The Voynich Ninja

[Edit KG: I changed the title of this thread so Geoffrey can discuss his Middle English theory here]

Recently in the You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. threads, I have tentatively raised the idea of a Germanic interpretation of certain possible readings of the Voynich manuscript text. To be clear, I certainly do not mean that the Voynich ms text could represent standard German, Dutch, etc., even in their late medieval forms as Middle High German, Middle Dutch, etc. In fact, in my tentative hypothesis I have made reference to an extremely divergent dialect, Wymysorys, strongly influenced by Polish in the area of southern Poland where it is spoken, and moreover exhibiting a curious mixture of High / Central German (like Standard German), Istvaeonic (Dutch / Flemish), and Low German (Saxon) features that even make it difficult to classify within the branches of the West Germanic languages. 

But in this thread I want to focus not on the arguments in favor of this latest idea of mine, but rather on its weakest and most problematic points. Among a number of these, I want to begin with the matter of word-final sounds, letters, and characters or units. This is a huge issue for any theory about the interpretation of the Voynich ms text: The number of distinct final units of Voynich vords, however they may be parsed and classified, is strikingly small and limited. The vast majority of Voynich vords end in EVA [y], [n], [l], or [r], with a significant but much smaller number of them ending in EVA [s], [m], [o], or [d]. EVA [n] is virtually exclusively vord-final. 

To justify a correspondence of such a distribution of vord-final units with any reasonably possible known language, I must attribute to EVA [y] a special nature, either as a "null character/unit", or at most as an ambiguous laryngeal "H", which may variously reflect an /h/ sound, a lenition or fricativization of the adjacent sound (as in Irish), possibly a vowel sound in certain environments, and/or possibly still a null unit. This analysis allows, for example, the vord-final sequence EVA [-ky] to possibly represent the unit [k] as the actual final unit, possibly modified in some way by the [y] unit, or possibly not modified at all if [y] is a pure null.

In this way, one can develop a system with [l] and [r] as consonants that occur very frequently in word-final position, and reasonably frequently elsewhere (more so for [l] than for [r]). The special nature of [y] allows a possible modest frequency of word-final occurrence for other consonants, but not a great frequency.

So the question then becomes, which is the consonant EVA [l] that is frequent in both final and non-final position, which is the consonant EVA [r] that is frequent in final position and modestly frequent elsewhere, and how does one account for the positional occurrence of all the other consonants?

In the Germanic interpretation, I propose that EVA [l] is "s", frequent in both final and non-final position, and that EVA [r] is indeed simply "r", frequent in final position (syllable-final as well as word-final) and only modestly frequent in other positions. This seems reasonable for Germanic languages.

But the problematic issue, then, is how to account for all of the other occurrences of word-final consonants? In particular, how does one account for and represent word-final "n" and "t"?

The only explanation I can think of is that EVA [ky], for example, represents "tH", and EVA [qoky], for example, represents "nH", and in these sequences "H" is merely a null unit. However, even this solution only provides for about 600 words with final "t" (or possibly "d") in the entire Voynich ms text, and only less than 200 words with final "n". 

Clearly there is no way that a Standard German text of tens of thousands of words will only have less than 200 words with final "n". Only 600 words with final "t" is a less extreme restriction, but still requires careful justification and explanation. These are the most problematic points, prima facie, for any Germanic theory of the Voynich ms text.

Now it does so happen that Standard German is rather extreme among Germanic languages for its very high frequency of final "n". Dutch / Flemish, Saxon, and even other High German dialects drop the final "n" in many places where Standard German retains it. Thus, a divergent dialect such as Wymysorys, an East Central High German dialect with an unusually significant presence of Flemish and/or Saxon features, is at least a much better candidate to be the underlying language of the ms than standard German itself. 

In short, any theory that the language of the Voynich ms text is Germanic will first of all have to account for the representation and frequency of all word-final consonants, in particular "s", "r", "n", and "t", and also explain how they are represented in non-final position and with what frequency. 

Geoffrey

There are strong points against any language simple substitution theory, and they have been constantly raised for I don't know how long.

One fresh observation is that certain vords (or maybe lines) appear as if sorted. For example, I You are not allowed to view links. Register or Login to view. first six quires (haven't yet got enough time to continue) of the MS, looking at the paragraphs beginning with k. Now, 91% of these paragraphs have their first three lines "sorted" by the first letter - that is, the sequence of line-initial glyphs complies to the deterministic ordered set.

(10-04-2021, 12:07 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.There are strong points against any language simple substitution theory, and they have been constantly raised for I don't know how long.

One fresh observation is that certain vords (or maybe lines) appear as if sorted. For example, I You are not allowed to view links. Register or Login to view. first six quires (haven't yet got enough time to continue) of the MS, looking at the paragraphs beginning with k. Now, 91% of these paragraphs have their first three lines "sorted" by the first letter - that is, the sequence of line-initial glyphs complies to the deterministic ordered set.

