A comparison of f116v with medieval charms

Abstract: In this article I analyse the final page of the Voynich Manuscript, denoted f116v. I examine a possible reason for the unnatural positioning of the text on this page, and look at the writing to identify temporal trends in the inking of the glyphs. A brief mention of possible languages are mentioned. I then compare the structure and format of the text with a series of medieval charms to try to identify a semiotic comparison with other such medieval works. Finally I examine the illustrations on this page to identify a possible correlation with the supposed intent of the text.

Description of the page: (From the interlinear description) This page is blank except for four lines of text at the top (numbered “1” to “4”) and some drawings at the top left corner. The text is rougly justified against the left margin, and ragged on the right. Line 1 is flush against the top edge of the vellum, lines 2-4 lie about 1/2 inch below it, with normal interline spacing. Except for two Voynichese words at the beginning of line 4, the text is written in a peculiar script (“Michitonese”) that seems to be intermediate between ordinary (Latin) alphabet and the VMs script. The handwriting is irregular and not very readable. The letters in line 0 is somewhat smaller but apparently in the  same handwriting as the rest. The area next to the upper left corner is hevily stained and wrinkled. There is a large hole in the vellum, 2-3mm wide, near the upper left corner, about 2cm from the edges. Lines 2 and 3 of the text start right next to the hole. In the dark region between the hole and the left edge, roughly aligned with text line #3, there is the drawing of a four-legged animal, resembling a dog with round ears, short hais, and a fat but pointed tail. The animal is facing left, with the snout right against the vellum’s edge. Below the animal is a female nymph (with breasts), naked except for some simple hat. She is facing seems to be seated on a sloping surface, right at the edge of the dark area, with arms stretched sideways and down. . [JasonDavies link]


Whilst examining f116r in an abortive attempt to decrypt it, it got to the point where I sat back in my swivel chair and actually started to think about it. I then got down some reference books, and started to do some old fashioned research.

To recap:

There are four lines of text in mainly standard alphabet on this page (with two Voynichese words) in two sections.

  • The first line consists of four or five words running along the very top of the page – the whole of this section has been water damaged.
  • Underneath are three lines in a paragraph. The first two lines have crosses between almost every word – I have come to suspect that these come from a folkloric tradition of writing “magic spells” or charms which included crosses within the words as a visual reminder to cross oneself whilst reading aloud.
  • The third line starts with two words of Voychinese and then goes off into standard Roman letters again.
  • All four lines are so far unintelligible (although there have been translation attempts).
  • There are also some doodles down the side of the page to the left, separated from the main text by a tear in the vellum, and a single three letter word.

I’m going to split this page into the following sections which are indicated in the article by “section (x)”:

  1. The very first line along the top
  2. The second three lines in which words are separated by crosses
  3. The images off on the left, along with a word like “lab”

I will furthermore indicate the four lines of text using the notation [1] for line one along the top, [2] for the next line reading +an…, etc.

The Michtonese on f116r (colour adjusted)
A screenshot of f116v concentrating on the text (colour adjusted to alter contrast for easier reading). Note jagged top of page above first line of text.

The first question that struck me: why write so close to the top of the page?

Positioning of the text – why it is where it is

Well, one answer would be to conserve space. But we haven’t seen evidence of this anywhere else in the manuscript, and the text on this page does not go all the way to the right hand margin, so space doesn’t seem to be at a premium; there is a whole page to write on, so why start so high? And how come the loop of the tall glyphs are so natural, there is no indication that the scribe was concerned about hitting the edge of the page? In other words, the handwriting is more fluid than you would expect when your glyphs are hitting the edge of the page. The natural solution is to assume that the writing was written before the rough cut of the top of the page was made, when the scribe had plenty of room above his inking to make his glyphs in a natural manner.

Well, we see water (or some other transparent liquid) damage (that’s the brown stain at the top of the sheet), and a rip, on this page. I suspect that the first line was written, and then the water spill occurred. We know the main spill was on this side, as the recto side of the sheet does not present the same staining (although there is some blurring of the ink on the recto side which could have been caused by the spillage, I will not enter into that here).

It was mopped up, the page may well have been trimmed at the top (maybe it started to swell?) – right to the top of the letters (the cut is still way above the writing on the recto side) – and then when the scribe returned to the page, he started writing afresh underneath the water damage. So we end up with a line right at the top of the page, followed by an unnaturally large space and a short paragraph. It is even possible that the first line stops abruptly because of the water spillage and is hence unfinished.


The text was originally inked onto the page when the page was larger than it is now. After inking the first line, a liquid spillage occurred which necessitated the trimming of the top of the page. The scribe later returned to the page to continue writing but decided to start a new line underneath the water damage.

