How are we ever going to resolve the mystery of the Voynich Manuscript?
Even sixty years ago, it was abundantly clear to William Friedman (widely believed to be the greatest ever codebreaker) that the Voynich’s cipher/language was of a different type to anything that had previously been encountered, and that normal cryptanalytical tools would be of little use in revealing its secrets. If Voynichese was ‘gettable’, Friedman and then Brigadier John Tiltman (who I personally place right up there with Friedman) would surely have got it, or would have begun to get it: but neither did. Yet high quality scans, several decades of study and the Internet’s million eyeballs have not yet yielded insights significantly beyond what Friedman and Tiltman managed.
And so we fast forward to 2014, where researchers continue to adopt one of two basic paradigms: (a) that Voynichese uses an unknown cipher, and so we must grind through an endless series of statistical tests that must surely eventually isolate an answer, and (b) that Voynichese is written in an unknown language, and so we must extract a set of individual word cribs to identify the language’s family etc.
After a lot of consideration, my opinion is that the Voynich Manuscript laughs in the face of both approaches: and that if you’re using either (or indeed both) of them, you’re almost certainly wasting your time. Moreover, by spamming people with the results of your research, you’re wasting their time too.
Please understand that I’m not saying that these two paradigms are worthless in all circumstances: rather, I’m saying as flatly and directly as I can that though they do have great value in other contexts, they are essentially worthless when applied to the Voynich Manuscript. They have not worked, do not work and never will work for it: hopefully that’s a clear enough position statement.
So… if they don’t work, what’s the alternative? Ay, there’s the rub.
The Block Paradigm
With all the above in mind, it became apparent to me a while back that what was actually missing was an entire paradigm, by which I mean a complete and systematic way of thinking about what we are trying to do with Voynich research, with what tools, and for what purpose. Hence I’ve spent most of this year trying to work out what that new paradigm should be, and to find a good way of communicating it.
(This has basically been why Cipher Mysteries has been so quiet as far as the Voynich Manuscript goes.)
My working title for this new way of thinking about the Voynich Manuscript is the “Block Paradigm”, because as its target it seeks to identify not a letter, a word or even a language, but a block of text. That is, the idea is to use three stages:
(1) identify blocks of the plaintext that stand some chance of appearing in other manuscripts;
(2) find possible matches for them in other manuscripts; and then finally
(3) attempt to reverse-engineer the way that the two blocks map to each other.
As such, the Block Paradigm is entirely neutral about whether Voynichese is a cipher, a shorthand or an unusual language (or any combination of the three): the purpose of the first two stages is to get us to the point that researchers can attempt to do the reverse-engineering third stage for a given block with reasonable confidence that they may have something that could well be the plaintext.
Over the next few weeks, I plan to introduce and discuss a number of candidate blocks from the Voynich Manuscript. Right now, I only have a possible matching plaintext block for one of them (and that’s a particularly complicated block that I’ve been working with for some years), so this is only really the start of what I hope will be a broadly collaborative and productive process.
I’ll start with a candidate block that has already been discussed by Voynich researchers, though not nearly as fully as I think it deserves.
The Poem Block
On 26 Apr 1996, Gabriel Landini posted to the Voynich mailing list:-
Folio 76R has a full body of text, that is, the page is almost completely filled with text. I can guess 4 paragraphs. However if we go to folio 81R, then the text looks like if it was written as a poem. Some lines longer than others and there is no drawing that is restricting the line lengths. I wonder if the pages like 81r are songs, hymns or prayers…
Rene Zandbergen replied (on 29 Apr 1996):
If I remember well […] f81r is unique is this respect. To me, it gives the impression as if a drawing was intended for the right margin, but it has been omitted. This is at variance with the generally accepted theory that all drawings were done first and the text filled in later. Another special feature of this page is that there seems to be a connection of the drawing in the left margin, through the binding gutter, to the opposite ‘page’ (which would be f78v if my above assumption for 81r is correct). […]
The ‘poem’ idea fits well with Currier’s observation that the line seems to be a functional entity. This again is strange in view of the fact that in many cases the text is clearly organised in paragraphs, i.e. several full lines followed by one short line and sometimes a somewhat larger gap between this and the next line. The length of each last paragraph line seems to be anything from one word to a full line. Note that when it is one word, it is sometimes centered on the line.
With the benefit of better scans and 18 further years of study, let’s look again at this possible ‘poem’ on f81r and annotate what we see:-
I’ve marked the two horizontal Neal keys (at the top of the two main sections of text) in red, some extraneous-looking text appended to the final line (possibly a date, or a signature?) in green, two almost identical gallows-initial words in blue, and various unusual features of the text in purple.
The text itself seems to have a line structure of 7 / 8 / 8 / 8 lines (as per the ornate line-initial gallows-initial words): so if this is a poem, my prediction is that the poem was originally 8 / 8 / 8 / 8 lines but that an entire line got omitted by mistake during the copying. (Copying ciphers by hand is surprisingly hard to get right).
As for the likely subject matter of the poem, that too isn’t too hard to predict: given that it is sandwiched between two drawings of naked nymphs in baths, it would surely shock nobody if the poem turned out to be about water or baths.
I would also expect the poem to be in Latin or Italian (OK, Tuscan), because those were (as I recall) the languages typically used for balneological poems pre-1450.
So what might the poem actually be? You might think it could be a section grabbed from Peter of Eboli’s famous early medieval Latin poem “De Balneis Puteolanis”: but that, according to C. M. Kauffmann’s “The Baths of Pozzuoli” (p.14) consists of thirty-seven sections (an introduction, thirty-five sections each covering a different bath, and a dedicatory section at the end), each section having precisely twelve hexameters. Hence I suspect we can eliminate De Balneis Puteolanis as a candidate simply because it has a different verse length – we’re looking for eight or sixteen line sections, but it is based around twelve-line sections.
All of which is where my preliminary account of this poem block comes to a halt: I simply don’t know the balneological literature well enough to take this any further forward. Were there any pre-1450 balneological poems written with eight or sixteen lines per section? I believe that this question should probably be the starting point for researching this particular block; but what do you think?
Dude,
It’s Copernicus… Hence the flip-flop, this way and that, helio-centric/geo-centric… all the languages (Hebrew, Greek, Latin, Slovak, you name it)… and in Church-proof code.
Please, at least consider it.
Scott
And if the hypothetical poem is of the Voynich Manuscript author’s own composing? What of your efforts, then?
Also I think “methodology” is a more appropriate term than “paradigm” here. Underlying assumptions do tie into it, but the important thing here is the method.
This is different what from I had in mind in my comment about progress; but we all do agree that there’s something fundamentally flawed about what the “Voynichology” field is doing and how it’s doing it, that has held us back from substantial breakthroughs in its 102 (and counting) year existence.
I often liken the personal efforts to the Columbus expedition. Large investments of time and resources into the deep blue yonder, with no real idea of what to find, how to reach it, or where to start looking for it. Could find nothing and have wasted a few months. Could find something but not realise it. Could hit a jackpot. No way to be sure, it’s a gamble. What I’m trying to say is, though I wish you luck and hope it works out, don’t be surprised if the Voynich Manuscript “laughs in the face” of your approach too. This thing has a nasty habit of rending the best laid plans of mice and men no matter how wise their conception.
Nick,
Admittedly, the compliment mayn’t be worth much, given that I’m ignorant of anything to do with ciphers etc., but all the same I think this approach is really a stunning new angle on something that has recently seemed stale.
Not to be depressing, but the imagery in this section is rather more sophisticated and allusive than mere illustration, and while I think it probably serves as aid to memory of the more detailed information in the written text beside it, it also stands alone – and my opinion of its content you may recall from 2009 or -10 or something.
Also, thanks so much for referring to Rene’s comment on paragraphs. I’ll refer to it in future as (possibly?) the first notice of a feature which is not common in western books made before the introduction of printing.
I think: you’d be lucky if any such pre-15th century poem has survived. Good luck tho’!
OK, I was thinking about this.
Medieval Spanish (Castillan and Catalan language) poems were often written in eight line verses called octavillas if they didn’t adhere to a strict rhyming structure or octava italiana if they did (generally ABBC’ DEEC’ structure) or octava real if they had a strict “arte mayor” style, generally ABABABCC structure. Anyway, the eight line poem was popular in the Middle Ages, 1100-1400’s across the Christian part of the Iberian Peninsula.
But there is no evidence of rhyming structure in the text. Let’s look at the ends of the first octavilla:
0x9
0M9
89
ax
aw
ox
aw
0x9
No evidence of rhyming structure there that I can see, but I know very little about poetry.
To link the VM text to a plantext we’d have to identify repeated words in the ciphertext and then find a plaintext with the same number of repeated words. If we found that, we’d have struck gold – but there just aren’t enough repeated words in the VM for that strategy to be feasible.
David: there is actually a large literature on medieval balneological literature, and it’ll take a little while to review it. So we shall see!
David: I understand what you’re saying, but there’s a very large presumptive leap in it – that you already understand the nature of Voynichese well enough to be able to reject things. Whereas the whole endpoint of this approach is to determine the nature of Voynichese.
What I’m aiming for (and I now realise that my post didn’t really describe this) is to find internal structural details that match the internal structural details of a matching text: by which I mean unusual words, jarring sequences, repetitions, whatever. Basically, things that would be awkward to express. It’s a tricky point to get across, so I’ll try to cover this in a follow-up post.
Brian: there’s methodology in there, sure: but there’s also a large amount of orientation and mindset that goes beyond that, that relates to what we’re trying to do when we look at the text. If you have a better all-encompassing word, please say. 🙂
The Columbus expedition had a very clear idea of what to find and how to reach it. It just so happened that the calculations were wrong and a quite different continent was in the way. 🙂
I fully expect the Voynich to laugh in the face of this approach too, but… the person or persons who made it couldn’t possibly know what other documents would survive to the present day. And we only need a single match to find a systematic way in!
Brian – but what if the hypothetical author of the Voynich Manuscript is a creature of your own composing? 🙂
Seriously, so much fantasy-history has been woven around the object that we can forget much of it. What we do know is that few of the remaining poems in manuscripts from before 1440 are written by the person inscribing them. Some, but not very many, and the fact they were copied means that more copies may have been made, and another might exist.. why not?
Nick – the word you’re looking for, although I’m not sure if you want so prosaic a description, is a crib. No need to reinvent the name – cryptologists have been using cribs to break unknown encryption systems since…. well, you know that as well as I do. Similar how the US code breakers used to match Japanese navy messages to the weather report for each day to see if they could match anything up, knowing that the Japs would send the weather forecast via the encryption systems.
David: I know the word well, but I wanted to differentiate what I’m doing from the low-level way people usually think of cribs (i.e. “HEILHITLER” in Enigma messages). That is, we’re not trying to find settings for a known cipher here, but to find a substantial block of parallel text and to use that to reverse engineer an entire writing / enciphering / shorthand system.
Hi Nick, I am not too sure what this new paradigm will add. If you think that trying to identify the name of a plant on a herbal page is perilous, this will be many times more. Not only will you have to argue that a whole text matches, but also that it is the right language, the right subject, and so on. I am not even sure f81r is a poem.
Also, I do not think that the finding of single word cribs is the only way to work on a linguistic theory. I don’t do that at all, and do not intend to do so for a long while. Maybe I should share my paradigm?
Thing: I would argue that trying to identify a block rather than a word is many times less perilous, because it should (I believe) be many times less susceptible to the kind of wishful thinking and soft-matching fallacies that plague the word-by-festering-word polyglottist approaches of people like Stephen Bax.
If you have an approach that is even better than the Block Paradigm and that also manages to avoid imposing lots of presumptions onto the text, please share, I’d be delighted to devote a post specifically to it.
Scott: thanks for your suggestion, I’ve now considered it. 😉
Nick, I feel it is more perilous because whereas a single plant name is likely to a) appear somewhere in the text and b) be translingual (so similar reflexes of the same word often occur in many language), you may end up bringing in a lot of text in the wrong language or which doesn’t appear. For example, you might find a likely poem in Latin, but even if the text on f81r is the same poem it could be a paraphrase in Tuscan. You could end up making many more guesses.
Methodologically speaking, when we’re looking to crack the kind of document that the Voynich manuscript likely is, we have three unknowns: script/cipher, language, and content. We should seek to hold two of those static while working on the third. When Champollion read names in the cartouches of the Philae obelisk and Rosetta Stone, the cartouches guaranteed the content and personal names made the language irrelevant. I don’t think blocks of text will hold either of these two static to allow careful work on the script/cipher.
That all said, I do not think that the single word method is workable either, at least not without a great deal of groundwork. Should I email you my paradigm, Nick? Critique from you and others would be very welcome.
So if I’ve got it right, put crudely the idea is:
a) deduce the likely content or origin of a section of text, primarily (but not necessarily) from the accompanying illustrations;
b) identify a likely original source from where this text may have been taken or adapted, and
c) work out how b) led to a).
Now, my memory isn’t wonderful, but I think I recall a discussion on here a couple of years ago which might be relevant. The question came up as to whether the illustrations in the Voynich Manuscript were sufficiently distinctive that they could be traced back to specific parallels. I’m pretty sure that the answer was a no, and from memory that ‘no’ was firmly and robustly expressed.
Does this ring a bell with you?
Thing: ah, the beauty of the Block Paradigm is that you don’t need to know the language. The idea is like Intellectual History or Art History, that the text must have flowed from somewhere (hopefully another document that still exists in some form): and if we can sufficiently narrow down the range of things we’re looking for, we test whatever we find. In the case of the Poem Block, I’d happily put forward balneological poems in any language for analysis and comparison.
Hence I think we can just hold the content and structure static and let the script/cipher and language take care of themselves. If we genuinely uncover the source text and there’s a mapping to be found, I reckon we stand a good chance of success, it’s as simple as that. 🙂
SirHubert: I think it would be true of the herbals in the Voynich Manuscript, because they have been closely analyzed by people with a great degree of domain expertise, Sergio Toresella and Karen Reeds being perhaps the two best known.
Arguably the closest thing there is to a “Herbal Block” is Rene Zandbergen’s example of f35v, which he says appears to be “inspired by an illustration seen in the earliest versions of the Tractatus de Herbis and copied in later issues“. But I don’t currently know what text accompanies that page in the Tractatus de Herbis or in the various versions found with alchemical herbals: perhaps that would be worth looking more closely at with this Block Paradigm in mind.
In keeping with your block theory, the Voynich may be a crib for an older oral story.
xplor: it’s possible, but I think it seems more likely that the Voynich Manuscript collects together information from a number of different sources, some of which may well be oral, but most of which are probably not.
Thing: hundreds of people believe this kind of thing, but why is it that we cannot find those plant names jumping out at us? To date, we can’t even find vowels jumping out at us, let alone entire plant names. Something’s wrong with that picture: it ought in theory to be true, but in practice it just doesn’t work.
All the same, please feel free to email me your paradigm – nickpelling a.t nickpelling d.o.t c.o.m etc. 🙂
Dear Nick, Xplor, & Thing:
All along (these past two years at least) I have been comparing various manuscripts which presently bear the names of the owners and/or final recipient museums (Borgia/Florentino/Badianus, to name a few). So, why would we settle for “Voynich” as a name for manuscript 408 which is currently in the possession of the Boenicke Library at Yale?
I continue my translations of that manuscript irregardless of who wrote it — though I am comparing every folio with Friar Sahagun’s manuscript which began with his diary and his early membership with the Franciscans.. His diary section and rough draft botanical section of his huge manuscript would have been separated from the portion dealing with the everyday lives, customs, superstitions, and work, of the native populations/tribes.
Terms like ‘tomatillo’, ‘osquash’, ‘scabiosa caucasica, ‘aconitilium’, psyllium, ‘dianthus’, ‘sericine’, and ‘Monte Alban’ appear in ms B-408. The name ‘Voynich’ appears nowhere.
Just trying to keep on track — and filing some 25 translations and downloads into my small file cabinet.
beady-eyed wonder
The 1450 era of Europe was a time of political conflict. The Byzantine Empire was about to fall, ending Islamic math and trading units as set down in Fibonacci’s LIBER ABACI and
Vulgate Latin had replaced by a higher form of Italian by the Dante movement. Dante had reconciled the cosmology of Rome and Islam, as Fibonacci has reconciled trade and the math of the region, an era that was about to end.
A new form of cosmology based on Copernicus and new forms of scaling rational numbers to decimal notations were privately discussed.
Machiavelle’s “Prince” by 1510 discussed ways to replace medieval serfdom run by Princes by creating Principalities such that space for Republics to be nurtured.
The arithmetic code topic was standardized by 1585 AD before Galileo standardized the cosmology model, that removed man from the center of the universe. Spanish writers like Cervantes added to the debate by placing Sancho as a leader and decision maker on penniless island. as Shakespeare discussed broader political topics.
