Scinderatio Fonorum

BNF Latin 13026, 9thC, Picardie, France, f11r

As I was researching a future post, I came across a set of techniques at the crossroads of grammar, mnemonics, cryptography and madness called scinderatio fonorum (the breaking/cutting of words).  It lies at the core of the wild “grammar” devised by Virgilius Maro Grammaticus in the 7thC, which Babino (2015) describes as “poetic-philosophical encryption”. Could Virgilius’ works have inspired the creator(s) of the Voynich manuscript?

The works of Virgilius were copied, either in part or completely, in several manuscripts between the 9th and 11th centuries, and he remained quite well known, quoted and incorporated into grammatical compilations, especially in Ireland and in France, but also Germany and Italy, until the later middle ages, after which his work was only rediscovered in the late 19th C.

Virgilius’ method is somewhat akin to the Cisiojani Nick Pelling has explored since 2009. If the authors of the Cisiojani had decided to make things fiendishly complicated for their readers, they would have written them following Virgilius’ teachings. Whereas the Cisiojani abbreviate words by keeping only the first syllable, Virgilius’ scinderationes can split the words into syllables, disperse these across the line of text, fill in the rest of the line with meaningless words… unless they condense text beyond recognition. There can be many refinements, among which using homonyms and writing words or parts of words backwards are only a best-case scenario.

Virgilius lays out these methods throughout his two books, the Epitomae and the Epistolae. There were probably more, but these are now lost. At first glance, they appear to be perfectly normal grammatical treatises. But upon closer inspection, it becomes apparent that he is actually distilling the principles of a method for disguising text throughout his books, spiked with a host of fanciful, eccentric linguistic inventions that seem to almost be a satire of the grammatical texts of his time.

Virgilius writes “maro-Roma-mare” to express his name : perhaps referencing the place he came from, but this is hard to ascertain. Elsewhere he writes that “Rome is Toulouse”. Was he Irish? It is almost certain that he lived there (Bracken, 2002), but was he born in Toulouse, as the Abbey of Fleury claimed in the 11thC? In Spain? Was he Jewish? His origins are debated.  He continues to employ this model in various places, and Dolezalova (2009) explains that throughout Virgilius’ text, “two variants of a word often existed side by side”, something that surely rings a bell for those familiar with the Voynich Manuscript. But we’re just getting started.

Mnemonics are thrown in as well. Thus, the biblical book of Ecclesiastes is referenced as “una hasta plena oculis” (a spear full of eyes) : oculis being vaguely homophonous with “Ecclesi”, while hasta corresponds to the “astes” part.  As Dolezalova notes, without the corresponding gloss, such images are mostly undecipherable.

Why devise such complex methods? Virgilius gives three reasons:

  1. “So that we establish the acuteness of perception of our students in searching and discovering these obscure things”.
  2. “Because of the ornamentation and construction of the speech”.
  3. “So that mystical mysteries and those which should be apparent only to the knowing ones, would not be by chance easily found by the inferior and the stupid, so that, according to an ancient saying, the swine would not tread on precious stones”.

At the end of the Epistolae, he again gives reasons, this time admitting that “elegance” was less of a motivation than the necessity “to hide these mysteries”. Clearly then, Virgilius’ concern was akin to those that motivate encryption, and there can be no doubt about the intentionality of the obfuscation of plain text that results from applying his “grammatical” method.

Roger (1905, pp 111-126) alternatively describes Virgilius as “ingenious” but “bizarre” and “insufferably pedantic”. Vivien Law sees his work as a “plea for diversity” and innovation in grammatical studies, but also as a puzzle that may conceal mystical elements. Dolezalova prefers to conclude that “making things difficult’ may have been a very conscious mnemonic process. By not being offered the message in the most simple and straight-forward way, the audience was encouraged to think, to wonder, to interpret. Once they found the solution, it stuck in their minds better because of the efforts they had to make”. In my opinion, Virgilius’ method is a parody of grammatical studies which hides a game, a challenge to those who think themselves learned, and a toolkit for those who have true “wisdom”.