An examination of the inking process

It struck me when reading the high quality scans now available that the structure of the glyphs… changes. The first and fourth lines are similar, as are the middle two, but there are modifications to the way the glyphs are inked. Let’s look at this a bit more closely. Let’s look at two major letters on the first line, second word which I shall call L and B. The cross-over of the L is pronounced, half way up, and the ligatures are all curved – look closely at the B:

f116r line1And compare with the same looking glyphs on line [2]:f116r line2

The L looks far more like a P (and it may be – but on the middle two lines this type of glyph only appears twice and the second time it’s far different from this example). The ligatures are harsh, with plenty of right angles. Look at the B, it’s clearly the same glyph but inked in with harsh right angles on the ligatures.

But if we look at the ‘a’ all the way through the text, it’s always clearly the same, as it is with several other of the more common glyphs.

In short and without delving more into the abstracts of the matter, what we’re looking at is the same hand writing with different quills. Look line [4] – the vowels are identical, but the longer looping glyphs are being marked with a quill of different sharpness, or a different quill altogether. The scribe may have come back to the page at a later date, or it may be as simple as he wrote section (2) after having mopped up the water spill and trimmed the top of the page.


The text is all written by the same scribe, but using different quills for different sections. This may indicate that there were breaks of time between the writing of different parts of the text.

A tentative discovery of Middle High German

In fact, look at the very two last words in line [4]: gaf mich, which are clearly inked with a refreshed and differently shaped quill, followed by a circle with four dots above it. (Note: I assume the first word reads gaf and not gas because of an apparent half ligature in the f ). Now, my German is sadly non-existent, but looking this up I find that both words are Middle High German:

  • gaf(fen): to stare at curiously, rubberneck (gaff is the imperative)
  • Mich: me (accusative of ich).

So the last two words in this charm could read I stare [read?] with curiosity, which after all is what most of us do with this book. Followed by a circle with four dots above it – this has been described as a “falling stone” and I discuss this in a separate section later on. (The middle high German translation is correspondent with the style of lettering on this page, but I won’t go into that here – it has been noted elsewhere). It’s entirely possible that the rest of this sentence has a meaning, but only for someone versed in middle High German, and that person isn’t me, sadly. However, a number of people have identified M.H. German words in the first and last lines (which I discuss later on), including of course the famous “pox leber” (goats liver) in the first two words.

So the whole thing was written by the same person, but with gaps of time in between the different sections.

Let’s concentrate on section 2. I’ve dropped any previous thoughts of this being in code, and my current theory is that it’s a magic charm, similar to the hundreds or thousands of other examples we can see from medieval works. I went off to find examples. It’s not Latin, so it must be in the vernacular, most likely something Germanic. Johannes Albus has translated it as a mixture of German and (very poor) Latin which reads something like:

[1]Billy goat´s liver for wet rot. [2]At the membrane you gave oil, then you bring a lot of the much(?) wax, in a [3]fixed mixture: 9 hands full, 9 morsels (from) the only just double mature [4]… … (two ciphered words), squash it into a paste, then take goat´s milk.

But the Latin of [2] and [3] is atrocious – without going into it here, it is badly misspelt (plenty of letters had to be added) and the tenses are all wrong. It just doesn’t fit. This doesn’t mean that he wasn’t on the right track though – I’m suggesting that we have lines [1] and [4] in vernacular, let us assume some type of German, and lines [2] and [3] (with their respective crosses) as a charm which doesn’t have to make sense, but which can be twisted by the modern imagination to make sense, as we can see in hundreds of other documented examples throughout the medieval and Renaissance West.


Lines [1] and [4] are entirely or partly in a vernacular, namely Middle High German. Lines [2] and [3] are “voce magica” based on the same base language.

A comparison with medieval charms

Now, Kieckhefer has identified three types of medieval Christian charms that were written in the vernacular: prayers, blessings and adjurations (or exorcisms). Most of these charms come from the north of Europe, and tend to be somewhat misunderstood Christian rites that lay people twisted around to meet their own needs and inserted their own local folklore into. Earlier examples (pre-1300’s) often came directly from monks, but the Church started to clamp down on this sort of simony. I shan’t delve into “magic spells” which would be a fourth group – these weren’t really part of the Catholic faith. The first three types invoke Christian power, whereas spells try to control Christian power to force something to happen, which is a whole different kettle of fish – for a start, it was invoking daemons which the Church wasn’t keen on, so you start getting into assumed witchcraft and the like. I’m going to concentrate mainly on the healing or lifestyle charms, which is what I think we’re seeing here.