Coding making was huge from 1500 to 1600, based on a range of purposes. Finding the correct one, or a small set of options, for this case under discussion may need to be discussed before a fresh approach can gather the needed team to solve this vexing set of coded documents.
Wow this thread has exploded!
Nick: The “continent in the way” is what I mean. They had some vague intention but no concrete facts or specific routes. And once they found something, they had no way to confirm if they found exactly what they were looking for (they didn’t). Once you identify a potential match, how will you decide if it’s worth attempting the mapping, based on anything other than number of lines? Even worse, what if you find multiple potential matches? How will you decide which one to try? I’m not so concerned with whether the method will work per se – that’s a concern that we all know and love – but rather any time-sucking pitfalls and how to avoid them.
If you want to include both mindset and method in a term, maybe “approach”? Sorry but “paradigm” sounds too grandiose in my opinion.
Diane: Don’t take that as an assertion that the poem was written by the author. I am simply raising the possibility.
Thing: Makes sense, but not sure how well we can hold anything static here.
bdid1dr: We call it the Voynich Manuscript out of convention. Chinese Checkers aren’t Chinese but hey, that’s what everyone calls them. Ask for “German Checkers” and you’ll get blank stares.
You left out (c) that it’s a hoax. Which would also account for why Tiltman and Friedman couldn’t read it, although I know it’s not a suggestion you like.
Brian: I’ll write up my thoughts as to how to identify matches in a future post. Lots to cover just yet, this is just the start! 🙂
Nick, as regards the ‘Poem Block’, why bother encrypting a poem? Unless the poem is more of a revolutionary message or something of the sort. In which case I would look for rally cries of the time. Additionally, would it be ridiculous to consider the Voynich MS may be a textbook?
Folio f81r contains a large stain in the empty area to the right of the text (also visible in f81v) and blotting on the 5th and 6th lines.
It looks as if the author spilled some liquid on the page, then covered the affected region with a cloth (was sticky perhaps) and continued on writing, thus producing the irregular margin.
Folio f112r also has extra space on the right margin and no apparent stain, but on the reverse side the extra margin is now on the left side.
In either case it’s unclear that the text is a poem, but it’s a possibility.
Dear Nick, you argue “that an entire line got omitted by mistake during the copying.” In my eyes there are three points speaking against such a statement:
1) The ornate line-initial gallows-initial words result in a different text structure. If “pchedy.qokeey” in line f81r.P.8 counts as markerI see no reason to omit “pchedy.qokey” in line f81r.P.26. But then the resulting text structure is 7 / 8 / 8 / 2 / 6.
2) If something doesn’t match you can argue that this is a copying error.
3) You handle the hypothesis that the VMS was copied as a fact.
As I understand you start your analysis with the block structure and with this structure you want to identify the function for some words. (In this way you named already a word “as a possible date or signature”.) With other words with the block paradigm you try to identify words from there context. Unfortunately the main problem for the VMS is that it is not similar to any other text or cipher. Therefore it is hard for me to see the difference to approaches starting with assumptions like “initial words on herbal pages should be plant names”.
P.S.: Did you already found the time to read my paper in detail?
Torsten: (1) the final gallows (that you think breaks up the final block of eight lines into 2 + 6) is smaller than the others, whereas the others are stretched gallows that extend over the letters. It’s my opinion, sure, and you’re right to flag your opinion as a viable alternative: but there you go.
(2) Out of the hundreds of pages of the Voynich Manuscript, I’m focusing here on a single page which seems to have a different structure to the others. I’m offering a slight revision on an opinion which was first proposed at least 18 years ago, and which tries to account for that structure. It’s a hypothesis grounded in a well-established discourse, with only a small step sideways.
(3) Yes, guilty as charged. But it is a hypothesis I have been debating, documenting and for which I have been providing evidence for over a decade.
You have missed the main idea about the block paradigm: which is that the point of taking the approach is not to try to identify individual words as a first step, but to identify an overall system as a first step. Given that, as you put it, the VMS “is not similar to any other text or cipher”, I think the primary challenge is to identify the broad structure of what is going on, and that is what the paradigm is all about.
I named already a small group of words “as a possible date of signature” based on the internal structure of the Voynichese. Had you asked (say) Philip Neal or Rene Zandbergen or indeed any other long-suffering Voynich expert to look at that line, I’m reasonably sure that they would have pointed out the same thing, that the last few words don’t seem to fit the text on the rest of the page. But feel free to ask them yourself.
Job: I discussed the empty space on f112r in “The Curse of the Voynich” back in 2006, and concluded that this was probably a copy of a space: that is, that the original vellum page probably had a rip that had been stitched up, and which the text had avoided on both sides. This was the kind of evidence of copying that Torsten Timm mentioned in a comment just now.
But on f81r, there is writing on the other side so this same explanation indeed wouldn’t work for it: Rene Zandbergen suggested that the gap might have been a space for a drawing that was never made. I don’t think this is likely: the way that the drawings at the top and bottom of the page go right up to the text suggests to me that the text was written first and that the drawings were made to fill the space. My suspicion is more that this text was copied from a much smaller source page, and that the drawings filled it out to a whole page.
It’s an interesting suggestion that there’s a subtle stain there: it wasn’t apparent to me when I examined it before, but I shall definitely have another look, thanks.
Dear Nick,
1) Unfortunately “pchedy.qokeey” in line f81r.P.8 also starts with a smaller gallow glyph.
2) Even if it is a hypothesis grounded in a well-established discourse I found it problematic if not fitting observations can be explained as a copying error.
The problem is still that the VMS is not similar to any other text or cipher. To identify an overall system for a group of blocks and to compare this system to other text seems to me as hard as to identify some of the plants.
In my eyes “ar or oraiin” in line f81r.P.31 fits very well to “ol ol ol aiin ol orain” in line f81r.P.5.
Torsten: (1) we obviously agree that the final line-initial gallows of the page is the smallest line-initial gallows of the page and the line-initial gallows on line 8 is the second smallest. Beyond that it’s just a matter of opinion whether the line 8 gallows is small or large. 🙂
(2) Of course it’s problematic! I’m just exploring the implications of the hypothesis that the rational basis for f81r’s logical structure is that it is a copy of a poem. There are a number of well-documented medieval balneological poems (the most famous of which I mentioned in the post), so I think it is hardly stretching the limits of rational discourse to propose that this page too might possibly be a balneological poem. But unless there’s a missing line, it doesn’t quite seem to hang together.
Torsten: and yes, I’d agree that the “ar or oraiin” in line f81r.P.31 does seem to match “ol ol ol aiin ol orain” in line f81r.P.5. Perhaps this will be prove to be an internal feature of the original poem… we shall (hopefully) see. 🙂
That pair: ‘ar or oraiin’ …./…. “ol ol ol aiin ol orain”
sounds so like the thumping rhythm of proverbs, or the pattern of saws and maxims, doesn’t it?
I didn’t want to publish this until I had done some more research, but Job hit upon my reason for believing that f81r might not contain a poem.
https://medium.com/@thingsnorthern/the-riddle-of-folio-81-c5ccdb749649
I must stress that the ideas in the article are a suggestion, and from somebody who doesn’t know enough about parchment. I hope people find it interesting, but I really am open to learning that I’m wrong with this.
Examples of likely ‘ver-i-si-mi-li-tude:
Prayer: Panis angelicus
Poem/Biblical reference: ‘Behold the lilies of the field…….
Baptism: Full immersion in water? Or oil/ointment?
Herbal remedies for everything from bowel movement (psyllium seed husk) to bathing soap and shampoo (yucca root).
What I would really like to see in manuscript B-408 is some mention of the messenger service known as “quipu”: A string ‘belt’ upon which many knotted strings had been tied.
We might not be able to find such an item in B-408 because the messengers may have died post-Columbian occupation of South America.
I still think B-408 was Fr. Sahagun’s memoir/diary which was put aside when it came to having some thirty years of research and work with the native populations published. One has to wonder, now: When, and by what means, did the manuscript end up in Europe. Before or after the fall of Granada? Before or after Busbecq brought it (and some 200 other manuscripts) to Vienna? Did it reach Rudolph before or after the battle of White Mountain and the beginning of the “Thirty-Years” War?
Then there are references to “Monte Alban” and “Alban Hills” — and Lakes Alban and Nemi. Lake Nemi, in particular was associated with the goddess/ Artemis/Diana.
There is a mountain in South America called Monte Alban, which apparently has been an archaelogical dig for several years now. I’m not sure if the current “Wari” dig has any relation to the Monte Alban region in South America.
Nick, do you have any archaeologist correspondents (who might be able to back up some of Diane’s contributions to your discussion pages?
🙂
Oh dear me, I’ve just returned from a ‘short’ break. I got sidetracked with a recent publication which you may find interesting:
The Lost Fleet – The Discovery of a Sunken Armada From the Golden Age of Piracy. The writer of this book, Barry Clifford, has written an earlier book: Expedition Whydah.
I’ posting this information on these pages because I’m hoping I haven’t created a ‘road block’ to some of your ‘regulars’. May apologies, y’all!
;-(
Thing,
It doesn’t take a professional to see that the parchment (according to McCrone and the University of Arizona) is not perfectly equalised. One can fairly easily distinguish sides or areas on sides that show hair follicles and feel (to quote Dana Scott) ‘like peach fuzz’. I suppose the point here is that German manuscripts had perfectly equalised parchment from about the thirteenth century, so these indications run counter to the ‘German hypothesis’ which was espoused by some Voynich people, though it is worth noting that few of the early expert appraisers thought so.
It seems to me that the best approach to making any sort of progress, is to study and collate all manuscripts contemporary to and prior to the Voynich MS that have similar imagery and probable content. After all, it is extremely unlikely that any document of this type would have been produced without precedent,
I’m sure this approach must have been looked at before now, but possibly a more organised and unified effort may produce some useful results.
Nick & Friends:
I call your attention to B-408, folio 42-r (which specimens I identified as radicchio (the large leaf which should have been colored RED) and the smaller specimen as cilantro/coriander (which leaves should have been colored green). The accompanying discussion of the two specimens also seems to have gotten mixed up . There is also mention of parsley.
I see a lot of potential for rhyming verses — if only the artist had been able to sort out which discussion went with which illustration!
Still smiling……..
‘Thing’: Meticulous — a good description of your efforts! I’m hoping you are following the discussion I just posted herein.
In regard to the balnealogical pages: Many of the ladies make their way through a series of baths (hot? warm? cold?, seawater? Freshwater?…….and eventually end up holding a mandrake fruit (B-408, f-83-v). The labels and discussion is about the care and good health of young women of childbearing age. In one of the various bath house folios, one can see a man, wearing a loincloth, waving at the occupants (women). I haven’t found any discussion regarding that male’s presence in the ladies’ bath-house.
😉
Diane, if it is true that you can easily feel the difference between the two sides of the Voynich manuscript parchment, then I accept that I am wrong. I was simply following statements from Rene and Nick that distinguishing the two sides was extremely difficult.
Why does discussion regarding the pictorial elements of B-408 almost inevitably get diverted to the date of manufacture, the type of animal skin used, ink, paint, dye….. ad infinitum?
I can understand some aspects of the women’s bath-house because some of them are holding emblems for their familial origins (Armenian and French being two). Other aspects of the folios in the numerical 70’s ( r and v) and 80’s (r & v) are discussing the care and nurture of young women, and issues dealing with conception, childbirth, maternity, and non-conception/abortifacients (mandragore fruit, juice, and root).
Some discussions relating to various floral specimens Dianthus andTurban Ranunculus in particular, are discussing the enslavement of women, with reference to the rising star ‘of the east’.
The so-called ‘sunflower’/’pincushion plant’ is Scabiosa caucasica. The full-page discussion which accompanies that illustration is all about using the plant with warm water to remedy mange for animals, and scabies/mites for humans.
I’m hoping these latest discussions aren’t ending up in Nick’s ‘back-pages’. I’m still squinting, Nick! Onward!
😉
The real paradigm shift is to stop imposing external solutions and to start following the internal evidence provided by the VMs.
Between known language and unknown text, there needs to be a bridge. That bridge begins with an unexpected language and that language is heraldry. However, when the text is intentionally disguised, the key to that disguise is also hidden. Nevertheless the trail of evidence can be followed in the VMs. Color, pattern, and pairing – once the optical illusion of radial orientation is removed. And any other difficulties are resolved by the addition of the red galero and the recognition of the unique historical situation represented in the White Aries illustration. And this interpretation is further confirmed by the proper hierarchical positioning of the relevant figures, their favored heraldic placement and the other independent elements I have listed previously, culminating with the use of obscure heraldic images and terminology to make a singularly definitive pun. And the really funny thing about the pun is that it is completely obvious to those in the know, and completely obscure to those who are not. Heraldry serves fairly well as a method of communication, particularly for those who were also literate. And in this particular case, for those who were also familiar with the early records of heraldry in the church.
Looking into the VMs illustration of White Aries, the language of heraldry identifies two historical persons somewhat obfuscated. Disguised at first, but never wholly denied, and then repeatedly confirmed by internal evidence and historical fact. The purpose of the papal identification is to transfer emphasis and validity to the papal keys, which are also known as Stolfi’s “start here” markers. And this transfer of significance is based on the internal evidence of the unique conjoined patterns found on the White Aries page. There are three such textual markers in the VMs Zodiac and they produce three segments of text. Given the complexity of this construction, there should be no question of its intentional significance.
Though this clearly is not a solution to the VMs, it is my contention that the path that leads to that solution was set to begin with the Genoese Gambit. Whether that path can still be followed remains to be determined. The internal complexity found so far would imply that the pathway is marked, but the evidence of papelonny would indicate that this is only for those who can properly interpret the markers. The signs are there for those who can read them. An unexpected language is in play. Author’s choice!
Given the broad spectrum of topics that have been tied to the VMs, I again would like to request that heraldry be given a separate discussion category. Certain topics, far more unusual than heraldry, have been given their place of consideration and in light of the sophisticated use of such an obscure term as papelonny, on top of everything else, there is a validated basis for such inclusion.
Dear Nick, on one side you argue that the Voynich manuscript was copied from an external source and that it contains copying errors like missing lines. On the other side you argue that the form of a single gallow glyph is enough to determine the starting point of a new paragraph.
There is no doubt that a new paragraph starts in line f81r.P.16. This is sure because of the gap between line 15 and 16 and because of multiple instances for “p” in line 16.
After Takahashi another paragraph begins with line f81r.P.24.
The reason for him was probably that “p” occurs twice for line P.24.
Because of the usage of “p” further candidates for new paragraphs are the lines P.8, P.20 and P.26.
My view is that the Voynich manuscript is an original and that it contains far less errors then you assume. But this is another discussion.
Torsten: no, I argue that if the long-standing hypothesis that the f81r text block is a poem block holds true, then I think the direct implication is that the source of that poem probably had an 8 / 8 / 8 / 8 line structure. Hence my “missing line” copying error prediction arises as a consequence of the f81r poem hypothesis, not as evidence. The whole thing is not an absolute claim about Voynichese that holds true for the entire manuscript, but is instead a localized hypothesis about the apparent internal structure of a single page, and the putative internal structure of its original source. Perhaps I’m right, perhaps I’m not: but if the source for that page does exist and can be found, then everything changes.
I’m not saying you and Takahashi are wrong for seeing what you see at the character and line level: I’m only exploring the macro-level (or, rather, block-level) implications of structure flowing from one document to another, and perhaps there will prove to be a deeper logic that explains both aspects at once.
Out*of*the*Blue: I started talking about cryptoheraldry and the Voynich Manuscript nearly a decade ago, and I fully expect that aspects of it will still be tapping at my windows in another decade. Right now, though, I’m trying to take a new (and extremely joined-up) approach to the Voynich Manuscript, one where all the individual pieces have to join together, an overall project that cryptoheraldry hasn’t yet managed to contribute to (in my opinion). But I will keep trying, and I hope that you will too. 🙂
I’m not sure what your cryptoheraldry theories proposed beyond hypothetical existence. It is interesting how close Stolfi came, but still didn’t see it. I had no luck finding any of your heraldic discoveries that carried over into this blog’s search capabilities. Where did they go?
I have laid out the various independent elements of evidence as taken from the VMs f71r illustration, which presents a recognizable representation of a specific historical situation involving two identifiable historical individuals. [The Fieschi popes] There were a good handful of independent elements that help to confirm this identification that were already presented some time ago. The proof that all this is part of an intentional construction is capped off by the recognition of the level of internal sophistication and intentional construction needed to set up the papelonny pun. If you can look at this construction in the first three pages of the Zodiac and say that you fully believe that the existence of the elements as I have cited them is really and truly random and irrelevant, then you are not interpreting the elements within the text in the way I have proposed.