Scinderatio Fonorum 101

Want to learn how to make any plain text completely unintelligible for generations to come? Here are the basics of how to properly scinderate your fonorum.  (All below adapted from Rochus’ 1931 review of Tardi’s translation and especially from Law, 1995, pp.83-97):

Virgilius claimed there are twelve kinds of Latin (by which he means there are twelve ways of writing Latin). Choose (one or several!) from the list below:

  1. Usitata: The one commonly used
  2. Assena: The one used by notaries. To represent a whole word it uses a single letter, and also represents it using signs [these are Tyronian notes].
  3. Semedia: It is not completely out of use, but not completely in use either. One writes mota gelus to signify mons altus
  4. Numerosa: It has its own names for numbers, like nim for “one”…
  5. Metrofia: This is the language of intelligence: dicantabat means “beginning”…
  6. Lumbrosa: The one that lengthens. For a single word in common language, it uses a whole verse. Gabitariuum bresin galsiste ion means “to read”…
  7. Sincolla: The one that shortens: A whole phrase is concentrated into a single word. So gears means “mend your ways and love good things”…
  8. Belsavia: The one that switches. It changes the cases of nouns and the tenses of verbs.
  9. Presina: The comprehensive one: A single word can be used to replace various words in the common language. Sur can mean “plain”, “mutilated”, “sword” or “river”.
  10. Militana: The one with multiple meanings. To replace a single word one can use several alternatives. For example, gammon, sualin, rabath… mean “running”.
  11. Spela: The lowest one. It deals only with purely material things. Gariga means a crane…
  12. Affina/Polema: The one for speaking of the highest things. Affla means “soul”, spiridon means “spirit”…

But wait, there’s more! Now that you’ve mixed up your plain text using your chosen blend of “latin”, you can also proceed with the actual scinderatio of your text, which can take three basic forms.

  1. “The first consists in cutting up the verses. When Cato* says: “Mare oceanum classes quod longue saepe turbatur simul navigant“, he should have said: ” Mare oceanum saepe turbatur, classes quod longae simul navigant“.
  2. The second consists in cutting up words or syllables, as Lucan* did: ge, ues, ro, trum, quando, tum, a, fec, om, ni, libet, aeuo. And it is read like this:  quandolibet uestrum gero omni aeou affectum.
  3. The third consists in cutting up the letters themselves. Cutting up letters seems pointless, however it is used by the most subtle writers, those who chisel their style. Obviously, short words are easier to cut than long ones. Cicero* says: RRR, SS, PP, MM, NT, EE, OO, A, V, I. Which is read: Spes Romanorum perit… Emilius the rhetorician* elegantly says: SSS, SSS, SSS, SS, PP, NNNN, GGGG, RR, MM, CT, TD, CC, CC, CC, CC, CC, CC, III, II, VVVV, VVVV, EE, AE, EEEE, EE. Here is the key: Sapiens sapientiae sanguinem sugens sanguissuga uenarum recte uocandus est.”

“All parts of discourse can thus be cut up.  Nouns; when you mean regmen, you can write germen… Pronouns can also be cut up: for is you write si. Verbs can also be cut up, like nodo for dono, and gelo for lego. Adverbs too, according to this example from Hilarius*: locutis summus rhei for rehi.”

Obviously, you may also indulge in dropping random syllables, but also in adding new ones within words, or even, in using only the ending of a verb instead of the whole word. Or you can scatter a word’s syllables randomly thoughout a text, like in this example from Cicero*: “omne visum ab homine CU delectabile natura stigante PI ipsi indita, ex qui non potest TUR cupidine carrere” (the scrambled word, cupitur, originally belongs after homine).

You may also scatter the letters of a word throughout a text, like Varro* does here with the pateremini which normally should go between viri and dolorem:


Bonus points: Using Virgilius’ unique “etymology”, you can also borrow characters from other alphabets to stand for words. For example, he claims that the word res is derived from the Hebrew resh, therefore you can use ר to represent res if you feel your text is still too easily accessible to the unrefined masses.

Virgilius does warn his students to apply his methods with care and not to take things too far, or it might not be possible for even a sophisticated reader to recover the original meaning: “for no problem can be solved that has not been correctly posed”.

… And there you have it, in a nutshell!

* Oh, those references I marked with a star? They don’t actually exist: Virgilius either flat out invented the authors he quotes, or made up the quotes he attributed to actual authors!


