in the document it was copied from
  • (I covered this in Curse 2006)
• Elmar Vogt flagged an empty-full star-colour pattern, which f103, f104, and f108 didn’t conform to
  • I wondered whether that implied f103, f104, and f108 were originally bound together
• Tim Tattrie pointed out (among other things) links between words on f104r and f108v
  • I wondered whether that implied f108v and f104r faced each other (original page order)
• Rene Zandbergen pointed out in 2016 that many of the stars on f111r look to be fake
  • I collected together more such comments in this 2021 blog post

TL;DR – even though earlier Voynich researchers usually thought of Q20 as a single ‘thing’, it instead seems to have started life as two or more separate ‘books’. There also seems to be a category difference between tailed star paragraphs and non-tailed star paragraphs, with the former possibly denoting the start of an item.

Quire 20 Item / Paragraph Structure

Note that f109 and f110 appear to have been the two halves of a central bifolio that got removed when the manuscript was a couple of centuries old or so (say, after 1600 but probably before 1700). (However, there’s currently no obvious reason to presume that it was a central bifolio in the original (‘alpha’) bifolio nesting order.)

• f103r: 18 x no-tail, then 1 x odd “top tail”
• f103v: 14 x no-tail
• f104r: 13 x tail
• f104v: 13 x tail
• f105r: fancy gallows, then 9 large tails
• f105v: 10 x tail
• f106r: 15 x tail (#3 has a tiny ‘child’ star’)
• f106v: 14 x tail
• f107r: 15 x tail (#11 has “…” next to it)
• f107v: 15 x tail
• f108r: 16 x tail
• f108v: 16 x tail (note that 7-8 & 11-16 seem fake, #10-#16 seem to be a single paragraph)
• f111r: 17 x tail (note that 2-12 & 14 seem fake, #1-#12 seem to be a single paragraph)
• f111v: tail, no-tail, 2 x tail, 4 x no-tail, tail, 5 x no-tail, tail, 4 x no-tail (#2-#8, #10, #17 seem fake)
• f112r: 7 x tail, no-tail, 4 x tail
• f112v: 5 x tail, no-tail, tail, no-tail, 4 x tail, no-tail
• f113r: 16 x tail
• f113v: 15 x tail
• f114r: 13 x tail
• f114v: 8 x tail, no-tail, 3 x tail (#5 seems emphasized)
• f115r: tail, no-tail, 11 x tail
• f115v: 13 x tail
• f116r: 10 x no-tail, followed by two large unstarred paragraphs (like a colophon)
• f116v: (end-page)

We can also rearrange these same lines by bifolio (rather than by sequentially numbered folio):

• f103-f116 bifolio
  • f103r: 18 x no-tail, then 1 x odd “top tail”
  • f103v: 14 x no-tail
  • f116r: 10 x no-tail, followed by two large unstarred paragraphs (like a colophon)
  • f116v: (end-page)
• f104-f115 bifolio
  • f104r: 13 x tail
  • f104v: 13 x tail
  • f115r: tail, no-tail, 11 x tail
  • f115v: 13 x tail
• f105-f114 bifolio
  • f105r: fancy gallows, then 9 large tails
  • f105v: 10 x tail
  • f114r: 13 x tail
  • f114v: 8 x tail, no-tail, 3 x tail (#5 seems emphasized)
• f106-f113 bifolio
  • f106r: 15 x tail (3rd star has a tiny ‘child’ star’)
  • f106v: 14 x tail
  • f113r: 16 x tail
  • f113v: 15 x tail
• f107-f112 bifolio
  • f107r: 15 x tail (#11 has “…” next to it)
  • f107v: 15 x tail
  • f112r: 7 x tail, no-tail, 4 x tail
  • f112v: 5 x tail, no-tail, tail, no-tail, 4 x tail, no-tail
• f108-f111 bifolio
  • f108r: 16 x tail
  • f108v: 16 x tail (note that #7-#8 & #11-#16 seem fake, #10-#16 seem to be a single paragraph)
  • f111r: 17 x tail (note that #2-#12 & 14 seem fake, #1-#12 seem to be a single paragraph)
  • f111v: tail, no-tail, 2 x tail, 4 x no-tail, tail, 5 x no-tail, tail, 4 x no-tail (#2-#8, #10, #17 seem fake)

Commentary: as with the Herbal bifolios, there are often unexpected consistencies to be found between the contents of two folios where they are part of the same (attached) bifolio. The most obvious example of this is f108v, f111r and f111v, which all have large paragraphs and what appear to be fake stars. Yet f108r and the first five paragraphs of f108v seem to be quite different (they’re more ‘metronomic’, small paras regularly followed by other small paras). I can’t help but wonder whether there is a change in recipe ‘style’ part way down f108v: and also whether the f108-f111 bifolio may have originally been the central bifolio of a guire / gathering.