Papelonny is in the VMs, it is in heraldry and it is in history. It is not well known. But that has no effect on the validity of the definition. Either the intention of the construction is acknowledged or the illustration presented is completely devoid of any intentional representation. And clearly there is a way in which the elements I have cited are most decidedly not devoid of all representation. The elements of the illustration are based on specific historical events. And if presented in a totally straight-forward manner, they would probably be patently recognizable to a fair portion of those in certain parts of Europe who were literate at the time of the parchment’s creation, let us say. Another option was needed. And nothing beats a little old optical illusion to completely befuddle the situation.
Do you think it will be another ten years before Rip Van Winkle wakes up?
Dear Torsten,
The manuscript as we have it has to have been formed from earlier bound volumes, because the signs of earlier and disjunct bindings are clearly evident (see the old ORF documentary, or if it’s easier, Nick’s post about same).
Secondly, the imagery repeatedly shows a clear disjunction between Latin European cultural, religious and iconographic parameters. Europeans simply didn’t know/do/believe what we find represented in the manuscript, and the use of a square model for the world is merely one very obvious example. (btw, some Arabian depictions also conceive of the world as square. Scholars are still uncertain of whether that reflects a common culture across the Great Sea or direct east-Asian influence).
Plus, of course, those two red-coloured characters on folio 1r. I never did get around to asking someone from East Asian studies to take a look at the one on the upper right which I suspect had been added, originally, with the vermillion brush. It may be a token of permission to travel, reproduced pretty well with quill and ink in the western style.
There’s really no chance that our present manuscript is the only copy of a single work having a single fifteenth century Latin author. Sorry.
B-408 is a section of a rough draft for an eventually published document. It began as a diary, which accompanied the traveler to New Spain. When it came time for the manuscript’s contents to be reproduced on amatl/paper (mulberry or fig bark) B-408 got separated from the paper-bound volumes.The “Florentine’ manuscript continues with the dialogues and illustrations; all of which have been translated from Nahuatl to Latin.
So, with three dictionaries (Nahuatl, Latin, and English) I am able to translate much of what appears in the so-called Voynich manuscript — illustrations or not. So, this morning I intend to do a full translation of B-408, folio 81r. I’ll get back to y’all when I’m done.
😉
Nick,
If ever you have a moment, could you offer some definition of what cryptoheraldry is? I have this idea of heraldry being used symbolically, to represent parts of a language – perhaps like della Potra’s symbols, using line after line of heraldic shields or something like that.
That has nothing to do with what I am proposing. Heraldry is just used as heraldry was intended to be used. Things are not always perfectly clear, but the traditional rules of heraldry are the standards which I try to use, without additional, hypothetical impositions. So I don’t quite see where the crypto- part applies to the Genoese Gambit. The failure of recognition that persists in our modern investigations results from the complete absence of all acquaintance and familiarity relevant to the specific topic of this one particular heraldic identification. Ask anyone for the blazon of the Genoese popes and let me know how that goes.
The radial interpretation is not the only way to see the White Aries illustration. Turn off the radial cloaking device and see that the orientation of the blue striped patterns now comes closer to being a match. With the red galero, the historical connection is validated by the correct hierarchical placement of the figures in the celestial spheres. There is something to be said in favor of having figures that are historically grounded. It adds a sense of validity. It then becomes clear that each independent element of confirmation is intentional. The use of papelonny in the VMs is intentional in its construction. If (and only if) you know what it is, then it does what the author intended it to do. It labels the popes in the illustration.
I don’t think that historical obscurity actually counts as form of cryptographic methodology. It sure as heck works, but it probably doesn’t count.
.
Bdid1dr – would I be right in thinking that you draw a parallel between the Voynich ms’ bathy- section and images in Codex Xolotl?
Dear Diane,
you misunderstood my statement. Nick is arguing that the manuscript is a copy of another manuscript.
Contrary to Nick I say that the manuscript is not a copy. Thats all.
Torsten: errrm… I thought you claim that the Voynichese text is a copy of itself? Or rather, that most of it is nothing more than a copy of text elsewhere on the same page?
Dear Nick,
indeed, my claim is that the Voynichese text is a copy of itself. But all I want to say here is that the text of the Voynich
manuscript was generated during writing and that the manuscript is not a copy of another manuscript.
Nick, Diane, Ootb,(and Torsten): I am not able to download page after page of B-408’s offerings. Each page/(folio) requires my hand-writing each letter/word form onto a piece of lined paper. It takes three lines of my hand writing to make sense of the first line:
First line is the text appearing in any folio.
Second line is translating into Latin terminology (botanical, balnealogical, religious, medicinal remedy or preventive medicine (scabiosa caucasis-scabies, lice) or maternal health & well-being (including the use of the FRUIT juice of the mandragore). Elsewhere in B-408 can be found use of the mandragore Root burning/smoke for anesthesia and surgery on battlefields.
I can (but won’t) repeat this particulary dialogue, Nick; you know where to find it.
I will try one more time to explain the non-code ‘alphabet’:
a is a
b looks like a P
c can be either c or e depending on its relative size and whether it has a bar or tail extending to another ‘c’– which are often found connected — and sometimes have a curlicue above the connecting bar.
d (and t) do not stand alone, but rather appear as tl
e looks exactly like a smaller c but has a small bar extending from its back
f (and V) appear most often as oe
g, j, or k appear as a large 9 —–x appears as a small 9 AND either character 9 can be attached to the ‘eus’ or ‘eas’ words
LL and TL are fairly well recognizable by their “loopy legs”
M and N are recognized by the number of half-bars attached to what look like parentheses.
o is obvious
P is the most ponderous and ponderable SYLLABLE in the so-called Voynich manuscript. It can indicate the beginning word of any dialogue, but requires some knowledge beforehand of what the dialogue will be. This is the best I can explain the elaborate character y’all call “gallows”.
Like the so-called “Brackets” character (t-l) or (l-l), it can be stretched over a whole section of dialogue.
Q is represented by the small, straight-leg ‘q’ as you see it here. No ‘u’ is necessary to the context. You can see the ‘q’ most often in the aquatic botanical discussions/displays.
R is the backward-facing S
S is the large figure 8 for the sound, most often, ‘aes’
T does not stand alone — see my comment for ‘dl’
U is eo
V is oe
W is eoe
X is explained above
Y is oi
Z is z…….?
One last character (please DO NOT focus on Tyronian notation) but do realize that it often finishes some dialogues:
itius, adieu, tius, dios, deus………
Adieu!
🙂
Torsten Timm,
What you write is against all evidence. The Voynich MS was not a new book, but a compilation of libellae. The individual libellae have been copied from older sources in the beginning of the 15th c. and have been rebound mid 16th c., when te meaning of the script and contents had been forgotten.
People: I refer you all to a soft-bound 8″ by 11″ book which portrays one section of Fr. Sahagun’s magnificent “Florentine Codex”. This one section (title, The Gods) has several pages of illustrations which portray the various god/godesses of “New Spain”. It would monopolize Nick’s pages if I continue discussion of this one book of twelve. I am now able to translate/transcribe many of the Nahuatl-Spanish/Latin scribal commentary for the 12-still-extant portrayals of Fr. Sahagun’s contacts with the Native South Americans.
So, I may be taking a lengthy pause from Nick’s gr-r-r-eat
blog! Have a Merry Whatever Season. I may be back online come New Year’s Eve!
Bye!
bd I’d 1-dr 😉
Dear Menno,
the question is if the text of the manuscript was copied from
another source and contains numerous copy errors or if the text was generated during writing. I didn’t see how the point in time when the manuscript was bound or rebound says anything concerning this question.
The only evidence for the hypothesis that the text was copied from another source is Nicks discussion of empty space on page f112r and f112f. Unfortunately the parchment is deformed at this place. Therefore this place could also be used as evidence for the hypothesis that the text was generated during writing.
Anyway, in my eyes it is problematic to assume that the text contains an unknown number of copy errors. For this reason I prefer to analyze the manuscript as it is.
Torsten: the empty space on the reverse contains paragraph stars shifted across, so the page would seem to be perfectly workable as a support medium.
Also: it is the normal case for historical ciphertexts to contain (often numerous) enciphering errors, so I think it would be unrealistic to assume there is no possibility of errors here.
Torsten,
You say:
“the question is if the text of the manuscript was copied from
another source and contains numerous copy errors or if the text was generated during writing.”
Since the pictorial text was certainly not first enunciated in the fifteenth century, to argue that the written text was must surely imply either than the text bears no relation to the narrative offered by the imagery, or that the text is subsidiary to it – as commentary or something of that sort.
In my opinion, and despite the long period – centuries – between first enunciation of the imagery and our fifteenth century copy of it, the copyist(s) were so careful that certain details survived intact: as example, the form given the scales in the calendar section which clearly shows the basket/tray suspended by cords which pass through the crossbeam. No medieval European balance that I’ve seen used that ancient form.
Since our fifteenth century work shows no apparent difference (as far as I’m aware) between the ink used for the imagery’s line, and that used for the text, then either (as you say) the text is newly created for the fifteenth century manuscript, or there’s a paradoxical care over imagery yet that care is being supposed abandoned when it comes to the text.
If there is a chance it is poetry then it is a great target for a block copied verbatim(possibly with errors). Is your current profile of the passage leading you to look at only the line length, topic, and time/region for comparison targets?
Character and word frequencies are hard to match if i’m not mistaken, that said are you thinking that by moving to the block level we have a better chance? In particular, do you think starting with a block that looks like it retains its structure will let us zero in on the source? In simplest terms, is the current objective merely vetting possible source material for structural matches, or do you already have a more refined methodology?
T Anderson: though I’ll try to lay out a longer answer before long, the short answer is that I’m hoping to find one or more internal structural matches and work back from there. That is, if we can find pre-1450-origin balneological poems with an 8 / 8 / 8 / 8 line structure, we can try to work with those to see if we can identify good candidates whose internal structure matches what we see.
The slightly longer answer is that whether it’s ultimately written in an unknown language, a cipher, or a shorthand, ‘clumpy’ (hard-to-encode) things like dates or numbers or unusual names (or even repeated names) should be structurally visible within the layout, even if at first we have no idea how the underlying mapping works. Find the block, and work out the details later. 🙂
Dear Nick,
both unusual observation can be made for the same page. The starting hypothesis is therefore that they belong to each other. On top of sheet f112 is a wrinkle. The best way to avoid that wrinkle during writing the first two lines on page f112v was to start where line f112v.P.1 and f112v.P.2 do start. As far as I see there is nothing unexpected on this place.
The question was if the text contains an uncountable number of copy errors. There is no doubt that the manuscript contain unusual words (see my paper p. 5.) . Maybe they are errors.
If the manuscript contains a ciphertext or not is another question. But if it is an ciphertext it would be very hard to copy such a text. It is a fact that all published transcriptions of the VMS frequently differ. In my eyes it is therefore unreasonable to assume a copied ciphertext.
Torsten: enciphering is already a form of copying, and I have long suggested that the original documents were enciphered onto a wax tablet (by the encipherer) and copied from there onto the page (by a scribe). All of which I suspect would comfortably account for the presence of various different types of copying error.
Picture: A man in a long robe, kneeling on the ground; a pile of dirt near his knees. The Nahuatl/Latin verse which accompanies this picture translates to ‘first dig a hole’ (a cavitl).
It was so long a go when I found this small black & white reproduction of a manuscript page (and posted to Nick’s blog) I’ve forgotten the manuscript number/name. I’m pretty sure, however, that it was that tiny photo and text which sent me exploring “New Spain” texts and manuscripts written by European-born monks. Sahagun’s works appear in several different forms because of his teaching while preaching and practicing medicine. Amazing what an octogenarian could produce before his death. What is particularly sad is that for a while he was under the scrutiny of the Inquisition for a while, himself. (Because of brother monks envy of the recognition he was receiving from “Headquarters/Inquisitioners”.
Dear Nick,
it makes sense to use some notes and code tablets for encoding a text. Also the presence of various types of encoding/copying errors would be explainable. Are there indications for such errors? One indication for errors are corrections. Unfortunately one remarkable observation for the manuscript is that corrections are missing [see Reddy: p. 79].
Furthermore I would expect that sometimes the space is running out at the end of a line. But the lines always fit perfectly into the available space [see also Currier]. Even the limited line length at the start of page f112v doesn’t mean that the scribe was running out of space there. The same is true for the pages f114v and f115v. During writing this pages the starting point was shifting to the right. Also the hole in page f107 did not result in any space problems. (For another interesting example see my paper p. 33-34). One possible explanation for such a behaviour is that it was possible for the scribe to select glyph groups which fitted into the available space. In my eyes this indicates that the scribe was generating the text during writing.
Torsten: I’ve previously emailed you various statistics (e.g. 4o/4a and aiiv/oiiv ratios, etc) that suggest to me the possibility of pervasive copying errors. I’m pretty sure I’ve posted about this too, but that was certainly a fair while back. 🙂
One of my long-held suspicions about the text and drawings is that they seem to have been arranged so as to duplicate the overall layout of an original source text. It seems to me that this is likely to have been achieved (in terms of process) by copying a line of text onto a wax tablet, and then enciphering it in such a way that the enciphered version takes up the same physical space. In that way, the overall layout is preserved, and the original layout of the document could potentially be reconstituted from the enciphered version.
In the case of f112v, it makes no sense to me that the text is right-justified in the way it is, except if the original document was also right-justified (working around a stitched-up rip or tear in the vellum on the outside edge). The paragraph stars, then, would seem to have been an addition, because they drift over towards the left, untethered to the text.
I accept your point about f114v and f115v: but would additionally point out that the same behaviour (i.e. that there are more words on the top line than on the bottom line, and that this seems to be a progressive change rather than an abrupt one) seems to be true of many other pages in Q20, and that this phenomenon isn’t something that seems to have discussed before.
Glen Claston, Elmar Vogt and I all debated the strong possibility that what we now call Q20 may well have been made originally composed of two smaller quires (I used the terminology ‘Q20a’ and ‘Q20b’ to differentiate them, as I recall), and it may be that one of these had – in its original folio shape – pages that were slightly wider at the top than at the bottom. f105r seemed (from its fancy gallows at the start) likely to have been the front page of Q20a, while f116v was the back page of Q20b: you get the general idea. Something to think about, anyway. 🙂
Nick, Torsten & All,
Quite often (throughout the discussions on any folio in B-408) one will find many ‘wraps’; the most common wraps being ‘aes ceus’.
As far as the large so-called ‘gallows’ figure is translated:
Depending on the number of loops and curliques which always extend over one or more syllables of text before looping back to, and sometimes behind, the upright stem, a person can read/interpret words such as ‘before’, brigantine, brilliant, beneficial, preference, previous……protest…..pri….(new) paragraph…….
Perhaps a person primarily perusing my presentation (herein) can paraphrase my latest contribution in “Voynichese”?
By the way, Folio 116v was Ambassador Busbecq’s sign-off reference to Ancyranum Augustus: his point of departure from Suleiman’s empire. He brought some 200 manuscripts given to him (some in very bad condition) — which he delivered to the Austrian Emperor in Vienna. Some of the manuscripts were sent to Rudolph II in Bohemia.
So, as long as you understand that the manuscript was most likely lost for a while, during the Inquisition’s investigations in “New Spain”.
I’m still reading Friar Sahagun’s magnificent body of works, called the Florentine Codex: Book one, “The Gods”.
Dear Nick,
indeed “qa” (“4a”) and “oiin” (“oiiv”) are uncommon. But are some statistics enough to classify them as errors? To answer this question it is necessary to look into the details.
The transcription of Takahashi contains six words using “qa”. This words are “qaiin” (2 x), “qairal”, “oqaiin”, “qakar” and “qaloin”.
The word “qakar” in line f33r.P.5 is a possible transcription error. The transcription of Currier and the first study group list this word as “qokar”.
“qaloin” in line f76r.R.19 contains two uncommon elements. This elements are “qa” and “oi”. This group looks indeed weird. Instead of “qaloin” we would expect a word “qolain” here.
Surprisingly a word “qolain” can be found within the next line. The word “qolain” exists only three times for the whole manuscript. Therefore it is unlikely that it is just coincidence that both word occur in subsequent lines. With other words “qaloin” belongs into the context of that page.
Even if you change “qa” into “qo” for “qaiin”, “qairal” and “oqaiin” this would result in an uncommon “oi” element. With other words it doesn’t matter if this words contain “qa” or “qo”. Because by the statics you would classify them as errors anyway. But then it is not possible to explain this type of words as a result of some copy errors.
How is it possible to explain uncommon words like “qaiin”? If we split this groups into common elements like “q”, “qo”, “al”, “aiin”, “iin” etc. “qaiin” would result in “q” + “aiin”, “qairal” would result in “q” + “air” + “al” and “oqaiin” would result in “o” + “q” + “aiin”.