Now, a common element in the early Christian works developed by monks (whose charms are usually based on religious works and are intelligible) is to insert crosses into the text to remind the reader to cross themselves at this point. Remember that until the Renaissance reading aloud was far more common than reading silently, and hence this type of semiotic was vital in reminding the reader to carry out an essential part of the Catholic rite. A random example of this (which I just found) can be found in a section of MS Harley 2558, which deals with “four recipes for coughs, a charm and notes on various herbs by 15th century English physician Thomas Fayreford.” The charm (right hand side of the above link) rambles on until it reaches the end, at which point we see Amen +++, a reminder to the chanter (the charm is designed to be read aloud over or by the patient) to cross themselves three times before God. Tom uses the same structure in other charms in his book, which is a nice example of the English genre of the time.

Kieckhefer in Magic in the Middle Ages gives a number of examples like this, and all contain the cross at specific intervals. I shan’t bother copying them all out here.

However, these charms are somewhat different from what we are seeing in the Voynich. They are intelligible, they are either in Latin or standard vernacular, and the crosses they contain are in specific portions of the text, not every single word, usually around the word Amen or similar. Let us remember that these are charms designed to be remembered and used afresh, often in conjunction with the giving of some medicine (or the collection of the ingredients for the medicine); typically, we find they have the short ritual formula (such as in nomine + pater+ filii et spiritus sanctus) before the incantation proper ending by an affirmation of effectiveness, and accompanied by directions for application to the patient. The f116v text is simple, apparently nonsense words and is mixed in with Voynichese words.

An interesting element is that the official charms as given by healers or monks tend to only use the crosses when invoking the power of Christ or using religious names. However, when we see the same charms in use by lay people (on amulets or tablets for example) we see the number of crosses in the text multiplying. This I attribute to the same effect that makes people take more medicine than the doctor prescribes: if a little is good, a lot must be great!

So there is a variant on the charm which fits what we are seeing – folkloric nonsensical abjurations.

The most famous of these that I can quote would be what we today know as “abracadabra”. This has been traced back to the Romans of the third century, and although its origins are lost, the pleasing word has often been found in charms written on amulets. In De Medicina Praecepta Saluberrima, a third century Roman tract, a charm is given for sufferers of malaria to write this word in a triangle like so and keep with them:

A – B – R – A – C – A – D – A – B – R – A
A – B – R – A – C – A – D – A – B – R
A – B – R – A – C – A – D – A – B
A – B – R – A – C – A – D – A
A – B – R – A – C – A – D
A – B – R – A – C – A
A – B – R – A – C
A – B – R – A
A – B – R
A – B

This sort of a charm is very common – nonsensical words forming a pleasing pattern. It is voce magica, sounds that have a ritual sense [purpose] but don’t make sense. Ritual incantations to sooth and infuse power into a body, medicine or amulet.

Another is the Roman-era Sator Arepo pattern. And Wellcome’s 188* Leechcraft has a number of medieval English Latin charms that delve into nonsense at some point – unfortunately, he didn’t seem to recognise them and so didn’t include either the crosses or the mumbo-jumbo Latin phrases into his translations, although he did leave them in his (somewhat sparse) original transcriptions or photos.

These charms are usually Northern European in origin (at least after the 16th century, I suppose the Inquisition stamped it out in the south – I can’t find any suggestion of this existing amongst curanderos down here in Spain) and tend to involve nonsense words that often sound vaguely Latinate. The important thing here is the formula, not the words.

Bengt af Klintberg and Ritwa Herjufsdotter in their Swedish work on folk charms against snakebites Jungfru Maria gick i gräset:Trollformler mot ormbet (which I found via Incantatio Journal, Vol 1) discuss some 50 folk charms (most after the 17th century, admittedly) which mainly consist of this exact formula: nonsense words which vaguely rhyme, strung together into a charm. You sing the charm over the snakebite as the poor chap writhes in agony on the ground, foaming at the mouth (or whatever the symptoms of being bitten by a Swedish snake are), crossing yourself as you invoke the power of the Lord.

The formula works as follows (for healing charms, anyway). The principle word is chosen upon and then strung along to produce a rhyme, which often sounds similar to a Latin prayer that people would be used to from church. There are two ways this can go: either a whole series of rhyming words are sung out, or the principle word is repeated through a series of transformations until it ends up as a little word, to symbolise the breaking down of the illness or poison that is supposed to be affecting the patient. The metrics often are similar to those of the Lord’s Prayer or other liturgy’s sung at Mass, which reassure the patient of the healing power of the Lord. Crosses are inserted at keypoints in the text to serve as a visual reminder to “cross oneself here”.

