The Voynich Ninja

A lipogram or a lipogrammatic text is a piece of writing in which the author deliberately avoids the use of a certain letter (or possibly letters, although this is not typical), using only words that do not contain that particular letter (or letters).

The concept goes all the way back to the Ancient Greek author Lasus of Hermione in the 6th century B.C., who wrote poems using only words that did not contain the letter sigma. This concept has reappeared in literature from time to time over the course of the three millennia since then. It would not be an unknown or unfamiliar concept to an early 15th century European author or student.

If taken to the extreme, lipogrammatic text could possibly explain some of the extremely unusual statistical properties of the Voynich manuscript text. Naturally it would likely require the avoidance of more than just one letter to achieve anything close to the extremely low entropy and conditional entropy values found in the Voynich ms text. The purpose of most historical literary lipograms was not to lower the entropy values of their texts, and I doubt the statistical analysis of actual historical literary examples of lipogrammatic text would find extremely low entropy values. But if the author of the Voynich ms carried this process to a certain extreme, it could be one part of the explanation for the statistical properties of the text that we observe.

For example, if an English writer wrote a lipogrammatic text using only words that do not contain the letter "a", very common words such as "and", "a", "that", "are", "was", "as", etc., would have to be absent from the text. (Full disclosure: This idea occurred to me while considering this very possibility for a Middle English text.) If the Voynich ms text is indeed lipogrammatic in this way, then we would have to consider not only the effect of the lack of a particular letter or letters on statistical properties of the text, but also the effect on the grammatical structure imposed by the necessity of avoiding, for example, such normally essential function words as "and", "a", "are", "was", etc., throughout an entire text.

Geoffrey C.

Just one more outrageous idea to add to this line of thought:

Imagine a (Middle) English lipogrammatic text without the letters...ABC !

This would actually also explain EVA [qokeedy] = "neece" ('niece'), [qokedy] = "nece" ('nice', or in Middle English 'foolish, naive; wanton, sinful; cunning; strange, odd, bizarre'), since without the "c", one would have to use "s", which I am already assigning to EVA [d].

Further, the character combination that would logically represent "b" could thus rather represent "f", or "v", which are otherwise difficult to include within the correspondence table according to my hypothesis.

Geoffrey S.

Amazing fact about a possible English lipogram without "ABC":

English number names do not contain any of the letters "ABC" at all until the word "thousand".

A more or less standard Middle English dialect that used oon, not an, for "one", and thousend for "thousand", if the very large number names such as "milliard" and "billion" were not in general use in the early 15th century, could have had none of the letters "ABC" in any of its number names at all !

(An aside: People whose surnames begin with "Z" must love this "no ABC" theory of the Voynich manuscript. Revenge on alphabetical order!  )

Geoffrey S.

That's interesting. In the Bennett's book I read about some guy who wrote a whole novel altogether without the letter "e", but I never knew that the idea dates back to centuries BC. Are there any historic examples where functional words would be omitted?

(15-04-2021, 05:13 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.That's interesting. In the Bennett's book I read about some guy who wrote a whole novel altogether without the letter "e", but I never knew that the idea dates back to centuries BC. Are there any historic examples where functional words would be omitted?

The late 12th and early 13th century French poet Petrus Riga wrote a verse commentary on the Bible (in Latin of course) titled Aurora, which has been called the verse Bible of the Middle Ages. Each canto of it is followed by a summary in lipogrammatic verse, the first without A, the second without B, the third without C, and so on through the Latin alphabet. You can read the lipogrammatic verses You are not allowed to view links. Register or Login to view.: the lipograms can be found on pp. 8-13 of this pdf file, which are pp. 32-42 of the work, Fragmenta ex Aurora.

(15-04-2021, 05:43 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.

(15-04-2021, 05:13 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.That's interesting. In the Bennett's book I read about some guy who wrote a whole novel altogether without the letter "e", but I never knew that the idea dates back to centuries BC. Are there any historic examples where functional words would be omitted?

The late 12th and early 13th century French poet Petrus Riga wrote a verse commentary on the Bible (in Latin of course) titled Aurora, which has been called the verse Bible of the Middle Ages. Each canto of it is followed by a summary in lipogrammatic verse, the first without A, the second without B, the third without C, and so on through the Latin alphabet. You can read the lipogrammatic verses You are not allowed to view links. Register or Login to view.: the lipograms can be found on pp. 8-13 of this pdf file, which are pp. 32-42 of the work, Fragmenta ex Aurora.

