The Voynich Ninja

While looking into this, I came across Français et dialectes chez les auteurs belges du moyen age by JULES FELLER.
You are not allowed to view links. Register or Login to view.

My French reading comprehension is slow and unreliable, so I asked Gemini for a summary:

Quote:For Jules Feller, medieval Belgian authors were not trying to write in their local dialects (Wallon or Picard); rather, they were striving to write in French. However, because they did not have a perfect mastery of the central language, they attempted to deduce or "manufacture" French words by applying logical conversion rules or patterns of resemblance.

The creation of the word octembre is a perfect example of this phenomenon: the human mind spots a regularity in a series of neighboring words (Septembre, Novembre, Décembre) and, through a pure logic of assimilation (analogy), alters the word octobre to align it with the others.

Feller thus categorizes octembre as "false French manufactured by analogy" or part of a "hybrid language"—a form that is neither pure dialect nor correct French, but an artificial creation born out of a well-intentioned (though clumsy) effort to write correctly.

The dialects in question are Walloon and Picard. 

This confirms our suspicions that there is some cross-linguistic/dialectical tension on these month names. 

So I wonder if the temporal component ties in to the dialectical: decreased influence of the dialects as the standard language became more widespread?

Excellent find!
So rather than a rare dialect, we might be dealing with a non-native French speaker trying to write French - which gradually was replacing Latin as Lingua Franca (sic)
If we accept Taccola as source, which already pushes the VM creation date into the 1430s, there is not much time left until a significant decline of Octembre.
So what would this mean for a likely origin of a Walloon or Picard speaking (writing) month name creator from what is nowadays Belgium? Valois Burgundy?

There is a trend in the dating, but it doesn't cut off entirely. In fact, one of our top three tied entries has an "octembre" from as late as 1495. This is from Cambrai, which was aligned with the Low Countries at the time. Other highly ranked entries are also post-1450, and a bunch of them are "15th century".

It's good that we now know about this trend, but I would not build further speculation upon it yet.

More generally though, what is up with these marginalia writers? The one who wrote the German on You are not allowed to view links. Register or Login to view. seems to have been struggling with German. The one who wrote the month names apparently attempted to write standard French but seems to have been more familiar with a northern dialect. And we can't even decide if You are not allowed to view links. Register or Login to view. is a Romance or Germanic language!

It might be that the person who wrote the months' names was an older individual who had learned to use 'Octembre' in their youth and stuck with it - despite linguistic changes.

1430 is the date you get if you accept a VMs artistic inclusion of the Order of the Golden Fleece.

It's a different investigation, but what about the letter "e" as presented in the VMs month names? It's a common two-part form, but the upper right part is separate and shaped like a mini lightning bolt. This structure may not be that unusual, may even have been fashionable to an extent. And if so, should be present in the high-scoring examples.

This writer provides some examples:

You are not allowed to view links. Register or Login to view.

(31-05-2026, 09:29 PM)R. Sale Wrote: You are not allowed to view links. Register or Login to view.This writer provides some examples:

You are not allowed to view links. Register or Login to view.

Bastarda (lettre bâtarde) was most common in French-speaking areas. It was someone who learned to write that way and used some of the extra spikes even in less formal writing.

(31-05-2026, 12:15 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.While looking into this, I came across Français et dialectes chez les auteurs belges du moyen age by JULES FELLER.

Quote:For Jules Feller, medieval Belgian authors were not trying to write in their local dialects (Wallon or Picard); rather, they were striving to write in French. However, because they did not have a perfect mastery of the central language, they attempted to deduce or "manufacture" French words by applying logical conversion rules or patterns of resemblance. [...] Feller thus categorizes octembre as "false French manufactured by analogy" or part of a "hybrid language"—a form that is neither pure dialect nor correct French, but an artificial creation born out of a well-intentioned (though clumsy) effort to write correctly.

Indeed that seems to be a quite plausible explanation for why it has been so difficult to identify a single language / dialect for all the month names.   Namely, the Months Labeler (whoever wrote those names on the vellum) was trying to write in a language which he was not fluent with.

Another possibility (which may overlap with that one) is that the Labeler copied the names from some book.  That could explain also the discrepancy (that, IIRC, some have noted) between the date of the handwriting style and the date range for the "octembre" spelling.  Namely, the book where the Labeler copied the names from was written in the early 1400s, but the copying happened at a later date.

Makes sense?

All the best, --stolfi

(01-06-2026, 03:19 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.Another possibility (which may overlap with that one) is that the Labeler copied the names from some book.  That could explain also the discrepancy (that, IIRC, some have noted) between the date of the handwriting style and the date range for the "octembre" spelling.  Namely, the book where the Labeler copied the names from was written in the early 1400s, but the copying happened at a later date.