With the concept that Voyniches can be split into common elements it is possible to explain even uncommon words like “qaiin” (“q” + “aiin”) or “qoiin” (“qo” + “iin”). Since it is possible to find a concept to explain uncommon words I would not count them as copying errors.
On one side you argue with pervasive copying errors. On the other side your argue that the layout is a duplicate of an original source.
The scribe didn’t use any lines as guides. This fact alone is enough to explain the shifting for some of the pages or direction changes like that in line f112v.P.30.
For me it makes sense that the scribe was considering damaged parts of the parchment like that on page f112v. See for instance how he used some drawings to solve the hole in f82r and f82v.
Torsten: qakar looks like a mistranscription, yes. But qaloin followed by qolain on the next line looks to me like a copying error, where the a and o got swapped over. And I’m guessing that the qaiin and oqaiin are copying errors where EVA l got miscopied as EVA q. So I don’t need anything complex to account for these oddities.
I’m not talking about the margins shifting or drifting, I’m talking about the observation that the lines seem to get shorter the further down the page you get on lots of (but by no means all) pages of Q20.
Speaking of Landini, Nick, could you or any other researcher please clarify for me whether Landini created the EVA true-type font? Many online sources credit him; some English language sources credit him first, and then add Rene Zandbergen’s name. The French wiki article credits both, with Rene’s name first.
I expect the font developed over time, which may explain the earlier references only mentioning one person.
Diane: I’m reasonably sure that Gabriel Landini created it on his own, but I could be wrong.
Dear Nick,
uncommon words are common for the Voynich manuscript. There are only two words contain EVA “qp”, only two with EVA “qs” and only two using EVA “qt”. Only four words use EVA “qa”, only four contain “ch”, only five contain “qy”, only seven words contain “qk” and only 22 words start with “oq”. Are this all errors in your eyes? If the answer is yes, this would change the text in a significant way. You would also reduce the already low entropy for the text. If the answer is no, what is your criterion to distinguish between rare letter combinations and errors?
Beside “qaiin” (2 x) the text also contains “qoiiin” (4 x), “qoiin” (3 x) and qoin” (1 x). To change “q” into “l” for this words would result in “loiiin” (2 x), “loiin” (4 x) and “loin” (0 x). How do you explain this group of words?
The shifting or drifting of the margin did result in shorter or longer lines (see page f112v). Anyway, even the entries in your blog here demonstrate that it is easier to let some free space at the end of a line then at the beginning. Therefore it is remarkable that for the Voynich manuscript the end of the lines always fit into the available space (with the exception of page f81r). As I understand you argue that this happens since the text was copied from an external source and that the length of a line was known from the source text. If this was the case copy errors like a forgotten word or letter would be easily detectable by the free space at the end of a line. Therefore the resulting layout can only mean that the scribe was able to avoid such mistakes during copying the manuscript or that the manuscript was not copied word by word from an external source. Therefore it is an contradiction in my eyes to argue that pervasive copying errors did not result in any observable layout change.
Torsten: I don’t know if they’re all errors, but even if they are, your assertion that “this would change the text in a significant way” would _not_ be held up by the numbers. These are a small number of words out of a very large corpus, which would not significantly affect the overall entropy.
I didn’t suggest changing ‘q’ into ‘l’ in those qoi[i][i]n cases: they may well have a different explanation.
I’m not going to agree with your “can only mean” logic (and the resultant contradiction you detect): there are plenty of mechanisms that could explain all these things, and – moreover – there are a huge number of combinations of mechanisms that can lead to the same effect.
1. If you look at the sequences daiiin – daiin – dain and daiiim – daiim – daim – dam in the herbal section of the VMS you may conclude that we deal with the same ‘word’ dam. Similarly the sequences doiiin – doiin – doin and doiiim – doiim – doim can be treated as dam. The matter is, that the VMS reflects consecutive spellings, sometimes mixed in the same ‘sentences’.
2. It has been argued, that the VMS originally comes from Roger Bacon (1214-1294) or one of his followers. I give it the benefit of the doubt as this would fit in the time table and seems to follow his scientific approach.
Greetings, Menno
Torsten: omitting words or letters would alter line length, but spelling errors wouldn’t.
Nick: entropy doesn’t matter – that’s to do with statistical tests which are ‘essentially worthless when applied to the Voynich Manuscript. They have not worked, do not work and never will work for it.’ So that’s all right then.
Torsten: I do not use the EVA for any of my translations. The lower case alpha-characters (iiiiiii iiiiv qokeedy ad nauseum) make no sense because they are a twentieth-century attempt to decode a non-existent code. Forgive me, Nick, if I am stepping on professorial (and your) toes.
I try not to condescend or offend anyone on Nick’s blog. I am retired from my work as a records management specialist. The instinct to translate anything I read began when I was four years old. Instinct is still present — and probably will endure until I am extinct!
So, y’all have whatever celebration you observe for the end of year 2014. I look forward to some new developments in year 2015.
N’as drovnya!
🙂
Word classification is another potential form of block-level analysis.
While we can’t determine the meaning of Voynichese words, we might be able to classify them as probably-noun, probably-verb, etc. This process would be repeated to produce further classifications (with decreasing confidence).
It’s an approach that can only yield probabilistic data, but it has some value.
Job: given that the massed armies of linguists who have attacked the Voynich can’t even begin to agree what are vowels and consonants, I’d be fairly surprised if they can do any better with words. 😉
For what it’s worth, the “obviously” vowel-shaped letters (EVA a / o / i) don’t get classified at all as vowels by automatic vowel detection algorithms: which kind of drives an awkward wedge between would-be linguists (i.e. those just trying to read Voynichese as an unusual language) and computational linguists (i.e. those who have the greatest respect for Boris V. Sukhotin etc).
Nick, the same is true for any type of block-level analysis, it’s unlikely that you’ll get a consensus.
The type of word classification i was suggesting would operate against a probabilistic designation of words in a language by length, recurrence, position in a paragraph, and position relative to other classified words.
The result would be an equally probabilistic classification of Voynichese words, with the stated initial assumption that the grammatical structure of the plaintext is preserved by the cipher or encoding process used by the author.
I recently read a paper by Tiltman where he describes a dictionary cipher from the 1600s. This type of cipher is well aligned with some properties of the text – moreso than any other, in my opinion – and would preserve the underlying grammatical structure.
In fact, i thought you’d be open to this idea because it gives linguists something worthwhile to work on. 🙂
It’s also well within the parameters of your block paradigm.
Job: that presupposes that Voynichese ‘words’ equate consistently to specific words in a plaintext language, so ‘daiin’ always stands for (say) ‘leaf’.
Nick: vowel detection algorithms are fine as far as they go, but will struggle with some systems of encipherment. Imagine a monoalphabetic substitution ciphertext where each enciphered word is interspersed by one or more ‘words’ made up of randomly assembled letter groups. Skews letter frequency for the document, and also disguises the behaviour of individual letters.
But as statistical analysis of the Voynich is futile, apologies for a futile and irrelevant post…
“For what it’s worth, the “obviously” vowel-shaped letters (EVA a / o / i) don’t get classified at all as vowels by automatic vowel detection algorithms: which kind of drives an awkward wedge between would-be linguists (i.e. those just trying to read Voynichese as an unusual language) and computational linguists (i.e. those who have the greatest respect for Boris V. Sukhotin etc).”
I seem to recall reading (maybe on Zandbergen’s site) that some of these had been found to weakly correspond to vowels.
Thing: nope, I think your information may be wrong there, sorry. 🙁
SirHubert: statistical analysis forms a hugely important plank of supporting analytical activity – it’s expecting it to hand you an answer on a plate that’s the futile part. 🙂
From the ‘words’ starting with one of the special signs (K, P, T, F – cKh, cPh, cTh, cFh) along with similar formations starting with o-, qo- we may decide, that the initial o is not merely a vowel, but has a special function, e.g. as word or group separator.
Menno: when I looked at this specific issue back in 2006, it seemed to me that (qo) (gallows-) (-word) was far more closely allied with (gallows-) (-word) than with (o) (gallows-) (-word). From that I concluded that (qo) was probably a free-standing genuinely short word (I later guessed it was steganographically hiding ‘lo’, the Tuscan word for ‘the’).
Even so, if the language driving Voynichese does have a short word that’s repeated (nearly) everywhere, my prediction is that it’s likely to be ‘qo’.
Look up the word ‘quoin’: Usually applies to cornerstones of masonry etc; but the possibility of using the word as “cornerstone of our religion, philosophy, beliefs……….
BD: thank you, I don’t need to look up “quoin”, and my house has several of them. But there is a Latin hymn Angularis fundamentum which has been translated as “Christ is our cornerstone.” Not sure of the relevance of quoins to the Voynich Manuscript but I mention it for possible interest.
f-81r: first line of text: bal-ne-ox-ece-geus aq-uo-tl-e-aes-geus
The above are only the first two words: the very first elaborate “P” (or what you call gallows) is saying ‘balne’ –and extends over the ‘ec’ — the last two characters of that one word are saying “e-geus”.
I refer you to another rather interesting website :
OlmosArtep32r.jpg
The photo image is of a manuscript written in mid-1500 ad. It portrays examples of words in singular or plural mode usage/writing. This one page also has discussion (of which I am just now beginning to read (en Espanol).
Fourth word of second set of examples of singular or plural:
ue-xol-trl — gallo –ueuexolo– roosters !!!!!!
See for your self, if you doubt my sanity!
beady-eyed wonder-er 🙂
SirH: You’ve been writing at the same pace; so we cross-post occasionally. That word quoin was used by Father Sahagun in his response to Inquisitional accusations of heresy. The office of Inquisition had received an ‘anonymous’ letter from another missionary stationed in “New Spain”.
I try to be brief. Concise?
bd
Nick, on Zandbergen’s site “Analysis Section (2/5) – Character Statistics”, he says:
“An algorithm for detection of vowels and consonants was designed by Sukhotin, and Jacques Guy has experimented with this in the 1990’s. Results indicated that the characters that look like vowels (a, o, y) also appeared statistically like vowels, though the confidence of the result was not that high.”
I haven’t seen the results, and I appreciated he says the cnofidence was low, but it stands at odds with your statement that they don’t look like vowels “at all”.
Jacques Guy gives a very lucid summary of Sukhotin’s algorithm here:
hum.uchicago dot edu/jagoldsm/Papers/GoldsmithXanthos/guySukhotin dot pdf
It’s based on the principle that vowels tend to come next to consonants, not other vowels, which is reasonable for many IE languages. Hardly surprising that ‘daiiin’ might be a challenge, then!
Given that EVA was, I think, designed to be fairly pronounceable, I’d expect the algorithm to have at least some succes in finding which letters are transcribed as vowels, so this still doesn’t tell us much.
A brute-force check of all possible vowel assignments indicates that EVA { a, e, h, o } are definitely vowels.
At least that’s the result when we reject vowel assignments that yield 10 or more words of 5 or more characters that contain only consonants.
Of course, this is a naive vowel classification that ignores character positions, but most vowel assignments that pass this simple test include EVA { a, e, h, o }.
The fifth vowel is unclear, though EVA y is the fifth most frequent.
In fact the only vowel assignment that doesn’t produce at least 10 all-consonant, four-or-more-character words is [a, e, h, o, y].
Dear Nick,
indeed, beside of EVA “oi” (335 times) we are talking about rarely used bigrams. For 97% of the words “q” is followed by “o”. The letter before an “i”-group is in 94% an “a” and in 5% an “o”. But if you use this statistics to argue that “q” must be followed by “o” and the letter before “i” must be an “a” this reduces the number of possible combinations dramatically. If the occurrence of “q” and “i” is in this way predictable what information did “q” and “i” provide? Or in other words the less likely an event is, the more information it provides when it occurs.
Other mechanisms are only thinkable if the scribe was able to define the end of a line himself. But in such a case the scribe made much more then just copying a text he was unable to read.
Anyway, since there is no progress anymore I say thank you for this discussion. I wish you and the readers of your blog a merry christmas and a happy new year.
Job: EVA is a stroke-based transcription, where ‘h’ is half a character (either ‘ch’ or ‘ch’), often with a gallows interposed between the two halves, and only extremely rarely occurring as a standalone character. How can half a character be a vowel?
Torsten: your presumption is that because these rare characters are entropy-rich they must be meaningful (and therefore it would be a bad idea to get rid of them). But this is just a presumption: copying errors that happen to break a strongly held letter-adjacency rule would look just the same, and yet would not be meaningful.
In my opinion, the challenge Voynichese presents is to find a way of explaining its pervasive low entropy: its high local predictability. People who concentrate on the statistics (such as my late friend Glen Claston) tend to get drawn to the rarities in the hope that this is the locus where the real meaningfulness happens: but I think this missing the systematic point, which is that these oddities may well be nothing more than scribal accidents that obscure the highly-structured nature of the rest of the text.
Just because qa is rare doesn’t mean that it is necessarily intentional or meaningful.
SirHubert: isn’t the real mystery of Voynichese that EVA – where the transcription letters were largely chosen to match the letters they resemble – is so pronounceable? Just something to think about, anyway. 🙂
Thing: I thought that other people had replicated the Sukhotin tests since Guy but with different (and contradictory) results? I’ll have a look through the archives, see what I can dig up…
Nick & friends: How would you write Christmas or Happy New Year in EVA?
The opening portion of the VMs is written in an unexpected language. A language well known to the author and his peers. A language of illustration, not of words. And that language is heraldry. For VMs purposes there is a combination of armorial heraldry, ecclesiastical heraldry and specifically relevant historical events and in the illustration of VMs White Aries the representation identifies two historical persons. Independent positional elements confirm the identification. Each of these elements must be properly aligned and each of the elements has been placed in the proper quadrant and in the proper sphere. This is a clear example of intentional construction.
While the heraldic references are quite specific in their details, a language dependent on illustration sometimes depends on simple basic representation. Connection is shown by actual connection. As in the example from White Aries where the armorial insignia and the patterned marker in the circular band of text are uniquely conjoined. Just two random rectangles that share a common side? No way!
What percentage of cardinals, between 1300 and 1500, would not have known the origins of the red galero? On the other hand, in the modern day, papelonny is a fur so obscure, it is currently omitted in various references, along with plumetty. Ignorance of heraldry blocks the correct historical identification and prevents relevant investigation in the VMs.
Heraldry is at least as relevant to VMs investigation as alchemy and astrology. And the initial examples of heraldry are not crypto-, they are entirely superficial. Take the first example on the first page (Pisces 12 o’clock). According to the hatching system of tincture designation, this can be identified as a proper heraldic pattern. Other considerations set aside, here is heraldry in the VMs. The all blue shields of White Aries are valid heraldry and identifiable to Berrington of Chester. And there are a few other patterns to be examined. Heraldry is also significant in church history, not only the origin of the red galero, but also the Fieschi position as two of the earliest arms-bearing popes, an oversight that was much later ‘corrected.’ Favored heraldic positioning is also an apparent factor in placing both of the blue-striped patterns together in the particular, favored quadrant, the heraldic upper right.
While I do appreciate the opportunity to post my investigations, the actual purpose, an inquiry into what the inclusion of the heraldic complex reveals about the author’s intentions, remains a one-sided discussion.
Sir Hubert: thank you for that link to Guy’s article. It seems that Sukhotin’s algorithm is simpler than I imagined. I would like to try it for myself, but sadly I think my programming skills are even simpler… I will give it a go some day though.
Nick: if you can find any outcomes from applying Sukhotin’s algorithm to the Voynich Manuscript, that would be great.
Sahagun and the Florentine Codex. You will find all you need for translating B-408’s contents (all twelve chapters).
Nick, it’s not even clear to me why a vowel classification would converge on a set of characters for any sensible transcription of what is otherwise an unreadable text.
But I am curious to know what the results become when we take “ch” and “sh” to be single characters, and similarly for the “i*n”. I’ll repeat the test for these modified EVA alphabets.
Also looking forward to checking which vowel assignment ranks highest under Sukhotin’s algorithm.
When we treat “ch” and “sh” as a single character – lets call it “u” – the results are unchanged.
In that case the most likely vowels are EVA { a, e, o, u }, with EVA “y” as a likely fifth (if there is one).
The same is true when we treat “in”, “iin”, “iiin” as distinct characters (e.g. 1, 2, 3 respectively), though EVA “d” becomes a more eligible vowel (possibly because of how often EVA “dai*n” occurs).
Here are the results of applying Sukhotin’s algorithm to the VM.
To make sure the algorithm was correctly implemented, i ran it through Dante, Pliny and the latin Bible:
Dante – [a, e, i, o, u]
Pliny – [a, e, i, o, u, y]
Bible – [a, e, i, o, u, y]
So it works pretty well. When applied to the VM’s EVA transcription, it identifies the following characters as vowels:
VM – [a, g, h, n, o, y]
The characters [a, h, o] that passed my previous vowel test also pass Sukhotin’s, though notably EVA “e” is now missing. Why? One reason is that about 1/3 of all occurrences of EVA “e” are followed by another EVA “e”, which decreases its character-adjacency count.