I have attached a MS Word file of the text of the most interesting of Petrus Riga's Aurora lipogrammatic verses, in my view: the ones written without using A, B, C, E, I, O, Q, and V respectively. (Ok, the one without B is not that interesting, but I must include it after proposing a hypothesis of lipogrammatic text without ABC.) Hopefully this format will make it more convenient for purposes of searching and performing statistical analysis of the text. I apologize for any typographical errors that may have crept into my transcription work; hopefully the number of them is minimal enough that it will not significantly affect any statistical analysis of the text.

Here are a few of my notes and comments about these medieval Latin lipogrammatic verses:

"Sine A" is 10 lines long. It does contain the ligature "æ" twice in the line "Ecclesiæ signum fuit hæc de corpore Christi". 

"Sine C" is 18 lines long. But it contains forms of "Christus" five times, and the editor explains this in a footnote as follows: "C. in Ch. aut saltem in voce Christus a littera C. ordinaria diversa habetur, scribiturque plerumque in codicibus veteribus mss. Ch. in Christus per Græcum X. Quamvis vero hæc vox excusari possit, tamen in aliis versibus auctor aut scriba adversus institutum pugnavit quo littera C. abesse debebat."
Further, in two other places in the "Sine C" verses, the text contains the words "crucemque" and "construxit", and the editor adds the following explanatory footnotes respectively: "Fidemque rectius ut C careat vocabulum." and "Ad C. tollendum potuisset scribi : Qui turrim Nimroth exstruxit, vel : Nemrothus turrim qui struxit."

"Sine E" is 12 lines long. Unlike the "Sine A" and "Sine O" verses, here the author does not use either of the ligatures "æ" or "œ". 
It is notable that the exclusion of the letter "e" in Latin precludes the use of both the word "et" and the suffix "-que", the standard ways to express the basic conjunction meaning "and". 
However, the text of this "Sine E" verse contains the word "et" once, and the editor adds the following explanatory footnote: "Rectius : ac."

"Sine I" and "Sine O" are 22 lines and 28 lines long respectively, suggesting that it is significantly harder to not use "A" or "E" in Latin text than it is to not use "I" or "O". 
However, the author does still use the ligature "œ" once in the "Sine O" verse, in the word "cœlesti". 

I find "Sine Q" very interesting, since here the author may not use a whole host of basic Latin function words such as "-que", forms of "quis", "quid", "qui", etc. 
(Only once in the 26 lines of the "Sine Q" verses did the author allow a single "-que" suffix to sneak in, in the line "Ecclesiam mulier signat Christumque propheta :". The editor does not comment on this either.) 
I sense a "choppier" feel to the syntax of the "Sine Q" verses due to the lack of these "q-" function words. This verse also includes "dialogue" such as "Clamantes : Crucifige Jesum ! crucifige !"

Finally, there is no "Sine U" verse! Moreover, the "Sine V" verse contains the letter "u" in the suffix "-que" three times, as well as in the words "profuit" and "Num"! Perhaps the author was just tired of the exercise at this point in the alphabet, and just didn't make the effort to write a verse without the vowel or glide "u", since the official standard Latin alphabet of the time considered "U" to be merely an allograph of the actual letter of the alphabet "V". Maybe the author was not up for making the same effort to avoid all "-qu-" words and forms that he had made in the "Sine Q" verses. I note that "Sine V" is only 10 lines long, and even those 10 lines contain the 5 occurrences of the letter "u" already noted. (By the way, I have great sympathy for Petrus Riga's plight, and I intend these comments to justify his compositional decisions rather than to criticize him for them.)

Perhaps the "Sine Q" verses may thus be seen as the high point of these lipogrammatic verses of Petrus Riga. Therefore I will present these 26 lines here in full, and others may evaluate these verses for yourselves:

"Igneus Heliam rapit ad sublimia currus :
Non ope non curru fertur ad astra Deus.
Hospes inops mulier Heliam pascit, et ille
Pascit eam : magis hæc pascitur , ille minus,
Non huic defecit oleum farrisve medulla,
Nec minui poterat ille vel ille cibus.
Ecclesiam mulier signat Christumque propheta :
Hunc cibat hunc reficit moribus illa bonis.
Est Christi corpus et chrismatis unctio victus :
Hinc cibat Ecclesiam gratia larga Dei.
Iste cibus semper expenditur; iste redundat
Semper; nec minui sive perire potest.
Sumptus in altari panis manet integer; illum
Semper sumis ; eum sumere semper habes.
Ascendat calvus, pueri clamant Helyseo ;
Sed necat ursus eos, devorat ille simul.
Jure notat calvus Christum : calvaria nempe
Dicitur esse loci pars, ubi Christus obit.
Ursi sunt illi duo : Titus, Vespasianus :
Jerusalem cives hic secat, ille necat.
Clamantes : Crucifige Jesum ! crucifige ! necantur,
Et patrum culpas tota propago luit.
Das famulo baculum, tradas Helisæe ministro :
Nil agitur surgit te veniente puer;
Non hominem salvat præmissæ littera legis :
Cœpimus adventu vivere, Christe, tuo."