Anton,
The points you raise in the linked "Regaining the lost order" thread, and in the several additional threads linked in your initial post there, are certainly important to consider and investigate, as you and others have done and are doing. However, it appears that the main focus of these ideas concerns the order, sorting, re-sorting, etc., of the existing vords and lines. But David Jackson made a very valid and critical point in that thread:

(28-03-2021, 12:56 PM)davidjackson Wrote: You are not allowed to view links. Register or Login to view.But even if you did this, you wouldn't alter the statistics of the individual words, and nobody has been able to break down the words.

Every point I made in my initial post here concerns the internal structure of individual vords in the ms text, not the order or sorting of the vords or the lines. You can reorder and re-sort the vords and lines in any way that you like, and it will not change a thing about the issue of the extremely restricted set of distinct final units of Voynich vords, the vast majority of which end only in EVA [y], [n], [l], or [r]. No matter how these existing vords and lines are reordered, re-sorted, and rearranged, they will still have these same final units. And it will still be quite difficult to account for the necessary occurrence of an actual language's words that end with, for example, "s", "r", and "n". I have pointed out how this will be a particular problem for any Germanic language, but such issues will also apply to just about any reasonable candidate language for the text of the Voynich ms.

That's true, but when any new "language solution" is laid before us, it usually assumes the natural left-to-right (perhaps right-to-left for Arabic or Hebrew) top-down text flow. I do not remember solutions which would assume re-ordering.

Fair enough, although I will observe that often one of the criticisms of such "language solutions", whether they be in Greek, Polish, Flemish, German, or what have you, is that the word order of the so-called "word salad" is wrong, scrambled, unidiomatic, not valid syntax for that language, and the like. If this were the main issue, then a vord re-ordering step could potentially resolve such problems and be part of the solution. However, at least in the case of my own previous proposed language solutions, I do not get the impression that this was the primary problematic issue -- that is to say, even if I had had the freedom to choose any vord re-ordering I like, my strong impression is that no such rearrangement could have satisfied and resolved the criticisms leveled at my proposed solutions. Serious issues with the logically necessary grammatical forms of the individual words/vords, as well as with the degrees of freedom of the phonetic interpretation of the Voynichese characters/units as sounds/letters in the target language, would still have remained. These are both fundamentally vord-level issues, and no amount of re-ordering of the vords would resolve such problems related to the internal structure of the vords.

Example: I could have added a vord re-ordering step to my interpretation of "bieh cu ipH Htilie[g]" (the [g] is still in EVA): "Transpose the positions of the first and last vords" in this four-vord phrase. Then I could argue that the phrase can now be interpreted to mean "Otilie(?) zuip(t) bie[r]", Dutch for "Otilie swigs beer". So there you have it now, Anton, a "language solution" which would assume re-ordering of the vords.

However, I am afraid that this little refinement will not convince Rene, Koen, and others that we are actually reading Dutch here. Despite the now apparently correct vord order and word order for Dutch, several difficulties remain, and they concern issues of the internal structure of individual vords. The biggest problem is that I have simply conjured up the final "-t" in "zuipt" out of whole cloth, yet it is precisely this letter that makes the whole phrase grammatical Dutch rather than just a "word salad" with a name, a verb root, and a related noun root. The next problems concern the final letters of "bier" and "Otilie(?)". 

Of course reading "H" as "O" in "Otilie(?)" is a big leap as well, all the more so because the same character (EVA [y]) apparently stands for the critical "-t" at the end of "zuipt". Now I could write a tortured but formally logical process even to account for this -- along the lines of "when "H" appears at the end of one vord and the beginning of the next vord, the letter after the second "H" (which is "t" here) is added at the place of the first "H", and the second "H" becomes a vowel (here "O")". This convoluted but formally logical cryptographic process solves the two biggest problems with this interpretation of the vords in this phrase. But again, I doubt I could convince anyone that this is the actual process by which the Voynich ms text was created. 

I note in passing that the problems with the final letters/characters/units of the vords and words in this phrase confirm the points I made in the original post in this thread about the problem of word-final letters and vord-final units in any Germanic interpretation of the Voynich ms text.

Yes I see what you mean - in fact, I agree with your point.

I recall Marco once wondering, if people have all these language theories for the Voynich ms, why doesn't anyone simply try to read it as English?

The simple answer is that very few people can really read Middle English anymore, least of all English speakers, except for a couple weeks in secondary school and a couple weeks in college when we pretend to be able to read my namesake Chaucer's Canterbury Tales with the help of very heavily annotated student editions where every other word has a note telling us that shoures means "showers", not "shores" (that would rather be strondes); soote means "sweet", not "soot"; swich means "such", not "switch"; flour means "flower"; and seke means "seek" but seeke rather means "sick". Most modern English speakers want no part of any of that monstrously difficult and confusing spelling, so why would they seke swich language, that they find more seeke than soote, among the floures of the Voynich ms?