Here are Sukhotin’s vowels for VM-A and VM-B:
VMA – [a, g, h, n, o, y]
VMB – [a, f, g, h, n, o, y]
It’s interesting to see that A and B folios share vowels, despite being quite different in some aspects (e.g. occurrence of the suffix “edy”).
Finally, as with my previous test, Sukhotin’s classification is the same when EVA “ch” and “sh” are treated as one character, as well as “in”, “iin” and “iiin”.
Apologies for the spam but i have a small correction to make on my previous post. Sukhotin’s algorithm, unlike my initial test, is in fact impacted by taking “ch” and “sh” as a single character “u”.
In that case Sukhotin’s vowels are the following:
[a, c, e, n, o, y]
Under Sukhotin’s algorithm, EVA “h” is a vowel if it is a distinct character, and a consonant if it is merely a stroke. Sukhotin’s result? Inconclusive.
Job:
The text may currently be ‘unreadable’, but it most certainly isn’t random or unstructured, nor is it necessarily meaningless.
As far as I can see, Sukhotin’s algorithm is based on the (empirical?) observation that, in many languages, vowels tend to be surrounded by consonants rather than by other vowels. If I produced a text composed entirely of nonsense-words from Jabberwocky, I’m sure Sukhotin’s algorithm would correctly identify the vowels in them, because although the words are meaningless they are perfectly pronounceable and share the characteristic vowel/consonant structure of English.
So, Nick correctly asks: “Isn’t the real mystery of Voynichese that EVA – where the transcription letters were largely chosen to match the letters they resemble – is so pronounceable?” It’s interesting, but I’m not sure it’s that mysterious. What this means is that some of the transcription letters are behaving as vowels, at least as far as pronunciation is concerned (their function within a ciphertext may be completely different). So, again, it’s not that surprising that Sukhotin’s algorithm picks some of them up.
With apologies if I’m repeating myself, Sukhotin’s algorithm will probably pick out the vowels in a monoalphabetic substitution cipher. It won’t work properly with transposition, including anagrams, and won’t work at all with polyalphabetic systems. And given that the text here is, as surely as God made little green apples, NOT a simple monoalphabetic substitution cipher, Sukhotin’s algorithm is not going to tell us anything very useful in these circumstances – except to point out that some characters in EVA do behave as vowels in some respects (which is interesting, but which we already knew).
SirHubert: for me, the mystery is that the letters that resemble vowels do vaguely / “weakly” work like vowels. If it was a language, you’d expect them to strongly work like vowels: while if it was a cipher, you wouldn’t normally expect anything like vowels to appear in the cipher alphabet.
So it’s something unlike normal languages, and something unlike normal ciphers. 😐
Hi O*o*t*b,
have you found heraldic imagery elsewhere in the MS or only in f71r? What about the non astro sections?
What about the other two people wearing identical hats to your “galero” in different colors on that folio?
Also, I looked at the arms of Berrington and they have hounds, a bird’s head and wheat, in a mix yellow, blue, red, black and white, not “all blue” as you claim, and nothing special is going on in the upper right quadrant (three bundles of wheat). How does that relate to 71r?
Nick:
“If it was a language, you’d expect them to strongly work like vowels”
Exactly, and they don’t, because it isn’t.
“So it’s something unlike normal languages, and something unlike normal ciphers.”
Exactly.
You see? Statistical tests can get us somewhere!
SirHubert: …which is where, exactly? 😉
Which is, or should be, to a position where we rule out a natural language or simple cipher, meaning that we can get on with looking for alternatives. Specifcally, alternatives which look superficially like a natural language rather than a ciphertext (try pronouncing an ADFGVX message) or anything occult or consciously mysterious (Devil’s Handwriting).
Now, if you’ll excuse me, I have a volume of balneological poetry to finish
Nick: Nope, it is your presumption that it is possible to “identify blocks of the plaintext that stand some chance of appearing in other manuscripts”. Your statement “my prediction is that the poem was originally 8 / 8 / 8 / 8 lines” implies that one Voynich line stands for one plain text line and that one Voynich word stands for one plaintext word.
If EVA “o” is the only possible letter after “q” this means that it is possible to replace all instances of “qo” with “q” without loosing any information.
The entropy of the Voynich manuscript is already low. Rene Zandbergen wrote about the word entropy “Voynichese has marginally less information than classical Latin” (see his page about the word entropy on his website). Prof. Craig Bauer also says something about the entropy of the Voynich manuscript in the youtube video “The World’s Greatest Unsolved Ciphers”. As more information you wipe out as less likely it is that the resulting word tokens still contain enough information to represent a word in any natural language.
Things would change under the hypothesis that the Voynich manuscript contains a code and that a Voynich word does encode a letter (see the work of Jürgen Hermes). Under such a hypothesis it would be a good idea to search for meaningless dummy letters. Eva “o” after “q” could be a used as candidate for such a letter.
Torsten: the block paradigm is clearly a hypothesis (i.e. that we should be looking for block-level source matches rather than word-level source matches) rather than a presumption. It is built on presumptions, though nothing like as clear as the two you mention – I suspect that one Voynich line stands for one plaintext line (though I’m neutral about line order), and expect that unusual words in a plaintext might well be sufficiently preserved through whatever process is involved to be recognisable as such in the Voynichese text (though again, I’m neutral about word order, and I don’t believe that a 1-to-1 mapping will be enough to tell the whole story here).
But seeing as that is a broad enough blanket to cover 99% of cipher and language theories, I think you’d have to be a fairly full-on hoax theorist to take exception to it. 🙂
Greetings Goose,
Yes, check of the patterns on the tubs of the characters in the outer circle of Pisces and also Dark Aries. Given a standard interpretation based on the designation of coloration by the hatching method, valid heraldic representations can be seen. Some are reasonable representations, some are rather sloppy and there are others that don’t really make sense. In addition the Petra Sancta dates to 1638 CE and further investigation reveals additional difficulties and also some discoveries. The point is that apparent heraldic patterns and designs are to be found found here and the investigation of heraldry in the VMs is a valid undertaking. This does not mean that every pattern has been used in historical situations, just that they are copacetic with rules of heraldry – more or less – rules have changed.
The heraldic investigation and the textual evidence that I have been able to relate to it is all contained in the Zodiac pages: Pisces, Dark Aries, White Aries and Cancer.
Ecclesiastical heraldry is a separate field from armorial heraldry (though in this investigation, both apply). Check out ecclesiastical hats. Color is used to designate rank in the hierarchy in the case of the cardinals’ red galero. Or it is used to designate a separate order within the church, where white galeros identify Premonstratensians (White Canons). There is some interesting history there. Their potential illustration in the VMs would indicate the author’s knowledge of their existence also.
I got Berrington out of a reference myself, but that is all I know. I can only suggest that like the Genoese popes and the early Premonstratensians, and other identifications, that the data represented in the VMs is quite early in the history of heraldry, and there is a lot of time were subsequent modifications might have occurred or a new insignia originated. The insignia you have discovered, though valid, may not be old enough. Accepting the reference as it was given, this is merely an indication that technically heraldry is present in the VMs, even when it doesn’t seem to be.
Favored heraldic placement is a separate topic, not related to Berrington. It is another indication how the influences of heraldry are shown in the VMs. By definition favored heraldic placement is in the upper right quadrant of the armorial insignia (in a smaller canton or in a chief – all the way across the top), Divide the White Aries illustration into quadrants. Not only are the identified figures in the proper hierarchical positions, they are also both in the favored heraldic quadrant.
It’s not one in a million, but it is still not an accident. You may still think, on the basis of this evidence, that there is still a tiny chance that it is an accident, but then there is the papelonny and that is no accident. None of the elements involved in the historical identification is present by accident. Each is in the proper quadrant and the proper sphere.
Heraldry is not important. What heraldry reveals is a glimpse of the author’s intent, and that is what should be important. Does the Genoese Gambit have a purpose? Why is this complex construction with these historical connections found in the VMs?
SirHubert: a cipher constructed such that its covertext resembles a lost language – what kind of idiot would have thought of such a plainly ridiculous idea? 😉
The interesting thing about vowel identification is that the results tend to be weak and inconsistent. Different algorithms tend to produce the same correct results on natural language, yet on the Voynich Manuscript they are different.
For example Reddy & Knight’s 2-state bigram HMM algorithm identified the word endings as vowels and everything else as consonants. Very different from Sukhotin’s results. Kevin Knight’s trigram generative model identified the following as possibly vowels: , , , , , , , , , and with plenty of borderline cases.
And yet these algorithms all identified a, e, i, o and u perfectly and unambiguously in English texts with only “y” as a borderline case!
Job: You asked about inter-word grammar; Kevin Knight also tested for this. His algorithm split Voynichese into ten types of word (parts of speech?) but this was rather weak and applying it to the text resulted in no discernible phrase structures (so pure codebook encryption seems unlikely).
Interestingly, he noted that each category had words that looked similar and that correlated with certain topical sections of the manuscript but not others. Perhaps an accidental precursor to Montemurro’s study about semantics in the manuscript?
Pardon my messing up of the list of vowels in Knight’s results. I should have avoided angular brackets.
The list is “y”, “r”, “ir”, “iir”, “n”, “in”, “iin”, “iiin”, “m”, “im”, “iiim” and “g”.
Nick: well, that would be one possibility of course. You must tell me where you found it :-). But I did phrase my suggestion quite carefully so as to encompass others.
I just wonder whether the pronouncibility may have mattered. And I also keep returning to Menno’s question: why encipher a whole book, especially when the images suggest the contents aren’t inflammatory, rather than just the key bits?
SirHubert: apparent pronouncability is one part of the mystery, the odd juxtaposition of apparently-mainstream-vowel-like letters with definitely-made-up gallows characters is another… put it all together, and there’s something that’s so consciously ‘staged’ about the mise-en-scene that it keeps feeling to me as though we’re missing a much bigger point.
Brian: thanks for that, that’s almost certainly the ambiguous vowel-identification results that I was imperfectly recalling. 🙂
I don’t think vowel detection algorithms could yield much of use when applied to the VM.
The problem with using the Sukhotin algorithm on the VM is it is relying on the implicit assumption that the VM isn’t encrypted.
The Voynich words are highly structured and it has been suggested the letters in the words are sorted into some order.
To test the effect of such encryption, I downloaded The Hound Of The Baskervilles from Project Gutenburg.
Using the Sukhotin algorithm on the unmodified text suggests the following vowels:
vowel: “a” 4233
vowel: “e” 9209
vowel: “g” 176
vowel: “i” 5694
vowel: “k” 16
vowel: “o” 3657
vowel: “u” 1568
vowel: “y” 419
vowel: “‘” 130
(the numbers are the sums at the point at which the vowel is detected). I included the apostrophe as a “letter”. As can be seen, aeiou are much more likely to be vowels than anything else.
If the letters of the words are sorted into order (apostrophe last), then the results are quite different:
vowel: “a” 1246
vowel: “e” 6768
vowel: “g” 274
vowel: “l” 927
vowel: “n” 3221
vowel: “r” 4525
vowel: “t” 1844
vowel: “y” 107
vowel: “‘” 18
Now aenrt are the most likely vowels. That a and e appear as vowels is probably because they are likely to be at the front of the words when you sort the letters into order.
So eliminate the vowels altogether?
From my point of view, the result may be as complicated as everyone suggests, but (as I see it) this book is plainly meant to be of daily practical use, and for the sort of person who would not want to spend time deciphering each page before using it. This especially of the botanical imagery.
So maybe the mode is weird, but I feel it can’t be so very complicated. A bit like Einstein’s comment on the natural world and its hypothesised creator – stranger than we can imagine, but never perverse.
Tiger: exactly, vowel detection algorithms require the letter order of a natural language to be maintained. Transposition messes this up, and it’s very helpful that you’ve demonstrated this.
Diane: okay, you find the vowels, and let’s get started. Or are you suggesting a purely consonental alphabet, which is possible but doesn’t explain why Voynichese is superficially pronounceable?
SirHubert, As long as it’s a game, given my non-existent qualifications as linguist or cryptographer.
But if there’s an algorithm for identifying vowels, is there not a complementary one for identifying consonants?
Many languages have been written ignoring many of the vowels, but none without their consonants.
Even in vowel-rich English, wds cn b rcgnsd wtht th vwls so long as the spaces between units are provided…
hmm not sure where I was going with that. Maybe Paris?
Now, now, Sir Hubert and Diane: Imagine the difficulties Sahagun and other manuscript writers encountered when trying to spell the names of some of the gods in “New Spain”:
Huit zil o poch tli
Tez cat li po ca
Ci ua coa tl
Tl al li o ll o
Chal chi uh tl i
X iuh te cu tl i
M ac uil xo chi tl
Just to name a few of the goddesses and a god or two.
Most interesting to me is Ch al chi uh tl i (the Jade Skirted goddess of the waters).
And then there is an item (download from the WWW):
Fig. Sahagun’s recording of the Tonalpohualli (Real Palacio Ms., fo 228r)
Table: Good or bad quality of the thirteen positions.
The cover of the soft-bound book (published by The School of American Research and The University of Utah: Monographs of The Schools of American Research, Santa Fe, New Mexico (First paperback edition 2012) has an illustration of a circular body of water. The midsection of the illustration is obscured by the elaborate title and names of the modern publication’s translators.
It has been immensely useful to me because several of the gray-black illustrations are accompanied by handwritten discussions of the illustrations: example 33 is identified as “Pantitlan” (Tepictoton). My translation of the word Tepictoton is Depiction.
So, my “Christmas Resolution” is that I resolve not to post again until my New Year’s Resolution ! Cross my heart, Nick!
TigerOfDarkness/SirHubert: Yes that’s exactly what I’m saying. I’m not applying these studies to actually detect vowels. I’m interested in using statistics to give us hints as to exactly what the text is.
By assuming it is a natural language and applying vowel algorithms, we get bizarre results that don’t match up with the results from actual natural languages. So it is less likely to be a pure natural language, or at least with vowels marked.
Also no need to trouble yourself with hunting and downloading text, I’ve pre-collected sample texts in 52 languages. You can obtain the sample pack in one easy download at briancham1994 dot wordpress dot com slash voynich-resources .
Perhaps running the algorithms on the abjad and abugida text samples in that pack might lead to some insights as to how it performs on non-alphabetic languages?
SirHubert/Diane: Reddy & Knight noted that the only way a pure natural language could fit their results is if the vowels were not marked, and indeed their statistics fitted better with Arabic (abjad) and English with vowels removed. And yet that has problems in itself – as you say, Voynichese is superficially pronounceable. And the set of glyphs that make it pronounceable are clearly designed to resemble vowels (a, e, i and o).
Nick is right about the mise-en-scene being very staged. Add to that his note about aiin/r blocks being fake page references.
Diane: I think a consonantal algorithm would just detect the vowels and give you the inverse set of glyphs. The patterns of vowels make them far easier to detect algorithmically.
Brian: yes. The biggest problem is choosing the right test and knowing what you’re testing for. Otherwise it’s a bit like running a brain scan to test for a broken leg.
You might be able to modify a test to find vowels in Arabic. (I dislike the term ‘abjad’, but it seems that Wikipedia recognises the term as an alternative to consonantal alphabet, so what do I know?)
Removing vowels causes more ambiguity than you might think. Mediaeval Arabic historical and geographical works are full of exasperating entries where we can’t fill out the consonantal skeleton of a non-Arabic name. There’s even an instance in the Quran, if I can say so without causing offence. A herbal, or a scientific work, doesn’t seem the kind of place you’d want to risk that. (English is even worse – you lose an indefinite article and a personal pronoun completely).
Diane is, in my opinion dead right that there is a point at which an obscure or enciphered text becomes too difficult even for its author or encipherer to use. But on the other side we have to postulate a text which is in such an obscure language or enciphered by such a tortuous process that the illustrious codebreakers and linguists who’ve studied it couldn’t nake anything of it. Not much space in between.
Finally, for anyone who might care, there is in fact at least one example of a script which doesn’t represent all the consonants in the language it represents. The Linear B script is a syllabary, where every symbol stands for consonant+vowel. But Greek has lots of words which end in consonants. So the convention in Linear B is to drop them, and thus ‘Knossos’ is written as ‘Ko-no-so’ with no final ‘s’. This also happens with some letters in the middle of words, which is why ‘labyrinth’ is written without the ‘n’ as ‘da-pu-ri-to.’ (If anyone is still awake and is wondering why the word starts with a ‘d’ rather than an ‘l’, the reason is terribly interesting, but possibly only to people who find historical linguistics terribly interesting.)
