After reading a post by Bluetooes about charms, I took a closer look at medieval incantations and related texts. The more I read, the more I realised that the formal characteristics of such texts could help explain some of the persistent problems we encounter in the Voynich Manuscript.
Statistical analyses have shown quite convincingly that the Voynich Manuscript does not behave like an encrypted information text in the classical sense. However, this raises an obvious question: how reliable are these analyses if a large part of the VMS consists of highly formulaic incantatory litanies?
Cianci (You are not allowed to view links.
Register or
Login to view.) says that these incantations have a perfectly coherent rhetorical structure. And The Pervinca charm (Clm 7021) appears in the medical section. This means that herbal + charm is a historically documented genre, not a special case.
These charms from the 14th–15th centuries show a combination of repetition, phonologically stable formulas, almost purely sound-magical sequences (voces magicae) and herbal-ritual embedding.
In other words, such incantations correspond significantly more closely to the statistical properties of the Voynich manuscript than medical prose or recipe literature could, and even more so than a hoax.
Let's take a closer look: the charms hypothesis explains several previously contradictory levels of the VMS at a stroke, without introducing any additional auxiliary assumptions!
For example, repetition with minimal variations would no longer be noise but part of the incantation.
The "Fix – marix – morix – vix" "gently fix mastic and myrrh" from f116 is an almost ideal example of this. Semantically loose, phonetically very close together, but nevertheless formally unambiguous. Language does not serve as language, but tips over into sounds and approaches a melody. And here, too, there are small shifts in individual letters, which we also know from the VMS.
And this is not an isolated case. This is exactly what we see in incantations, spells, litanies and apotropaic sayings: not information transfer, but performative effectiveness through rhythm, repetition and echo, as well as linguistic phonetic similarity.
Further examples:
Komt ge van God sprekt
komt ge van den duivel, vertrekt
If you come from God, speak!
If you come from the devil, leave (Dutch incantation)
Eloim, Essaim,
frugativi et appelativi!
Eloim, Essaim – those who drive away and those who call (names/formulas)."
Eloim, Elohim, Elohim, Essaim Elohim, the Hebrew word for ‘God/deity’
Essaim : God, [Lord of Hosts]
Heilig, Heilig, Heilig ist der Gott Sabaoth u. durch die allerschröklichsten 'Worte.
Soab, Sother, Emanuel, Aden, Amathon, Mathey, Adonai,
Eel, Eli, Eloy, Zoag, Dios, Anath,Tafa, Uabo, Tetragramaton, Nglay,
Josua, Jonas,
Calpie, Calphos, So erscheine mir N. sanftmüt in menschlicher Gestalt u. erfülle was ich begehre
Holy, Holy, Holy is the God Sabaoth and through the most terrible “words”. Soab, Sother, Emanuel, Aden, Amathon, Mathey, Adonai, Eel, Eli, Eloy, Zoag, Dios, Anath, Tafa, Uabo, Tetragramaton, Nglay, Joshua, Jonas, Calpie, Calphos, So appear to me N. meekly in human form and fulfil what I desire.
(German incantation: You are not allowed to view links.
Register or
Login to view.)
This is close to some lines of the Voynich text, if one looks at the frequency of repetitions.
The problem with the Voynich manuscript, that words are repeated and/or only one letter is changed, could easily be explained by this – they would then be incantations.
If we then assume a greatly reduced phonetic Bavarian and liturgical or formulaic Latin mix, as we see it unencrypted on 116, many other peculiarities of the VMS also fit into the pattern. (See Stolfis' approach to Chinese, based in part on possible monosyllabicity, which exist in Bavarian Speech too You are not allowed to view links.
Register or
Login to view.)
Such hybrid forms are particularly well documented in the southern German Alpine region of the late Middle Ages: Latin provides the typical sacred part, while the dialect ensures proximity to the individual.
If one then assumes that these incantations were written ‘by ear’, precisely because the linguistic and phonetic characteristics of the incantation give it its actual power, this explains not only the unusual orthography, but also the extreme positional binding and stability of sound clusters with simultaneous semantic vagueness.
But that would also explain the question: Why would someone encrypt a recipe text? When it comes to pure recipes, encryption makes no sense, as I have already noted several times. But if it is, so to speak, ‘ultra-secret, almost esoteric knowledge (esoteric in the sense that something is intended only for an inner, usually small circle of initiates or particularly knowledgeable or ’chosen" people) – then it was almost obligatory to encrypt it, because it could be danger, too.
In short: Much of what we see in Voynich,
repetitions,
sound shifts,
formulaic structure, words that look as if they were derived from the previous words (marix / morix vix fix) fits almost perfectly into the structure of an incantation.
Unfortunately, it could then be extremely complicated to decipher something like this. Because incantations can also become very incomprehensible: Here is an excerpt You are not allowed to view links.
Register or
Login to view.
De Voces Magicae
ysaac bapsiul
afilo anaba floch bilo ylo sandoch az
achel topharie fan habet hyy barachaist
ochebal trach flamaul moloch adach frach
aiam ustram bucema adonay eley elenist
gorabraxio machatan hemon segein ge
mas iesu
"I couldn't get any further with the translation, so I turned to ChatGPT

:
"
You can identify individual anchors:
• adonay – clearly the Hebrew name of God, Adonai (‘the Lord’).
• iesu – Jesus.
• moloch – a traditional demon/god name, used here as a word of power.
• barachaist – most likely derived from beracha (Hebrew for ‘blessing’).
• aiam / eley – phonetically similar to ehyeh / el, also a name of God.
The rest – gorabraxio, machatan, bapsiul, flamaul, etc. – are ritualistic artificial names. Some are reminiscent of well-known magical names (e.g. Abraxas cluster), while others are purely sound constructions. Their function is authority, rhythm, intensification."
I cannot judge this; I think I can recognise several other German word fragments (segein = segen / trach = tragen, floch = flach/ flechten ,etc ), but as a whole text it actually makes little sense.
In academia, ‘Voces Magicae’ is described as Christian magic formulas with Hebrew elements: inspired by Hebrew names of God and prayers, but not encrypted or normal Hebrew sentences – rather, sounds are imitated. (And that might be the reason why a group of researchers came up with Hebrew?).
And what does that all tell us?
When examining Voynich with this background in mind, it is clear that we have typical sound shifts, word repetitions and, at the end, a striking number of identical ending sequences. All of this could be an indication that these are often line-by-line or longer incantations.
But does that mean that VMS could consist solely of incantations? Probably not. I have now read up on it, and most charms also contain instructions and other information.
But even then, if a certain part of Voynichese were incantations, this could
influence any statistical evaluations to a greater or lesser extent, distorting them to such an extent that they would not yield any meaningful results. Especially if a lot of ‘Voces Magicae’ were hidden in it....