A lovely example of this comes from Cambridge MS R. 14.30, f. 146r (in Occitan):

A trastotas febres: Escrieu aiso en pargamin verge […]: + on lona onu oni one +onu onus oni one onus

‘For all fevers. Write thus on virgin parchment [and place on the altar beneath the chalice until three masses have been sung over it] +on lona onu oni one +onu onus oni one onus [and then attach it to the patient’s neck]

The same manuscripts give the following recipe for a cure for a woman suffering from excessive menstrual bleeding:

A curamen de sanc de las femenas que per natura lo perdo. Escrieu aquestas caractas ins en .i. tauleta que sia d’estang e lia la sobrel ventre: apoono

‘Write these letters on a small tablet made of lead and tie it on her abdomen: apoono.’

Here’s another one:


Here we have first a frame of liturgical expressions to set the ceremonious atmosphere: “In the name of the Father etc” (et spiritus santus), loaded with intelligible, ‘normal’ power words taken from Christian mythology, which have a bearing on the situation of childbirth: LAZARUS, because of his resurrection compared with childbirth here, COME FORTH and CHRIST CALLS YOU. Thirdly, there is a series of nonsense words, which, however, may betray a wordplay on REX = king, triggered by the spell CHRISTUS REGNAT: “Christ rules”. Like late antique charms medieval Latin charms display a variety of linguistic forms 1) nonsensical sounds, 2) Latin verse, 3)strings of powerful names, 4) narrative themes, 5) performatives of adjuration and conjuration and prescriptives & 6) the semiotics of the cross. (See the analysis by Olsan, o.c. 124-139 (and another very complex example on p. 137) and B.K. Halpern & J.M. Foley, “The Power of the Word: Healing Charms as an Oral Genre”, Journal of American Folklore.)

Now: imagine that we discovered the “small tablet made of lead” inscribed with apoono in a tomb. What would we make of it today without knowing the original recipe from Cambridge MS R 14.30? And yet it made perfect sense at the time.

Note: half way through researching this, my attention was drawn towards the blog of J.K.Petersen who has a similar explanation for this page. He has an excellent example from a manuscript called Der Neusohler Cato, which isn’t quite what we see here – his example is of a word that is reduced to a single symbol through transformation, in the process of which the “illness” associated with the charm is symbolically reduced to naught. This could apply to line [2] which gets progressively shorter and shorter until it ends up as a single glyph, but line [3] has much more of a repetitive chant to it.

So far I have given examples of charm recipes, examples that have come down to us via medieval herbals and medicine books. They come with their introductions, their instructions and the associated assurances that they work.

So what are we seeing in the Voynich? We’re seeing an example of a charm in action. Someone has taken a charm and used it in the book! We’re not seeing all the associated introductions and explanations, because this is the charm in action, used by a lay person.

See the second line?
Line [3] has a repetitive rhyming structure to it with the same glyphs repeated four or five times (see my notes in the image), possibly based on Christian liturgy familiar from Mass

This could be something as simple as the owner of the book trying to invoke the mysterious power of the book by adding his magic charm to the very last page.


Lines [2] and [3] are a medieval charm in action. They are “voce magica” and whilst may have an underlying theme, they do not make readable sense for a modern reader.

The illustrations

So we now see something sensible emerging from this page!

The first line is in M.H. German and could well read something along the lines of Billy goat´s liver for wet rot (as per Albus’s reading).

We then have a specific charm on two lines, with additional crosses for additional strength.

Finally, we see two Voynichese words followed by some more M.H. German words, and what is either an “o” or a doodle of a falling stone.

Is it an illustration or a glyph?

Certainly there are examples of “o’s” with several dots above them in other manuscripts, where the dots represent distinct types of diacritics. But I don’t think that is what we are seeing here. For a start, there are no diacritics in the text (although the i’s are dotted). Secondly, the o is out by itself, with a space separating it from the preceding word, and what language has a single o with a diacritic being used as a word? None that I know of. Nor do I know of any word that ends in an o with a diacritic (leaving aside accents, but if accents were involved we’d see them elsewhere in the text). No, I suspect this is an illustration, a doodle, of something specific, and the scribe later looked at it, thought it looked too much like a letter, and hence added the four dots above it to give it some meaning. I am going to call it a stone, for want of a better word.

Let us look at this stone. I suspect it is linked to the doodles on the left, amongst which is the word “lab”. Anton Alipov has saved me having to investigate this too much by pointing out that “lab” is Middle High German for rennet-bag, or the fourth stomach of a ruminant. The animal below is a goat (note the hoofed feet); above we see the stomach with a stone in it.