Between cipher and natural language exists the codified language as is seen in mediëval scientific herbal medicine, Chemistry, astronomy/astrology and even politics, called nomenclatura. I guess words with special signs K, T, P, F and cKh, cTh, CPh, cFh indicate such items in a nomenclatura.
Roger Bacon has been the first to systematize sciences using codified names for chemical substances for scientific exchange purposes.
Greetings, Menno
SirHubert,
Does your saying ” ..we have to postulate a text.. “(and so on) imply that there is an expectation the written text should reveal either a slab of fairly grammatical prose, or formal poetical form(s) or both?
If so, is the expectation merely pragmatic because non-standard uses of language can’t be reconstructed, or is it an assumption derived from the history of cryptography?
SirHubert: The results of running a brain scan on a broken leg would at least indicate that what you just scanned was probably not a brain. If the body part was a complete mystery that would be useful information in itself.
There’s a variety of writing script types in my sample pack for people to test on.
Brian: by that definition, Voynich research is absolutely awash with useful statistical information, an ocean which people seem to add to every day without the tide ever getting higher. Good luck swimming in it! 🙂
Nick: All information can be useful, it just depends on how you decide to interpret it. Unfortunately without a clear frame of reference to start you off*, it is indeed very much like being thrown into the deep end! 😛
*Well there’s “this is clearly an old cipher” and “this is clearly an unknown language” but those didn’t work out as well as expected.
Brian: the problem comes when your data are misleading. My patient turns up with a broken leg. I run a brain scan. This gives the all clear. I conclude that the patient’s leg is not broken.
I have no problem with validly obtained data which don’t necessarily have an application yet, of course.
On a different topic, browsing through some digitized manuscripts I found this plant very much in the “VM style”:
http://www.e-codices.unifr.ch/en/csg/0754/164
Job: did you see the pen trials at the back? 🙂
http://www.e-codices.unifr.ch/en/csg/0754/168/0/Sequence-659
If a patient of mine were to arrive at the hospital with a broken limb, a scan would also be ordered for the brain and lungs.. Blood clots have a way of migrating at the speed of the heart beat.
PS cod sang 754 is very much similar to B-408’s squash and psyllium specimens.
Pen trials — of course! I won’t bore you with the ‘trials’ a southpaw (left-handed) person would have to scratch in order to get the ink flowing to the nib.
Sir Hubert (?) or ThomS (?) may remember my previous plaint?
SirHubert: Agreed, it’s all in the interpretation.
Job: Yes that illustration does look similar. It even includes the messy colouring outside the lines. 😛
p164 – What I want to know is why “Agrimonia” is repeated, with the second instance looking more like Voynichese. The lines are more sloped backwards (resembling “i”) whilst the rest of the manuscript’s writing and even the “m” and “n” on the same page are upright.
p168 – The doodles (and maybe some of the handwriting) do resemble some marginalia style if that’s what you’re saying. In the block near the middle there are three words in a row with word endings resembling Voynichese “in”. I know nothing of medieval handwriting, what would that mean normally?
Nick, I glanced at it, was busy looking for “pox” adjacent to one of the many “leber”s.
How about the doodling on this folio?
http://www.e-codices.unifr.ch/en/csg/0754/153
Elsewhere i found a familiar figure drawing:
http://www.e-codices.unifr.ch/en/fmb/cb-0103/124r
A Libra, a Sagittarius with crossbow and lobster cancer:
http://www.e-codices.unifr.ch/en/csg/0827/18
http://www.e-codices.unifr.ch/en/csg/0827/17
A few phase-of-moon diagrams reminiscent of f67r2, an apparently inverted T/O map and a depiction of what i hope are the winds:
http://www.e-codices.unifr.ch/en/csg/0827/259
http://www.e-codices.unifr.ch/en/csg/0827/172
http://www.e-codices.unifr.ch/en/csg/0827/257
http://www.e-codices.unifr.ch/en/zbs/SI-0167/26r
Also, a 7-rosette page:
http://gallica.bnf.fr/ark:/12148/btv1b8529692z/f27.item
Finally, a cipher in the correspondence of Ferdinand of Naples:
http://gallica.bnf.fr/ark:/12148/btv1b8478962p/f115.item
Haha “Lake Constance Region”…
From 1661 but a complete wind-rose, giving names in most European languages.
http://www.swaen.com/zoomV2.php?id=24547&referer=antique-map-of.php
It’s zoom-able
* 1650. The 1661 version was published by Cellarius.
Brian, the i-like stroke is intriguing.
The “Voynichese” has added annotations throughout the manuscript – apparently rewriting a few words.
On the last page, we also see the same “x” that occurs in the VM’s f116v, in “vix”. It’s repeated three times in a row.
Nick, I am down-hearted. This morning I visited Rene’s discussion of folio 35 (r or v, hard to tell). He displays what I have identified and translated to be the saffron crocus. I’ve also referred you to Saffron Walden (a small historic village somewhat in the vicinity of your town). (Yes, they were able to grow saffron corms in England!
So, today, can you show us what folio 35 v and 35 r contents look like? Just so we all might be on the ‘same page’ ?
BTW: The defining features of the illustration on folio 35 are the sex parts being prominently displayed (from which we get saffron). Also prominently displayed are the flat-bottomed bulbs, correctly called ‘corms’..
So, the first line of discussion, (Latin) below the blossom, begins with “Speci-am” and ends with “cro-cos–ce-aes-ce-am”. on the stem.
No code.
So, add a little ‘seasoning’ to your Holiday-Season dining event!
Yassou!
Eph-a-ris-to!
ps: I seem to recall that the Avery Hill Glass-domed Winter Garden at Greenwich University is hosting a ‘Holiday” event? Maybe poinsettias surrounding the palm trees?
😉
Job: Yes there are a few intriguing things in that particular manuscript. I see what you mean with the pseudo Voynichese text throughout, though none are as close as the “Agerimonia” next to the illustration. I’m mainly wondering about the purpose of labelling it twice.
By the doodling you mean the dotted loopy-loop thing? That’s the first thing I noticed. The rest, not so convincing as a direct correspondence.
Hans P. Kraus, there’s a familiar name. Aegidius Tschudi, that one isn’t.
Description mentions a relation to “2° Cod. 572 of the Staats- und Stadtbibliothek Augsburg” which I would like to see, but is not available on their digital collection. Only available in person? 🙁 Can only find second hand information that it’s about “wonder drugs” derived from plants, snakes and vultures.
Job,
I like the ‘familiar figure’ – not much like any in the Voynich, but very like some in earlier western manuscripts, with a hint of Byzantium. Is it ‘nature red in tooth and claw’ do you think? 🙂
Sorry – seriously, of course its an ‘Angel of the winds’, and its having three good quarters with one ‘bad’ (southern) would alone predict that this marginal drawing came from the pen of a medieval Christian..
For what it’s worth, p32 is entirely in the backwards-sloped writing.
Brian, wonder what codex 572 looks like.
From i could gather, it’s an anonymous German manuscript from 1446 written in the Swabian and Bavarian dialects.
The contents are classified as “Alchemy Magic Medicine”, and described as “Manuscript dealing with colours, cosmetics, alchemy and medicine”.
The following is given as a transcribed sample of its contents:
“Eÿsen vergulden Wiltu machen ain wasser da man daz eÿsen inn kupfert daz es golt facht So nÿm ainen kupferinn kessel vnd gusß gu(o)ten wein essich dar ein vnd ain tail salcz vnd daz ander tail alun vnd wirf eÿsen dar ein waz du vergulden wellest vnd lasß es wol dar inn sieden so wirt daz eÿsen kupfer rot vnd lat sich gerne vergulden”
Couldn’t find any images though.
Some more similarities on the last page, not sure how typical they are:
– The word “dadaiiid” north of the gigantic “D”.
– The beginning of the word that is like “antifmens” or “mafmens” (third line from top of page). It looks like the thing at the beginning of “michiton”.
– Almost entirely gibberish in pseudo-Latin with a bit of pseudo-German. Very few meaningful phrases (I think).
– Repetitions of the letter “i” in a row.
– The words “maria” and “gas”.
Nick, the last several days I’ve been translating page after page of Fr. Sahagun’s fabulous manuscript — hand-written in two languages, Spanish-Latin and Native South American Nahuatl :”General History of the Things of New Spain”
folio 132/pages 265/266 of 506:
de las getas – nanacatl – : edible and/or toadstool
folio 134, page 269 of 506:
de las yeruas – azoquilitl – isphagol – psyllium/plantago ovata
folio 171, pages 283, 285, 343, 344:: tz-pi-pat-li, te-cz-om-pa-tli, czompatli, quetz-ue, quetzal uex otl, haquatzin (corn or maize in all its identities and uses.
I had to stop, here. I’ll give you a few days to ‘digest’ my latest endeavor. I shall now proof-read b4 ‘sending.
Cheers for the New Year !
beady-eyed wonder
I’m referring to a two inch thick bound book (500+pages) of handwritten script (two columns per page, one column Spanish/latin and the other column Nahuatl. Depending on the subject matter, the columns are accompanying illustrations of insects, birds, bees & ants, wasps, termites, mushrooms, toadstools, cultivated plants (corn/maize most important). I am just now venturing into his series of discussion dealing with the wildlife (jaguar and ocelot, coyotl, rabbit, to name a few…..If you take a good look at that book, you should be able to recognize a good many of the mystery discussions in the so-called “Voynich” manuscript (B-408).
;-^
Yesterday (and today) I’ve been reading “Culltivated Landscapes of Middle America on the Eve of Conquest”, Thomas M. Whitmore and B. L. Turner II.
Much more validation of some of my “Voynich” and Florentine Manuscript’s translations of botanical discussions.
Nick, I’m getting ready to do a page by page translation of Boenicke ms 408 botanical illustrations and discussion. I’ve done my best to clarify the difference between the tomatillo (folio 1v) which has a white papery husk, and two other types of tomato (one of which we call the ‘cherry tomato’ (no husk, and not much larger than a cherry) and the standard tomato which can be the size of a tennis ball, and even the ‘beefsteak tomato: huge, heavy, fruit which quite often has a sling to support it until it is ripe enough to pick — and feed a large family.
As I’ve mentioned before, the triple-layered commentary (ixtomatl) (physalis ixocarpus) and (physalis Peruviana) all refer to the ‘husk tomato’. Confusion arises when different professionals and their assistants get thoroughly confused because they apparently have not been able to sample the fruits being described by Friar Sahagun and his assistants.
One of my earliest childhood memories were the large vegetable and flower fields of the Burpee Seed Company. Burpee’s fields stretched from Lompoc to nearly Santa Maria in California. Their headquarters were/are in Pennsylvania. Lompoc has a flower festival every spring (in April, I believe.
What came to be a recent shock to me was that Lompoc shares a border fence with Vandenburg Air Force Base. VAB is still an active air base . Burpee flowers still bloom near the first Spanish Mission, La Purisima, which eventually was crumbled in an earthquake, and was rebuilt shortly afterward. Both “La Purisima” mission sites were serviced (gardening, field work, hauling fertilizer, carpentry adobe brick-making), by enslaved Native Americans. Very few of the laborers ever had a ‘day off’ or festival (except a mandatory daily church/catechism).
Their living quarters were windowless cells into which they were locked every evening. Pallets on a dirt floor and a chamberpot in a corner.
I know all this because our fifth-grade teacher arranged a field trip. She was also our tour guide (and independent of the Mission).
Another enterprise which used local labor, in more recent years, was the diatomaceous earth mines. I think the mines are still operating — because diatomaceous earth is/was used in swimming pool filters, primarily, but I believe the mine still supplies the powder for other uses.
19th line of the specimen page above this comment space: that strange figure which looks like an ampersand leaning on a crutch is a formulaic word ending for the sound of ‘it-ius’. An example of a whole word would be ‘pro-pi-tius’. A whole word could also be said with that one strange syllable: ‘deus’ or ‘dios’. See — just how stu-dious I can be?
tongue in cheek, as usual — but always respectful !
bd
Otherwise, I think what we are seeing is the product of a very weary scribe, who probably lost track midway in the dictation and just gave up. That — or maybe he just wanted a drink of water! (Maybe day-dreaming? After all, the discussion appears to be all about bathing ‘babes’ and waterfalls).
Remind me to tell y’all about our visit to a hot spring waterfall near the Colorado River — and a Boy Scout troop who came trooping through……
😉
Nick, I’m hoping that I add a little humor (wry or rye –punny or otherwise — what I can read with only one of my eyes) .Poem-wise?
bdid1dr
“1421 – The Year China Discovered America”: Gavin Menzies
doesn’t go over the ‘same ol’ same ol’; but he does discuss some stops along the way (river and waterfall with ‘ladies” on duty to help with the bathing).
Interesting!
😉
Well, I finally got a round tuit:
Line 19 of the document displayed at the top of this discussion, can be roughly translated to latin for the words or syllables for:
aquatiligeus cre es aes geus ceceaesgeus —
aquollegeasgeus ox-a-tius
“Psalmodia” Another of Fray Sahagun’s publications written for the benefit of his Native American parisihioners. Maybe a type of hymnal? Perhaps Kevin Knight is familiar with this publication?
Could it be that some of B-408’s folios are, again, rough draft notes for a book written in the native language so that they can participate in the readings/songs?
I truly am 1-dr-ing
bd
The poem block appears to be a classic rune song heavy on the trochaic meter, alliteration, and depth symmetry. Torsten Timm and Jurgen Hermes certainly picked up on the repetition inside blocks of text but gave little context to that phenomenon. Rune songs and joiks have their own sort of internal logic which show up even if you don’t know the meaning of the text. The manuscript reflects the way knowledge had been transferred for hundreds of years previously, as if it sprang directly from a long-held, formal oral tradition. Although Voynichese is most likely an extinct language, it does contain loan words from neighboring languages. Its base Finno-Ugric, it is charged throughout with derivations of Old Norse, some Slavic, and a few Latinate words like epais. Alkeisa, for example, is utterly Finno-Ugric as is ukkopekka, samalla, and taikuus. But alla, ellar, and “som som,” which are still used by the Swedes, appear very commonly throughout. Yet there is also, in this case only in the plants section, forms of kepka, a Slavic (Russian) word for cap. Eparlasai takes a shade from Latin for liver and tacks onto it a suffix derived from Old Norse/proto-Germanic: las – to lock. Found on one of the herbal jars, its translation may be a fixative for liver of sulphur, which is an obsolete chemical term. All this, along with myriad visual clues, strongly suggests a language related to Elfdalian, Meankieli, or Kainuu, possibly Kainuu Sami.
Limoges,
I would appreciate it if you could point to some examples of imagery in the manuscript which have close counterparts in works produced by the northern peoples you would prefer had produced our manuscript.
There is no necessity to posit a germanic or norse origin for the manuscript’s content. .
Limoges: I have followed your webpages for some time, but am as yet unconvinced by anything to do with Finno-Ugric and Old Norse. I cannot honestly understand how it could be a valid approach to write off 99% of the Voynichese text as “most likely an extinct language”, and then relying on a plausible-feeling transcription reading to select the rest: the same approach would easily yield as many valid words (1%? 2%?) in almost any world language, of almost any era. As an example, the list of “names” on your site (“It’s in the Names” etc) makes no sense to me at all as a proof of anything: it seems linguistically hopeful from start to finish.
On the one hand, I think it’s fair to say that Voynichese is a particularly tangled mystery, and that all current accounts of it (including my own) do fall short in numerous different ways. But I do honestly think the answers to it will depend on our building up a rather more nuanced way of reading it than as an exotic lost language.
Nick, I’m hoping you will take a look at Sahagun’s Psalmodia. There is one example of the HANDWRITTEN booklet — loopy LL’s and half-looped TL’s, figure 8’s and all — I just can’t find it again. I bookmarked it, but may have signed off before the bookmark registered.
Basically, it confirms my alphabetical assignment of the various ‘mysterious’ script which appear in the Very Mysterious Manuscript (I refuse to call it the “Voynich”).
I’m proceeding ‘full tilt’ ahead!
😉
A similar theory might be applied to smaller text segments, if some limiting factor for the selection of specific examples could be applied. Here is an EVA transcription from the outer circular band of text on VMs f71r. [voynich,nu]
The text runs from beginning to end without any line breaks, but sending it that way seems to cause formatting difficulties. I chose to put breaks before ‘aiin’ because it helps to show the high degree of internal repetition present in this segment. That would imply that it would connect to another text with similar internal repetitions.
olkeeody.okody.okchedy.oky.eey.okeodar.okeoky.oteody.oto.otol.oteey.ar.ykooar.
aiin.aekeeey.okeo!keo!keody.okeodar.chy.s.
aiin.oto!keoar.or.ar.al.otol.al.shckhey.oteeeodar.oteody.otol.
aiin.shoekey.sal.al.ald.cheeokseo.q!orky.choly=
Words repeated:
‘al’ three times plus ‘ald’
‘otol’ three time plus ‘oto’
‘aiin’ three times
‘okeodar’ twice plus ‘oteeeodar’
‘oteody’ twice in comparable phrases ‘oteody.oto.otol’ and ‘oteody.otol’.