Among other things, for modern English speakers, the Great Vowel Shift of 1400-1700 really means that we are "orphans" with regard to medieval languages. At the time of the Voynich ms, no one spoke a language with vowels pronounced the way we do in modern English. So all medieval languages are foreign to us modern English speakers, in a way that does not apply so extremely to speakers of other European languages.

But since this is a Germanic thread, and I happened to stumble across some interesting Middle English words in my research of my latest ideas about possible Germanic readings and interpretations of Voynich characters, units, and vords...well, Marco's wish is hereby granted.

I was thinking about and looking into the famously repetitive vord [qokeedy] [qokeedy] [qokeedy] throughout the bathing women section and in a couple places in the recipes section, and my latest Germanic reading system gave me "neecH neecH" for this vord. Well, it so happens that neece is a Middle English spelling of "niece", an alternative form of nece, and it can also mean a woman family member more generally. Moreover, this word nece happens to also be a Middle English alternative form of nyce, which in that period could mean "foolish, simple, naive"; or "scared, weak, lazy"; or "fussy" (14th c.); or "wanton, sinful" (14th c.); or "cunning" (15th c.); or "fragile, delicate" (15th c.); or "strange, odd, bizarre" (15th c.).

So...by this interpretation, [qokeedy] [qokeedy] [qokeedy] could be a Middle English play on words, "nece, nece neeces", meaning "foolish, bizarre women" or really any other combination of the various possible meanings of Middle English nece as adjectives describing the neeces. 

Even better, according to this system, the very similar Voynich vords [qokedy], [qokchdy], and [qokchedy] could possibly represent alternative Middle English spellings of these words: nec(e), nyc(e) / nic(e), niec(e). Hmm, the latter two spellings, although the most familiar to us in modern English, actually would have been more current in Middle French than in Middle English in this time period. Either way, these various spellings of these two words could account for the roughly 750 occurrences of these [qoke/ched-] vord forms in the ms text, the vast majority of them in the bathing women and recipes sections. (A search for [qok*d] on You are not allowed to view links. Register or Login to view. is a quick and simple way to visualize the occurrences of all of these forms in the ms.)

As Marco would say, the English interpretation is no worse or less logical than any other claimed reading of the Voynich ms.

[Note: both Middle English words come from Old French, they are the origin of "niece" and "nice" in modern English, and the latter's original meaning was actually "ignorant", being derived from Latin nescius, as in ne scire, "not know"!]

I can read Middle English. It was still largely a Germanic language in the Middle Ages (just switch the brain into Norse and the "odd" words make sense). The influence of French vocabulary wasn't as strong then as it is now (modern English has a huge infusion of French words).

A Middle English perspective doesn't make the VMS look any more comprehensible, as far as I can see.

Yes, JKP, I surely include an experienced medievalist such as yourself in those "few people who can really read Middle English anymore"! I just meant that for the average modern monolingual English speaker, Middle English is just as much a foreign language as any other. I would argue, more so than Middle Dutch for a modern Dutch speaker, Middle High German for a modern German speaker, and likewise with French, Italian, Spanish, etc. For an interesting perspective, check out these Memrise courses on Middle English: You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. . It is really striking to me how unusual the pronunciations of Middle English "I", "thou", "he", "she", "we" sound to my modern English ear. And in terms of vocabulary, "underyete", "menske", "reed", "swage", ... even the average college educated English speaker will be totally lost just with these first few words. This is truly a foreign language for modern English speakers.

I think I am mostly joking about the idea of Middle English as the language of the VMS. Nevertheless, it would not be completely absurd, since we know for example that Wycliffe had a great influence on Jan Hus and his followers in Bohemia in that time period: it is believed that students brought Wycliffe's heretical doctrines and ideas from England back to Prague by 1399. And Wycliffe was famous for his vernacular Bible in Middle English. Surely such students would have become familiar with Chaucer's works in England at that time too. So it's not absurd to imagine students in Bohemia in the early 15th century with both a knowledge of and an interest in the Middle English language and some very significant works of literature in the language.

As far as I can tell, the only so-called "Middle English Voynich theory" was mostly based on supposed numerical gematria readings of letters, and the only serious English theory of the VMS attributed it to Anthony Ascham, a mid-16th century writer and thus Early Modern English, not Middle English. But that was an old theory, proposed long before radiocarbon dating identified the parchment as early 15th century.

(11-04-2021, 11:22 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.I would argue, more so than Middle Dutch for a modern Dutch speaker, Middle High German for a modern German speaker, and likewise with French, Italian, Spanish, etc.

This is true, I remember my "history of English" professor at university saying as much. English beame harder to recognize than other Germanic and Romance languages, for a number of reasons. The fact that some originally Germanic words were gradually replaced by Romance alternatives, the Great Vowel Shift...

That is not to say that I can just pick up a Middle Dutch text and understand every word (or get the correct meaning of the words I recognize). But for people who only speak English, it takes objectively more practice to read Middle English. Calling it a foreign language like any other is certainly a bridge too far though. I bet that an English speaker with no experience can still read more Middle English than Finnish