No Stars upon Thars

Capture d_écran (804)

One of my favorite things about the Voynich Manuscript is that no matter how many times I’ve looked at it, every time I do there’s always something I hadn’t noticed before that jumps out at me.

In the Zodiac section, each of the signs is surrounded by circles of nymphs who hold stars… Except, some of them don’t. In this post, I’ll try to stay away from speculation and from further references to the Sneetches, and just present a brief census of those nymphs who don’t have stars.

Throughout Pisces, both Aries, and both Taurus pages, each nymph has a star, either connected by a tether (tail) or holding it directly. Things start to go awry in Gemini, where three nymphs are depicted without stars: the male-looking nymph pictured above in the outer circle, and two female-looking nymphs: one in the inner circle and the other in the outer one, both of which appear to be also missing an arm.


In two of these three cases, it appears that whoever applied the yellow ink tried to make up for this by awkwardly adding a roughly star-shaped blob of yellow ink, although this is much fainter for the one in the header image.

In Cancer, two nymphs in the second and outer circles are also each missing both a star and an arm.


Everything is back to normal in Libra, but in Leo, almost indiscernible due to the fold in the parchment and severe fading, we find another such nymph (I had to tweak this image slightly to make the lines a little clearer): No arm, no star.

Capture d_écran (805)

In Virgo, one nymph in the inner circle is again missing both her star and her left arm.

Capture d_écran (799)

In the inner circle of Scorpio, we find another instance of a starless, one-armed nymph, which the yellow painter has again attempted to remedy, this time also adding the tail on the star to connect to the nymph.

Capture d_écran (803)

Finally, in Sagittarius’ outer circle, one nymph has no star and no arm. I had doubts about the nymph behind her too, but although her arm and star are faded and covered by a green blob of paint, their outline is still definitely visible, whereas for the next nymph, the parchment was clearly never marked.

Capture d_écran (798)

EDIT: As OutsiTer pointed out in the comments, there is another nymph with an arm but no star in Sagittarius, and here the star (with tether) has been added by the yellow painter:


And that’s all. A total of only 10 out of 244 nymphs in the Zodiac section don’t have stars. Out of these 10, 8 are also missing an arm.

Initially I was inclined to search for some significance in the absence of stars. But the fact that along with the star, the arm is also missing most of the time, as well as the fact that on several occasions the yellow painter has attempted to correct this by adding stars back in later, leads me to believe that that their lack of stars is possibly an oversight by the illustrator.

In the first five pages of the Zodiac section, the nymphs are more elaborate, each drawn with various tubs and clothing, and there is still some of this in Gemini, whereas afterwards the illustrator mostly adopts a simplified, more generic style for the procession of nymphs that takes place in the second half of the Zodiac. Whether or not this transition has meaning is uncertain, although I tend to think it probably does, due to the occasional reappearance of nymphs with tubs/tubes and/or elaborate headdresses and clothing afterwards. However we can imagine that the artist grew tired of the tedious, repetitive process of drawing nymph after nymph and star after star. Illustrating the second half of the Zodiac section must have been a rather monotonous task, and I can easily envision him muttering some variation of these medieval scribes’ complaints as he completed the section.

This is not to say that the stars and arms of these nymphs are necessarily devoid of meaning. On the contrary, in this section the illustrator seems to make a point to differentiate between tailed and un-tailed stars, and even the yellow painter knows when to add a tail to a star or not when he tries to correct the illustrator’s omissions. The fact that he knows the difference is one of the reasons why I believe it is likely that the yellow painter was an integral part of the Voynich’s original creation and understood the subject matter.

If the missing stars are accidental, the results of this census might shed some light on the illustrator’s process. In my opinion, from Gemini onward, the fact that it is not just the stars but also the arms that are missing seems to indicate that unlike the more complex nymphs in the previous pages, the nymphs and stars on these pages may have been drawn in two passes: in a first pass, the body (I would guess this may have been in order to secure correct placement, to ensure that they would all fit in their circle while leaving enough room for the labels), and then in a second pass the artist went around adding each nymph’s left arm and star, forgetting one every now and then. It may even have been nymph->label->arm and star.

It seems to me that if each nymph had been drawn as a whole, left arm included, in a single pass, the arm and star would probably not have been forgotten.

But none of this is really certain, and the possibility that the absence of arms/stars is significant cannot be ruled out.