Limoges:
“Kepka” is not Slavic, it’s a derivative from the Roman family (French “kepi” etc).
Nick,
You originated this thread with the question: “How are we ever going to resolve the mystery of the Voynich Manuscript?”
Your proposed method here then is to match the formatting of the long and short lines on a page of text from the VMs to something of a similar block structure from historically known texts. And something that might exhibit such structure would be the repetition of verses in a poem. We should be so lucky. And then, one would hope, in step three, to discover the process by which VMs text can be understood. If only it were that simple. What are the criteria for selecting the VMs text? A long text, perhaps, with long and short lines either in an internal pattern of repetition or presenting a pattern that matches a known text. That seems problematic.
A second factor might be word repetition, A pair of words and their relative locations in the text might also be of some significance. Paradoxically, perhaps, it would seem that word repetition is favored in shorter texts relative to longer texts.
So as a variant of your block paradigm suggestion, I propose the short block paradigm, Not based on line length patterns, not that these are eliminated, but on internal word repetition. But beyond that, where in the VMs text does one find the right example of such repetition? Surely, there’s a bunch of them.
Another criteria is to follow the information contained in the text itself – provided by the author, The perilous path of Pisces, the path of pairing. The pairing of papelonny. The pairing in armorial, Catholic ecclesiastical history, used in the VMs to designate the pope’s golden key. This is all in the text because it is the author’s work. So if you were to ask me, the problem is not finding the VMs key. The problem is turning the key. See text last posted above. How to make it work? There are any number of ways to split the VMs text based of repetitive words. It needs a plain text with a similar set of repetitions. Got a couple historical possibilities handy? Anything liturgical per chance?
word repetition is a good place to start. What if the Voynich was written in doo wop ?
Ooooo wah, oooooo wah, ooooo wah, oooooo wah,
ooooo wah, oooooo wah, Why do fools fall in love
Why do birds sing so gay
And lovers await the break of day?
Why do they fall in love?
Why does the rain fall from up above?
Why do fools fall in love?
Why do they fall in love?
xplor,
Give it a try. And if you could just dial that doo wop back about 600 years or so. Or better yet find another VMs text segment with a fair amount of internal word repetition.
How did you like the third criterion: authorial guidance? There is a trail of paired clues, because single could be mistaken for accidental,leading from Pisces to a representation of the papal keys that only depends on a standard Medieval interpretation of the VMs Zodiac illustrations. Papelonny is certainly exclusive to a good knowledge of heraldry, but it is verified by standard information. In the VMs the pattern is paired, positioned in the proper quadrant and sphere and also arranged sequentially, pope first. Not to mention historical connections to the particular individuals in question. The VMs Zodiac cannot be fully understood without a recognition of the purpose of papelonny. It is an absolute affirmation of identity and it is a complete confirmation of intention on the author’s part. And then there is the pun, if you get it.
Doo Wop is a metaphor for Ambrosian chants and other singing. Are you sure the Vonich zodiac is correct ? shouldn’t. there be 13 signs.? Maybe the scales you see are from Ophiuchus . My own shield has three Marlots (small birds ) without legs.
xplor,
Hang on a sec. You’re the one who is trying to set this to music. Now it’s “Ambrosian chants and other singing” by the fabulous Three Whatevers, past, present and future. All I said was that the presence of a pattern of word repetition might be more helpful in matching two texts than an absence would be. Internal repetition would certainly help augment any similarities in block paradigm formatting. Repetition of phrases or whole lines of text, as in your example previous, at corresponding locations would surely add further verification. But that’s all hypothetical, if no sample texts have been examined or even selected. Authorial guidance would sure help with that part.
I realize there are other hypothetical interpretations as to how the VMs Zodiac could be arranged. Such possibilities play no part in my investigations. I simply use the material of the VMs Zodiac as it is. I’ve been looking into the use of heraldry in the VMs for a while, so I have seen the use of marlets elsewhere. Are you acquainted with papelonnny? The scales of a butterfly’s wing. The story of Chotard de Chateaubriand.
Papelonny is a discovery I first brought to VMs discussion about a year ago. Prior to that it was not something I would have recognized. Therefore, for me, it is a prime example of how we can look at something repeatedly without really seeing what it means. But then, after the identification is established, things take on a totally new perspective. This applies to the whole of the construction of pairing illustrated by the astrology and heraldry in the VMs Zodiac, supported by a list of elements taken from the Zodiac illustrations – blue stripes and red galeros, etc. Since all this is too precise and complex to exist by accident, it can only be the result of an intentional construction and, therefore, it is an area for further investigation – attempting to follow the path of the author’s intention.. As I see it, this whole construction is designed as a text delivery system, intended to supply a specific segment of text – with that text segment internally designated by the use of historical identification with the pope’s golden key. Should we try it first or just pick something else at random?
A brief survey not only shows that the primary segment of text, discovered by following pairing, heraldry and historical representations in the VMs illustrations, has several examples of triple and double word repetition, it is by far the one segment with the most internal repetition of all the segment transcriptions in the entire VMs Zodiac. Most of these segments have no internal repetition of words, though there are a few examples where one word such as ‘aiin’ or ‘otol’ may be repeated. What are the chances that authorial guidance would lead to such a unique specimen if this was not part of an intentional construction?
Therefore, this is my recommended candidate for Phase 1) Find section of VMs text.
Now on to Phase 2) Find corresponding plaintext with similar internal word repetition. Actual Medieval suggestions appreciated. Language not yet specified.
The Pairing Paradigm:
Pairing is one of the simpler forms of pattern repetition.
There is a path of pairing that starts with Pisces in the VMs Zodiac.
The path pairing is built on pairs. Without an ongoing series of pairs the pathway ends, the paradigm is broken. But at the same time, as long as pairing continues, the pathway is validated. And the examples found follow the criteria of common topic and restricted location.
The dark and dangerous trail of potential VMs misconceptions is now a smooth walkway with pairs of garden lights on either side. As far as it currently goes.
Dear O.O.T.B
Whenever I’m tempted to feel disenchanted by what I see as the increasing rigidity and fossilisation of Voynich studies, I read your communications, which have been allowed to flow sans flaming for so long. Buoys my spirit, I must say.
What is the cosmology of the Voynich? Does it follow Ptolemy or the pseudo-science of astrology ? Are their signs of Sexagesimal cosmology ? What is the best understanding of it all?
Diane,
Well, obviously, Nick runs a proper, tight ship here. Anything considered flaming just wouldn’t make it. But I have experienced the warmth elsewhere, on another list, sometime before your time of participation there I believe. Though any personal attacks are regrettable, some seeds need exposure to the flames before they can sprout. And that is what the investigation of VMs heraldry has been like for me. Regardless of the negation and the frequent silences, the VMs keeps saying yes. Yes with the red galero. Yes with proper hierarchical placement of the two designated figures. Yes with their favored heraldic positioning. Yes with the choice of White Aries. Yes with the papelonny pun.
VMs f71r, aka White Aries, is an interesting page. It is by far the most elaborately painted illustration in the VMs Zodiac. But the question for some is why it has been done this way. I see two reasons.
If the intent is to paint some part of the illustration, but not to have the painted part stand out as completely obvious, then other parts of the illustration need to be painted. It’s easier to hide a tree in a forest, in this case a pair of trees, than it is to hide two trees on a prairie.
The second reason is to reenforce the idea of whiteness. In an unpainted illustration, what color are the objects? Okay, sneaky, zen-like question. They are the color of the substrate: parchment, paper, etc. Although we might call the color null. Now look at f71r. Almost everything that can be painted has been painted – with the most notable exception of the central animal. The illustration has been set up to affirm White Aries’ whiteness. Given that the heraldry found here is patently religious in its representation (behind a radial disguise), there is no question of the role white animals traditionally occupy in religious sacrifice (celestial versus chthonic). Is that why these two elements are combined in the illustration or is it just by chance?
xplor,
My investigations of the VMs originated with an inquiry into the potential use of heraldry in the VMs Zodiac. That is to say, the investigation of potential heraldry, which happens to be located in the zodiac, rather than the wider investigation of the VMs Zodiac, which happens to contain prospective heraldry.
More recently I have been following the path of pairing, which not only includes the heraldry found in the VMs Zodiac, but now includes the pictorial identities of the VMs representations of the Zodiac signs themselves, as presented with some unexpected (non-traditional) manipulations. And again here we are only looking at the use of the pictorial illustrations and not at any deeper (astrological) interpretations.
My only attempts at a broader cosmological investigation concern the matter of placing Pisces first in the VMs Zodiac. along with the unexpected halving of Aries and Taurus. Instead, I’ve been looking at calendar reform and the problems caused by precession in a time when these problems were known, but before the actual Gregorian reforms were put in place (1582). Roger Bacon’s ‘Opus Majus’ cites these problems in 1267. Regiomontanus’ attempted correction was cut short circa 1475. And still there is another whole century when the Julian calendar was still used and precession moved further and further along. But I have made no particular investigation into other aspects that you mention. I would only say that it is my opinion that the VMs author is fully capable of taking standard tradition as we know it and manipulating it to suit suit the necessary purposes – creating a series of pairs while making something that flies in the face of convention. It is the same level of knowledge and skill by the author that accounts for the paired presence of papelonny and for the particular locations of their placement. Or to disguise heraldic insignia with radial distraction and yet retain the essentials needed for identification – of which pairing is the primary part. And it is the same capabilities that account for the presence of the pairing paradigm in the first place. As long as there are pairs, the investigator is on the author’s intended pathway.
Having established such a path, the goal then is to follow it. In other words, explore, but follow the path.
OOTB:
Your posts here have always been well-written and courteously expressed. The reasons I don’t comment on them very frequently have nothing to do with whether or not I think they are interesting, and have an awful lot to do with the fact that I don’t know very much about heraldry!
One would expect the heraldry to be on the ex-libris not buried in the text.
SirHubert & xplor,
SirHubert, first, that is perfectly understandable. Each of us have our interests, and heraldry is generally not on the VMs agenda in prior investigations. It has no individual reference listing. I believe that constitutes a serious oversight. Heraldry in the VMs Zodiac seems to be my own little branch of investigation. And most of the people, who do respond, take a sort of dissuasive position: “Don’t be ridiculous. You can’t really believe that.”
Heraldry is indeed a vast topic, with only a rather small (but interesting) portion of it relevant to the VMs. Best to start with the general rules from Wikipedia. Also compare what the VMs has to offer. It is not much, only three zodiac pages, just a selection of simple patterns, for the most part. So as far as heraldry is concerned, all the chapters about charges and mottoes and crests etc. are not relevant to this investigation.
*
xplor,
If you have captured the standard expectation, then apparently the VMs usage of heraldry does not conform to it. Yet in both cases, heraldry is used for identification, which is the standard usage. In your case, to confirm ownership. In the VMs, something a bit more subtle, and more complicated. It is something highly dependent on the reader’s recognition and hidden by layers of obfuscation (2 intentional, 1 otherwise). A superficial investigation reveals nothing in particular. And looking into hatching lines in the VMs examples, one would think the work was done by an incompetent imitator. [[Later, from the other side of papelonny, it looks a good deal different, crazy like a fox, as they say]]
So, giving this further consideration, there is a second view. Look at White Aries with all radial influences removed. Look at the blue-striped insignia in relation to each other and to the innate page directions. A single image of an insignia can be difficult to place, but a paired example, where the lower figure wears a red galero, hierarchically placed and heraldically positioned, and one starts to wonder it the whole story of the Genoese popes is represented here. But that is not the point. Look at White Aries and see the upper figure uniquely attached to a patterned marker in the middle band of text – and also reaching for a similar patterned marker in the band of text just above. Remember that the papal keys come in pairs – gold for heaven, silver for earth. This character is clearly identified with a historical pope and now presents this unique, internally repetitive segment of text. Right out of the blue, there it is. A lot of complex construction in the VMs Zodiac illustrations, if it’s all for nothing.
The pathway to the present juncture is clearly lit by pairs of garden lights. The pathway forward, is about a black as pitch. How does internal word repetition promote a potential solution? How does language choice affect word order?
It’s definitely gone beyond heraldry,
Dear Nick and All,
I used a variant or special case of Nick’s block paradigm to try to match a single text source to a single folio of the VM. I set out some parameters for successfully applying your method: 1) availability of the primary text source, 2) independent secondary text sources & commentaries supporting the findings and 3) cohesiveness of the suggested results. By cohesiveness I mean a networked character, i.e. close proximity in the text source should be also close in the imagery display.
I summarised the methodological approach, the findings and some further thoughts in a document here ( https://figshare.com/articles/_One_picture_is_worth_ten_thousand_words_The_Rosette_folio_s_exhibition_of_elements_and_phenomena_described_in_Aristotle_s_Meteorologica/4287446 ).
In essence, I think that (parts of) the Rosette folio suggest a link to Aristotle’s Meteorologica. After reading quite some of Aristotle’s work (the natural philosophy work mostly) and others in that area, I suggest that some meteorological phenomana are visually displayed in the Rosette folio and I compare the microimage with the primary text source. I focussed on the winds so far -It is by no means finished, so if anyone has any thoughts or comments, I would be grateful for any feedback. Particularly, does the method of block paradigm work in such a context or am I completely on a wrong track?
Nick,
I stumbled up David Jackson’s comments of 7th December 2014 which I will repeat because I think they are relevant to the application of your block paradigm to the VM.
“Medieval Spanish (Castillan and Catalan language) poems were often written in eight line verses called octavillas if they didn’t adhere to a strict rhyming structure or octava italiana if they did (generally ABBC’ DEEC’ structure) or octava real if they had a strict “arte mayor” style, generally ABABABCC structure. Anyway, the eight line poem was popular in the Middle Ages, 1100-1400’s across the Christian part of the Iberian Peninsula.
But there is no evidence of rhyming structure in the text. Let’s look at the ends of the first octavilla:
0x9 0M9 89 ax aw ox aw 0x9 …….” No evidence of rhyming structure there that I can see, but I know very little about poetry……”
I am enthusiastic regarding the octavilla structure described by David but I can’t agree regarding his conclusion that there is no rhyming structure in the putative poem. There are many forms of poetic rhyme and the final characters or string of characters would not have to be the same, or even similar, for there to be a poetic rhyme. To my eye there is a possible rhyming structure, AAABCBCA. I see a possible crib if the text on folio 81r is assumed to be poetry (as a working hypothesis, to be proved or disproved). If f81r is poetry then possible homophones suggest themselves. There are quite a few rhyming schemes, but the number of these is probably manageable.
Byron: to be precise, I don’t believe that the way Voynichese appears to work (truncated words, stylized line-endings, etc) is amenable to looking for rhymes, because those things you would look for as line-endings are precisely those things that Voynichese doesn’t really have.
For me, the point of finding the matching plaintext block would be to match the whole block, not just the rhymes: and thereby crack the whole system.
Nick,
I remember that I have seen manuscripts of Peter of Eboli’s De Balneis Puteolanis with twenty four lines per folio. And the possible poem(s) on folio 81r of the Voynich Manuscript total twenty four lines. Maybe the possible verse on folio 81r consists of 2×12 lines rather than 3×8 lines?
If so, then there are thirty five possible cribs that can be tested.
Byron: good observation! I’m away from my notes on this, but will check as soon as I can…
Nick,
De Balneis Puteolanis manuscript Ross.379 (Vatican library) has two twelve line verse per folio. Manuscript Cod.Bodmer 135 has twenty four lines per folio but these are split lines (two split lines=one full line of verse). Gallica has manuscript BNF Lat. 8161 which has twenty four lines per folio (split lines again).
It has been previously noted by others that whoever drew the illustrations of the VM probably had a copy of the Ross.379 manuscript by his/her elbow.
Hello Nick, briefly on this topic.
An example of a block paradigm can be seen in the manuscript itself. The ONLY place where the two voynichese words of the sentence in f116v can be seen exactly in the same combination is the eighth paragraph of f104r.I have found this “reading” carefully the entire manuscript page-by-page and “word”-by-“word”. When a Voynich Query Processor appeared on
the Internet he gave me the same result.But there is something more interesting.Taking exactly the crosses and the words between them, the two
words are in the same place in both sentences(I exclude the cross above “maria”).For more clarity, see the PDF file below.Whether it can be used
and how I leave it to the specialists.
https://drive.google.com/file/d/0B_T290o0O-dPX0FhN1ZXeWJsYlk/view?usp=sharing
Best regards,
Milen
I have obtained a copy of the 1967 reprint of Cappelli’s “Dizionario di Abbreviature latine ed italiane” and I note that the VM gallows like characters also appear in the bottom line of the 1172 manuscript as well as in the top line.