This has been a hunch i've had for a while, namely that the month names may be roughly contemporaneous with the VMS instead of being far later additions. Nowhere near enough examples have been found in enough contexts for this to be convincing yet. The octembre data is just an interesting trend that somewhat goes against expectations, but it is not suggestive that the month names must have been written earlier. There are countless reasons why Books of hours (most entries) may follow a different trend to other texts, especially considering the fact that the months were headers surrounded by illuminations, not scribbled labels. 

So it's possible, but as Koen mentioned there are examples of octembre up to the end of the 15th century. Additionally, there is 1 example of an 'o' present in june up at around 1470-90AD, and critically it is being used in a table, not in a formal calendar. July is the biggest question mark as we've only ever encountered 3 examples of 'oi' (not even a lone 'o'), never with the use of 'll' instead of 'ng', and never after the 14th century.

Okay. An absolute ton of entries have been added now, and I got to the end of the list I was sourcing from. The source is JONAS, the "Directory of French and Occitan texts and books (850-1550)". Most manuscripts/books in their database are given variables or descriptions to show the contents of those manuscripts. By chance while looking at an entry from gallica, I accidentally clicked on one of the variables for that manuscript (You are not allowed to view links. Register or Login to view.). 

Doing so listed basically every single manuscript in their database that was recorded to contain a french calendar, across any/all seperate libraries worldwide. Each one tends to provide links to descriptions or digital content (when available). Over the last few days, I have checked every single link on that page and added every one that I could, around 150 in the end I think. There are so so many that are not yet digitised, so this page may one day contain more examples as more and more things get digitised.

It may be off-topic, but there may be other variables on this site that could be used in the same way (zodiac?, astrology?, illustrations?, language?) for lots of other areas of VMS research so it's well worth looking into.

-----------------------------------------

A super interesting (even if not the highest scorer) entry was this one. You are not allowed to view links. Register or Login to view.

mars, auril/auris, may/mais, joing(et)/iugnet, auoust/auost, septembre, octoiure, nouembre, decembre

On f4v, the months are discussed, with iugnet and iulet. In the calendar on f7r, it switches to joing(et) and juillet. JONAS lists the number of hands as 1, and the supposed origin for this manuscript is Italy, with a potential author being "Pozzolino da Basilea". 

There were definitely many others of note, which I will probably post at some point, but right now I can't formulate enough about them until I take another look.

That Marciana entry is interesting indeed, and I found it quite confusing. JONAS refers to the 2008 catalogue work of Bisson. As far as I can tell, this is not easily accessible online, but Gemini was able to read it (it wouldn't give me a direct quote since the text is not freely available). It summarizes:

Quote:Bisson outlines that Fr. Z. 2 stands out sharply from the rest of the Fondo Francese Antico because its language is not Franco-Italian (the typical hybrid franco-veneto found in the library's epic poems). Instead, he classifies the linguistic layer of the narrative text as a highly conservative, authentic representation of northern Old French with a distinct Picard dialectal coloring. 

When discussing the two-folio calendar (ff. 4r–5r), Bisson addresses the dialect in a highly specific way. He isolates the month names and liturgical markers, noting that they are spelled using rigid Picard orthography (such as Genvier and Juignet).

However, Bisson adds a critical caveat to this dialectal profile: he notes that while the language points cleanly to a Picard archetype (traditionally associated with the ecclesiastical centers of Tournai or Arras), the calendar itself is content-sparse. It omits the critical regional saints and major local red-letter feasts that would typically allow a paleographer to pin a Picard calendar down to a precise diocese or town.

Bisson concludes that the text represents an interesting paradox: a manuscript that is linguistically pure Picard North-French, but materially and historically 100% North-Italian courtly art, commissioned to satisfy the aristocratic tastes of the Gonzaga family.

Take this with the necessary precautions about AI reliability. But Eggy, if this is correct, then it looks like you were able to single out a linguistically Picard manuscript from an Italian collection.

(31-05-2026, 12:15 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.While looking into this, I came across Français et dialectes chez les auteurs belges du moyen age by JULES FELLER.
You are not allowed to view links. Register or Login to view.

[...]

Quote:For Jules Feller, medieval Belgian authors were not trying to write in their local dialects (Wallon or Picard); rather, they were striving to write in French. However, because they did not have a perfect mastery of the central language, they attempted to deduce or "manufacture" French words by applying logical conversion rules or patterns of resemblance.

The creation of the word octembre is a perfect example of this phenomenon: [...] .

The dialects in question are Walloon and Picard. 

[...]

Very promising!, but of course now what about the 'b' in Aberil?
I once read that this could have been used in Luxemburg (which was quite a bit larger then), but I have not seen that appearing in discussions here.
Still interesting, as there would be a geographical overlap.