One more significant fact about You are not allowed to view links. Register or Login to view., the author of these lipogrammatic verses:

He was an influence on You are not allowed to view links. Register or Login to view., the 14th and early 15th century (1330-1408) English poet and personal friend of Geoffrey Chaucer:

From the first link: "See Robert P. Miller, Chaucer: Sources and Backgrounds (1977), p. 205; Beichner."

From the second link: "MacaulayYou are not allowed to view links. Register or Login to view.:xxx-xxxiii and other critics have observed that he [John Gower] must have spent considerable time reading the Bible, You are not allowed to view links. Register or Login to view., You are not allowed to view links. Register or Login to view., You are not allowed to view links. Register or Login to view., You are not allowed to view links. Register or Login to view., You are not allowed to view links. Register or Login to view., You are not allowed to view links. Register or Login to view., and others.You are not allowed to view links. Register or Login to view."

Sources for the latter statement:

"G.C. Macaulay (ed.). You are not allowed to view links. Register or Login to view. (PDF). The Complete Works of John Gower, Vol 4 The Latin Works. p. vii–xxx."

"George L. Hamilton (1912). You are not allowed to view links. Register or Login to view.. Modern Philology. University of Chicago Press. 9 (Vol. 9, No. 3 (January 1912)): 323–346. You are not allowed to view links. Register or Login to view.:You are not allowed to view links. Register or Login to view.. You are not allowed to view links. Register or Login to view. You are not allowed to view links. Register or Login to view.."

(15-04-2021, 05:13 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.That's interesting. In the Bennett's book I read about some guy who wrote a whole novel altogether without the letter "e", but I never knew that the idea dates back to centuries BC. Are there any historic examples where functional words would be omitted?

For anyone interested, there have been at least two: Gadsby by  Ernest Vincent Wright and La Disparition by Georges Perec.   La Disparition has been translated into English, and the translation (A Void) also omits 'e'.

Yes it was Wright whom I meant.

Another idea in this line of thought: There is a logical reason why an author's attempt to write an extreme version of a lipogrammatic text, such as one without using any of the letters ABC, could have compelled him to invent his own new writing system in order to do so:

It is extremely difficult to compose a lipogram, any lipogram, without making a mistake and allowing the prohibited letter or letters to slip by one's notice and sneak into the text. I recall that Ernest Vincent Wright talked about having to literally tie down and disable the "E" key on his typewriter in order to write Gadsby without making a mistake and letting a word with "e" slip into his novel. Full disclosure: last night I tried my hand at writing my own English extreme lipogrammatic text without using any of the letters ABC. I can do it, but it is a brutally intense, exhausting process. I managed to write 953 (nine hundred fifty-three, see, no ABC's there!) words of such text. The point is, I kept making mistakes, catching them with the convenient Ctrl-F "search and find" function, and going back and fixing them. You saw above in my analysis of Petrus Riga's lipogrammatic verses that he made occasional mistakes as well. The editor of the edition I read and linked to above caught a few of them and offered fixes of his own, but it seems both author and editor missed "Christumque" in the "Sine Q" verses, which should have been somehow fixable by rewriting the phrase and clause with "et" instead of "-que", albeit the literary quality, such as it was, would suffer I imagine. The point is, it is extremely difficult to compose a lipogram without making mistakes.

Well, the author of the Voynich manuscript did not have the convenient Ctrl-F "search and find" function. They didn't have typewriter keys they could tie down and disable, or printing press letters they could remove from the character set when printing, as an extra final safeguard against allowing the prohibited letter or letters to slip into the text. They couldn't erase mistakes as easily as we can now. From an extreme "purist" lipogram perspective, even the appearance of the prohibited letter or letters on the manuscript page, even if corrected, might have been considered unacceptable. 

The author was not satisfied with such flawed lipograms, so he found a solution: Invent his own writing system, and leave out any letters for A, B, and C entirely! Now he could write his lipogram and not have to worry about an overlooked a, b, or c mistakenly slipping into his text and spoiling his beautiful lipogram. And thus the writing system of the Voynich manuscript text was born.

Such a person may well have also been the kind of person who simply could not stand to spoil the appearance of his manuscript and text by making any corrections to his writing at all. And so he didn't.

I can tell you from my own extreme "no ABC" English lipogram writing experience that eventually one decides to allow words that normally contain a prohibited letter, but one writes them without the letter, with a substitute letter, or in some abbreviated or other different way. That could explain, for example, [qokedy qokeedy] as "nese neese" standing for "nece neece" (modern "nice niece"). It could also explain other perhaps strangely abbreviated words.