It also seems that we have three categories of starred paragraph to wrestle with: starred ‘item’ paragraphs, non-starred paragraphs, and starred non-paragraphs (i.e. fake stars, which may or may not have a tail). So I think we have to be very much on our toes when trying to draw inferences about stars.

More generally, there are occasional changes in ‘tempo’, e.g. when dense small-para pages with tiny tight stars (such as f111v) get followed by not-so-dense pages with larger paras and fewer stars (such as f112r). These give me the strong impression that we’re not looking at a single, uniformly-structured list of items, but rather at several different kinds of item (i.e. with different text styles) that have ended up interleaved. Moreover, the presence of fake-looking stars (as flagged by Rene) looks to me as though the author may have been trying to conceal some aspects of the very structure I’m trying to discern. But maybe this is a good sign, and that – as Sherlock Holmes once said – “The game is afoot! Not a word!“

Jehan le Begue Bibliography

The manuscript itself (Lat. 6741) is in the BnF, and is accessible online here. It’s marked “Ex Libri Lud[ovico] Martelli Rx 1587). It starts with a long table of synonyms, often with alternating red-blue capitals, e.g. part of fol. 2r looks like this:

This is transcribed on Merrifield p.18, but just so you can get acclimatised to the writing, I reworked it below to include the (entirely typical) early 15th century scribal abbreviations visible above:

• [A]zurium vel lazuriu[m] est color ; aliter celestis vel celes[-]
  tinus, aliter blauccus, a[li]t[er] pers[us], et a[li]t[er] ethere[us] dic[itur].
• [A]uru[m] est nobilius metallu[m] croceu[m] colore[m] habens et
  tenuatur in petulis, quo carentes utunt[ur] stanno
  attenuato et colorito colore croceo et in petulis tenuato.
• [A]rgentu[m] est nob[i]le metallu[m] album colore[m] habens, quo
  qui caret utitur ejus loco de d[i]cto stanno tenuato, non colorito.
• [A]uripigmentum est color croceus qui al[i]t[er] arsicon dicit[ur]

This is followed by:

• Experimenta de coloribus
• Experimenta diversa alia quam de coloribus
• Liber Theophili admirabilis et doctissimi magistri de omni scientia picturae artis
• Liber Magistri Petri de Sancto Audemaro de coloribus faciendis
• Eraclii sapientissimi viri liber primus […]
• De coloribus ad pingendum capitula scripta et notata a Johanne Archerio seu Alcherio anno Domini 1398 […]
• Capitula de coloribus ad illuminandum libros ab eodem Archerio sive Alcherio scripta et notata anno 1398 […]
• Aultres receptes en Latin et en Francois per Magistrum Johannem dir Le Begue […]

Mark Clarke’s (2001) “The Art of All Colours” p.101 includes a half-page description of Lat 6741, and also notes a complete 19th century transcription in BL MS Add. 27,459. The most important reference (crazily omitted from the BnF description) is the major part of Volume 1 of Mary Philadelphia Merrifield (1849), who not only transcribed 6741 (adding a voluminous introduction), but also had her sons translate it into English (“except [for the] Theophilus portion”, Clarke points out). Clarke also notes that a transcription appears in van Acker (1972) “Petri Pictoris Carmina”, pp.143-198 and 242-246. A more recent edition was by Inès Villela-Petit (1995), presumably in her PhD dissertation (which I haven’t seen).

For accessible sources on Alcherio / Alcherius, I’d heartily recommend:

• “The Recipe Collection of Johannes Alcherius and the Painting Materials Used in Manuscript Illumination in France and Northern Italy, c. 1380-1420”, by Nancy Turner
• “Copies, Reworkings and Renewals in Late Medieval Recipe Books”, by Inès Villela-Petit, and translated by Jilleen Nadolny. (Available on academia.edu.)

Navigating Lat 6741

The recipes in Lat 6741 have been sequentially numbered (with occasional gaps) in the left margin, which gives anyone discussing them a helpful starting point. For now, I’m going to restrict this discussion to those 118 recipes that Jehan le Begue copied from Alcherio (or else this would end up insanely large):

• Experimenta de coloribus
  • 1-46: Latin. Copied by Alcherio in 1409 from “an unbound [quire] lent me by Brother Dionysius […] at Milan.” Note that these recipes use an alchemical-sounding code for colours: “Sol” is for gold (i.e. yellow), “Luna” for silver (“the rust of which is azure”), “Mars” for iron (“the rust of which is violet”), “Jupiter” for tin, “Venus” for copper or brass (“the rust of which is green”), and “Saturn” for lead (“the rust of which is a white colour”).
  • 47-88: Latin. Copied by Alcherio from a second unbound [quire] lent by Brother Dionysius. #48 onwards is “Experimenta diversa alia quam de coloribus”.
  • 89-99: French. Copied by Alcherio from recipes lent to him in Bologna by an embroiderer called Theodore of Flanders, who had in turn procured them in London.
  • 100-116: Italian. Copied by Alcherio in Bologna in 1410 from a book of Magister Johannes de Modena. (Jehan le Begue copied these, and then had a friend translate them into Latin.)
  • 117: Latin. Copied by Alcherio in Venice in 1410, from Michelino di Vesuccio, “the most excellent painter among all the painters of the world”.
  • 118: Latin. Copied by Alcherio in Paris in 1410, from Master Johannes de […something…]

This is also because these 118 recipes seem to have the highest “trade secret” rating of all the recipes given by le Begue: most of the rest were either centuries old or French recipes added by le Begue himself.