I further note that, judging from the transcription of this document given in Cappelli page LXIV, the gallows like characters stand in place of various letters (d,s,n,i,f,l,g,c and h) within words and for four digraphs, ii,br, st and de. One gallows like character stands in place of the latin word sancti.
I have underlined the letters that have been replaced in the first (topmost) line by the gallows like characters.
“Die martis tertio decimo mensis iunii in Placentia infra monasterium sancti Savini in solario …”
Similarly, for the bottom-most line.
“Ego leo de Turre sacri palatii notarius interfui et rogatu lamscripti Abbatis hoc breve investiturae scripsi.”
At present I can’t see any rhyme or reason in the use of these gallows like characters in the above example. But I did note that Cappelli lists perhaps 15,000 abbreviations and I noted that in many cases the “abbreviations” are as complicated as word they replace. So why use them? This suggests to me that the “abbreviations” were intended to confuse any casual reader of Latin. Enciphering of sorts.
Byron: the closest we can get to a collective expert opinion is that we’re looking at the raw materials from which the Voynich was very likely fashioned. But the rest… remains a mystery. 🙁
Milen: it’s a good observation, for sure, one that I’ll definitely have a think about. It’s just a shame that we’re more sure about the Voynichese on f116v than about the rest of the text. :-/
But the rest… remains a mystery.
I personally do not find it so mysterious. The whole book follows an intelligible logic.
What do I need, (plants)
When will I get them, (Zodiak)
How and what do I make of it, (bathing women, plant parts)
For what I use it. (Possible recipe part)
The only thing is, I can not read it. ( Not yet )
Nick,
you mentioned that you could not find many poems dealing with balneis. I noted the following example that may not be known to you. I note that the epigrams “…. each consisting of twelve lines, on the baths of Pozzuoli …… the same person composed rather earlier a work on the triumphs of Henry VI, and another on the actions of Frederick II, to which the epilogue of the poem De Balneis Puteolanis alludes.”
Alcadinus “was a physician to the emperors Henry VI. (1190 —1198) and Frederick II. (1212-1250) during their residence in the kingdom of Naples, and died at the age of fifty two. It was at the command of Frederick II. that he composed a poem, ” De Balneis Puteolanis ” (” On the Baths of Pozzuoli,”) in elegiac verse. Of this poem, however, eighteen strophes, or epigrams as they are called, are ascribed to a certain Eustasius or Eustatius de Matera, who is said to have lived under Charles II. of Naples ( 1285-1309), and to have written a work, ” De Natura et Temperie Hominis” (“On the Nature and Temperament of Man”). A manuscript at Naples, written on parchment in the thirteenth century, and beautifully illuminated, contains thirty-four of these epigrams, and is merely entitled ” De Balneis prope Neapolim.” Two manuscripts in the Vatican library, (one of the fourteenth century on parchment, the other of the fifteenth on paper,) both men-
tion Eustatius as the author, and say nothing about Alcadinus ; while on the other hand a manuscript at Naples of the seventeenth century on paper ascribes the work partly to Alcadinus and partly to Eustatius. A paper manuscript of the end of the fifteenth or the beginning of the sixteenth century, in the university library at Marburg in Hesse Cassel, contains thirty epigrams without making any mention of the author’s name. Jo. Elysius, in the beginning of the sixteenth century, (Collectio de Balneis, Venet. fol. 1553, p. 212.), mentions Alcadinus as the author of thirty-one Jo. Franc. Lombardus, who wrote somewhat later, but in the former half of the sixteenth.”
http://www.ebooksread.com/authors-eng/society-for-the-diffusion-of-useful-knowledge-gre/biographical-dictionary-volume-12-ico/page-72-biographical-dictionary-volume-12-ico.shtml
Nick,
you previously mentioned that you were interested in any balneological themed poetry. I have obtained a copy of “Eorum, quae de balneis aliisq miraculis puteolanis scripta sunt. Auctore ioan francisco lombardo neapoitano. Adiecus balneis aenariarum necon locis obscurioibus non inutilibus scholiis. Opus ad Auttore denud recognisum, & lacupletatum. Venetiis MDLXVI”
I obtained my copy (print on demand) from SN Books World, India for US$8.15 postage included. The book (128 pages) appears to contain a significant amount of poetry and I note there is one chapter “De Baln, Solis (an abbreviation I don’t recognise) Lune.” (page 59). Judging from various chapter headings material dealing with the thirty plus baths at Pozzuoli is included, including what appear to be poems. There are several detailed appendices.
Byron: thanks very much for pointing me at that, much appreciated! As with a huge number of print-on-demand books, searching for the incipit reveals various scans on http://www.archive.org/ , such as this one here: https://archive.org/details/bub_gb_D1DfPJSRIwsC
It’s an interesting little book, because it appends a section of scholia (notes and summaries from different authorities) to each of the verses, giving it more of a “balneobibliographical” feel than it would have had it just recited the verses. Finally: I believe the sign you refer to at the top of Cap. XLII (page 59) is simply a decorative ‘et’, nothing more complex. 🙂
If I was looking for a block-paradigm, given my analysis, I would start with the library of the Abbey of St. Gall in St. Gallen. Next on my list would be archives in Basel, Geneva, Zurich, Interlaken, Konstanz, Pavia and Milan. Then finally archives fanning out from those areas or with a preponderance of archival material linked to or originating from those areas.
There has been discussion of the need to find a Voynich Champollion, understandably so. However I think really what Nick is saying with the Block-paradigm approach is that we are still need to find a Voynich Rosetta Stone before progress can be made. As I have stated before I have always been reluctant to accept the idea that in practice the Voynich can be deciphered without other clue (s) than we have now. That is why I have devoted very little time to trying to crack the text. Even if more progress is made to illuminate aspects of the text by means of statistical or other methods then my suspicion is that the range of possible decipherment “solutions” will still be so large as to make the identification of the correct “solution” infeasible in practice (given contemporary technology). So the necessary clue may come in the form of a block-paradigm or something else. My focus has been on questions of authorship and geographical markers as a way of narrowing down the search for clues.
We may just have to wait for someone to stumble upon an important clue in the archives somewhere, but who knows when that could be.
There are many clear precedents in herbal literature for the content and arrangement of information relating to plants and yet no one has been able to relate these readily available resources to the VMS text on plant pages.
This means either that it’s a unique way of representing information OR that the text on the plant pages is something other than plant information. If it is something other than plant information, then how would one recognize the right block-paradigm if the images give no clue as to what the text represents?
My concern, as I have previously stated, is that they may well be herbal pages, but have some original herbal related content and therefore not constitute a block-paradigm. This has been a slight concern in general of mine with the block-paradigm approach is that if the book is merely a reproduction of content from other texts then it seems ridiculous to have enciphered it, so I think it makes sense to assume that there is some original content. Having said that I doubt that everything is original, so I believe there is scope for the possibility of finding a block paradigm.
JKP: all the block paradigm candidates I’ve proposed so far have been in non-plant sections, for that precise reason.
Mark: if the Voynich is a collection of secrets, then it would seem unlikely to me that the author invented over 200 pages of secrets. In which case, some must surely have come from elsewhere. Hence the block paradigm search. 🙂
Nick: I concur with what you have said. As I have said, I think the block-paradigm search is a worthwhile endeavour. I am inclined towards the view that there must be some kind of block paradigm somewhere in the manuscript. However I am a little sceptical that there are many such blocks. And frankly I have no idea how easy it will be or how long it will take to find such a block; it could happen tomorrow or in 40 years.
Nick: I do wonder about the possibility of finding what I term as a constructed block-paradigm. You may well have discussed this concept. By this I mean a scenario such as where one can see how 2 non-Voynich astronomical drawings could be combined to produce a Voynich block-paradigm. So it may not be necessary to find a block paradigm as such as it may be possible to construct a viable block-paradigm from other non-Voynich sources. I hope I have made myself clear.
Mark: the first step is to understand the original ‘alpha’ state of the manuscript, because that may well reveal hitherto unknown structures within the text. For example, when you reconstruct the original order of Q9, the sun page sits next to the moon page, which may offer a block paradigm match into another 15th century document. Similarly, when you remove the additional confounding markings on the dense circular diagram on the back of the nine-rosette page, you find that it’s a magic circle with four people in, extremely similar to the other magic circle with four people right beside it. And so those two close-together magic circles might yield a block paradigm match with another 15th century document.
And we should only need a single decent qality block paradigm match to be able to work out the mapping between the two.
Nick: As you know my core research at this time is not focused on looking for a block-paradigm. However it looks like you are doing an excellent job of exploring that line of enquiry.
Obviously a difficulty with the block-paradigm search is knowing in which archives to look. Narrowing down the search zone is vital. I would think to do this knowing where the author was geographically situated and which libraries he/she would have come into contact with is important. I obviously have my own ideas which archives are most likely to contain these texts, though there is clearly the possibility that they could be anywhere.
I note the following poem describing the healing properties of thermal baths in caves that just might be a crib for one of the poems on folio 81r of the Voynich manuscript. The poem comes from page 69 of Synopsis: Eorum, Quae De Balneis. Auctore Ioan, Francisco Lombardo Neapolitano
Venetiis MDLXVI
De Hoc balneo, Alcadini leguntur versus
Ultima thermarum laudes spelunca meretur
Cuius aqua poterit simplice nemo frui
Ingenio faciente modum capit unda calorem
Sic intrabit aquas ingeniosus homo
Cuiuseunq velis perimit symptomata gutte
Hic fugit hydropisis, tussis iniqua perit
Vt galenius ait drachmas si quinque calentis
Quisquam quottidie sumere aqua
Et super, et subtus qua sunt diaphragma medetur
Rheumatos excludit quod nocet omne genus
Non salutarem continet intus aquam
Byron Deveson: that’s a nice little find, thanks! “De Hoc balneo, Alcadini leguntur versus” appears to be the title of the poem, and the last line is missing – it should be “Crypta salutarem continet intus aquam”. This also appears in Caput XLVII of Thesaurus Antiquitatim et Historiarum Italiae here – https://books.google.co.uk/books?id=1lxLWlJ7r2UC&pg=RA2-PA45&lpg=RA2-PA45&dq=%22De+Hoc+balneo,+Alcadini+leguntur+versus%22&source=bl&ots=zeMya-JCCH&sig=NEOlJd6Xa9H4oeEouY7vtTzC52Y&hl=en&sa=X&ved=2ahUKEwj32sa77ujeAhVRQcAKHZtJAfYQ6AEwAHoECAAQAQ#v=onepage&q=%22De%20Hoc%20balneo%2C%20Alcadini%20leguntur%20versus%22&f=false
Hi Nick,
“De hoc balneo leguntur Alcadini versus” translates (loosely) as “Concerning this bath, one reads the verses of Alcadino.”
Alcadino of Syracuse is credited on Italian Wikipedia as another possible author of De balneis Putoleanis, instead of Pietro of Eboli.
http://www.textmanuscripts.com/medieval/alchemy-medicine-cosmetics-recipes-60946
Recipes and Extracts on Alchemy, Medicine, Metal-Working, Cosmetics, Veterinary Science, Agriculture, Wine-making, and other subjects arranged in 520 numbered sections In Latin, manuscript on paper. Northwestern Italy, c. 1425-1450 (probably before 22 December 1438)
Nick, what piqued my attention was the dealer’s comment that the book is “The direct ancestor to sixteenth-century Books of Secrets, collections such as these appear to be still little-studied in the scholarly literature.”
“This is a diverse and well-organized recipe collection with more well more than 520 different recipes on topics ranging from alchemy and medicine to solutions for simple household problems such as fleas, designed from the outset to be used with its extensive alphabetical subject index. Its outstanding (unique?) contemporary binding adds to its interest….” sixteenth-century Books of Secrets, collections such as these appear to be still little-studied in the scholarly literature.”
The dealer states the following regarding the provenance.
“Watermark evidence, together with the script, suggests an origin in Northwestern Italy, possibly in Bergamo, Milan, Pavia or Como, between 1427 and 1447; the date, 22 December 1438, copied inside the front cover may or may not be the date when the manuscript was completed, but it does suggest the manuscript was copied before that date…….”
Nick,
I note that page 42 of Synopsis Eorum, Quae de Balneis, Aliisq; Miraculis Puteolanis Scripta Sunt, Giovanni Francesco Lombardi 1566 appears to deal with astrological and astronomical matters in relation to one of the thermal baths, “Tripergule” which may be the modern day place Tripergola close to Pozzuoli.
The association of thermal baths and astrology/astronomy suggests that the twelve line verse attributed to Eustace of Matera on page 43 might be worth looking at as a crib for the possible balnological poems in the Voynich manuscript.
Byron: thanks! The specific difficulty here is that we seem to have a poem structure of [8]/8/8/8 rather than of a multiple of 12, and almost all of the De Balneis poems are multiples of 12 lines. 🙁
However, there is at least one poem there with a line structure closer to what we are looking for, “Balneum Subcellarium”:
1 Est subcellarium lavacrum quod convenit Egris.
2 Lucida quo multum dulcis et unda manet.
3 Pondus et arborem vesice solvit · et egris
4 Provocat urinam tencio cuius obest.
5 Hac hominis varii pertolus curatur in undis;
6 Dentes gingivas in mundificando iuvat.
7 Passus in estate triteam vel cotidianam
8 Aut tepidas febres, senciat eius opem
9 Pulmonis, Iecoris vicium splenisque medetur.
10 Tussis ob hoc lavacrum pectore pulsa fugit.
11 Appetit et stomacus ista pertolus in unda:
12 Nam bene digestus redditur inde cibus.
13 Lentigo, scabies, turpis pannus faciei,
14 Hec curantur aqua quando lavantur ibi.
15 Reddit prolixos et claros ipsa capillos,
16 Et totum corpus exhylarando iuvat.
Looking at Kauffmann’s “The Baths of Pozzuoli” (p.17), he lists “Balneum Subcellarium” (“Situated on the shore of Lake Avernus, to the left of the huge bath known as the Temple of Apollo” – [2] “This and the other ‘temples’ of the Baia area are described and copiously illustrated by A Maiuri, “I Campi Flegrei (Itinerari dei Musei e Monumenti d’Italia”, 32, 1934), pp. 142 ff. and 66-74)”) as #10 in the second of Sebastiano Bartolo’s three epigraphs, covering “the baths to be found between Pozzuoli and Baia”).
Nick,
regarding the two possible poems on folio 81r and suitable material for a block paradigm, I note that the first possible poem on f81r has fifteen lines. There is a 15th Century French poem form, a Rondeau Cinquain which has 15 lines consisting of three stanzas: a quintet, a quatrain, and a sestet with a rhyme scheme as follows aabba aabR aabbaR where R is a refrain. Lines 9 and 15 are short, being a refrain consisting of a phrase taken from line one. The other lines are longer (but all of the same metrical length). I note that the last line of f81r (line 16 of the second possible poem) appears to have an additional bit tacked on at the end and I wonder if this is a refrain.
Nick,
Would I be right in thinking that what you call a ‘Block Paradigm’ is pretty much the same as what one of the chaps in Malta describes as
“Semantic similarity analysis”?
As you know linguistics isn’t my thing so I may have misunderstood. I’m going by the dictionary definition which has:
“an analysis used to indicate whether the content or meaning of a text is similar to the content of another or not”.
In the absence of a translation/decryption I don’t know quite how such a comparison could be tried out. As so often, I wish people would clearly state what they adopt as their basic premises – and why.
The other term used by the same chap in Malta – sorry, his name escapes me at present – was
Stylometric analysis.
If my dictionary definition is correct, this is an even more ambitious plan. The dictionary defines it as
” an analysis of linguistic style, used to suggest whether a text belongs to a certain author based on linguistic features”.
Wouldn’t that imply certain assumptions too? I mean how many ‘certain authors’ are they intending to test, and in how many languages? If you have a text that you know was first composed in thirteenth-century England, you might run such a test to see if it were written by one, or by another, very well known English author.. but if your initial guess is wrong, surely you’d just get ‘computer says dunno’. Or is it more sophisticated than that?
If others readers wish to reply, instead of/as well as Nick, remember to make clear whether you’re speaking to Nick, to me (D.), or to the world at large. Thank you.
Diane: the way I try to describe a block paradigm ‘match’ is more in terms of “matching the structure of an unknown ciphertext with the structure of a known text”, in order to establish a block-level equivalence between them (e.g. matching paragraphs in one to paragraphs in the other), despite not knowing what cipher system was used to encipher the ciphertext (working that out is basically Step #2). As such, it’s a bit ‘deeper’ than either semantic similarity analysis or stylometric analysis, but all are clearly trying to use related means to achieve analogous outcomes.
Nick,
Thanks. That helps me appreciate the distinction.