Speculation And what prized medieval medicine came from the stomach of a ruminant? A bezoar, which is a stone believed to cure any poison. Which we see at the end of the fourth line and in the doodle!

René Zandbergen has pointed out in the past the similarity of the woman on this page with the “dead woman” elsewhere in the VM. Is she appearing on this page as a reminder of the need for an antidote against the poison that killed her? Is she the real life subject of this charm?


The final page of the Voynich is now clear. It is a charm written by someone in Middle High German against a specific illness.

By why the Voynichese?

Well, the EVA transcription is  oror (or aror) sheey and it only appears twice in the main manuscript (René again). Once on f76v where it appears next to a woman holding a big red object (it could be a stomach – someone suggested a giant vulva, but we won’t go there) and again on f104v where the only illustrations are stars. Make of this as you will. The writing on this line, as I have mentioned above, whilst by the same scribe, appears to be with a different quill which could indicate it was written at a later point in time. Maybe the scribe associated this illustration on f76v with the object of his charm and assumed (or knew) that oror sheey was the name of the object? Pure speculation this bit I’m afraid.

The actual translation of the first and fourth lines really needs to be passed over to someone knowledgeable in Middle High German than I. Taking random words and putting them into dictionaries isn’t really going to get me anywhere.


  • The page appears to be a medicine charm against some specific illness written as voce magica.
  • The first and fourth line appears to be written in Middle High German, as is the single word amongst the left hand illustrations.
  • The middle two lines are a folk-charm.
  • The illustrations appear to reference a bezoar, the gallstone found inside the stomachs of ruminants, which were highly prized as universal antidotes to poison and medicine. There is a faint possibility that this is linked to the “dead woman” illustration in the middle of the main manuscript.

11 thoughts to “A comparison of f116v with medieval charms”

  1. Hi David,

    I agree that with many similar examples in contemporary literature, the “charm” hypothesis is the best match for explanation of lines 1 and 2. I would be cautious nonetheless in linking all elements of f116v together. It is quite probable that they are not interrelated.

    1. Hi Anton,
      I’ll admit I got a bit over-excited with the connection between the illustrations and the text, and that this is the most speculative part.
      However, I do feel that all four lines of text were written by the same hand, even if there is a difference in time between them (as I explain above in the text)

      1. Yes I did not dispute that. I believe there has been some analysis in the past showing that all the “Michitonese” stuff is written in the same ink as the main body of the VMS.

        I have some considerations upon f116v (no sensational discoveries or conclusions so far, just consistent reasoning which may appear useful), but I ever postpone publishing those because either I get some new idea or something does not fit into the picture. But I promised elsewhere to do that in spring, so I’ll have to keep my promise 🙂 The “spell” concept and the German context are central in those, so it’s good that it is supported by other authors independently – that suggests that we are going in the right direction. You know, Voynich studies have seen so many theories that any new theory is met quite skeptically by the community. But the “German” theory is not a classic Voynich “theory”, because it provides no solutions per se – it is rather an attempt to narrow the context.

        To illustrate the possible absence of interrelation between different portions of f116v, I just published part of my research today: http://athenaea.net/index.php?id=57

        1. Hi Anton,
          A most interesting read. Sadly, my German is not up to evaluating the linguistic arguments (and I have no intention of starting to learn it, just yet!).
          As I’ve said elsewhere, I dropped the Castillan theory and personally feel a German / central European origin is the most likely. Nor do I think it is encoded text of any form.
          The key most likely lies with the last line of text and the first couple of words of line [2] and [3], have you had any insights there?
          The McCrone report, by the way, simply establishes that this ink is similar to the rest of the black ink used for the main body of text. The coordinates in the report locating the sample site appear to be erroneous (more on the VMS mailing list, one of the threads is http://voynich.ninja/#id:12825 )

          1. I think my German is not better than yours, e.g. I can’t read freely without a dictionary. Let alone that I am not an expert in MHD. That’s why I say that input from professionals of the field would be most valuable, not simply from German-speaking people. In the absence of such input, the resource http://woerterbuchnetz.de/ is most helpful. It comprises several dictionaries of German of various times and places, and provides a unified search across them.

            Regarding lines 1-3 (if we count from 0) – yes, as I said above I have some considerations, I will try to sum them up and publish asap. The problem (with the VMS) is that each time when you develop a certain hypothesis, some thing still does not fit.

        1. Your first post was very interesting – I linked to it from my original article.
          I wasn’t aware of your later discovery, thanks for the update.

  2. I agree with J.K. Petersen and you about the magical words used on the last page. But not only do the voces magicae and the so called charakteres appear there but everywhere else in the manuscript.

Leave a Reply

Your email address will not be published. Required fields are marked *