Recipes #1 to #46 (fol 22r to fol. 27v)

Diving straight in, recipe #1 looks like this:

Note the recipe number in the left margin, and the title of the section / book embedded in the top line in red. Merrifield transcribes this recipe on her p.47, which (reconstructing the abbreviations above) would look like this:

1. [N]ota q[uod] auree Experimenta de coloribus
li[tte]re scribu[n]t[ur] sic, cu[m] ista aqua ; accipe sulphur vivu[m] et
corticem int[er]iorem mali granati, alum[i]nis, saltis, et de plu[-]
via auri, tantu[m] q[uan]tu[m] vis, et aqua[m] g[u]mmi liquide, et modi[-]
cu[m] de croco, et misce et scribe.

Note, the number of lines in Lat 6741 for #1 to #46 (to the nearest half-line) I counted are:
4.5 4.5 3.5 3 5 6.5 11 2 7 17 13 19 11 14 5 3.5 13 4.5 19.5 6.5 5 9.5 3 7.5 5.5 4.5 9 4 6 4.5 11.5 6 10.5 3.5 5.5 6.5 6.5 6 10 8 8.5 6 5.5 11 6.5 6

Recipes #47 to #88

The number of lines for #47 to #88 (note that le Begue has some numbering gaps) I counted are:

20 4.5 5 3.5 9 5.5 8 4 4 15 9 4 12 9.5 13.5 4 8.5 7 16 4.5 4 (#69 missing) 2 (#71 missing) (#72 missing) (#73 missing) (#74 missing) 5.5 3 1.5 (#79 missing) 3.5 5.5 4 8 8 9.5 3 4 7 3 2

Recipes #89 to #99

These are in French, and have a noticeably different format, with a header preceding each recipe:

Merrifield transcribes this (p.85) as follows:

89. Pour faire l’eau noire. – Prenez une pinte de l’yaue de dessoulz la meule sur quoy on meult les courtesaulx, et la mettes sur le feu, et gettez ung voire de vin aigre, et ii onces de galles, et prenez demie onche d’alon, et une onche de coperose, et le faitez tant boulir, qu’il apetice du tiers, et puis le laissier reposer un jour.

Recipes #100 to #116

These were originally written in Italian, but were translated into Latin by a certain friend of Jehan le Begue.

Recipe #117

25 lines long

Recipe #118

122 lines long, which – compared to the rest of the recipes – is a bit of a monster.

I hope everyone who attended the Voynich Conference 2022 hosted online by the University of Malta enjoyed the presentations and the Q&As.

In Lisa Fagin Davis’ final presentation, she mentioned her recent theory that p/f were in fact ke/te: and mentioned that she’d thought this up, but then found it on Cipher Mysteries. If you want to see the original page I put up in September 2020 suggesting this idea (along with her comment near the bottom), it’s right here, along with the August 2020 page where I started exploring the behaviour of single-leg gallows.

There’s an additional aspect to the set of gallows/e/ch groupings I discussed in 2020, which is that you can usefully compare the (parsed) ch:chch ratio in the text as a whole (which is 10616:18 (0.17%)) both to the (parsed) ratios of strikethrough gallows preceded by ch…

• ckh:chckh = 634:242 = 38.17%
• cth:chcth = 766:139 = 18.15%
• cph:chcph = 185:27 = 14.59%
• cfh:chcfh = 58:15 = 25.86%

…as well as to the (parsed) ratios of strikethrough gallows followed by ch:

• ckh:ckhch = 871:5 = 0.57%
• cth:cthch = 902:3 = 0.33%
• cph:cphch = 211:1 = 0.47%
• cfh:cfhch = 73:0 = 0%

This, too, is a strikingly asymmetric result; and would seem to suggest that the ch:chch ratio is practically identical to the ch:c<gallows>hch ratios, yet completely unlike the ch:chc<gallows>h ratio.

I would take this as reasonably good support for the idea that c<gallows>h is actually a visual proxy (and it doesn’t really matter whether this is for scribal, cryptographic, or steganographic reasons) for <gallows>ch, because Voynichese seems to want to avoid c<gallows>hch almost exactly as much as it wants to avoid chch.

Perhaps combining this result with the pe/fe result (and other “forbidden” Voynichese combinations) might be the start of something really positive…